diff --git a/platform.txt b/platform.txt
index 1b92b64efa8..03ad7846273 100644
--- a/platform.txt
+++ b/platform.txt
@@ -23,7 +23,7 @@ compiler.prefix={build.tarch}-{build.target}-elf-
 #
 # ESP32 Support Start
 #
-compiler.cpreprocessor.flags.esp32=-DHAVE_CONFIG_H -DMBEDTLS_CONFIG_FILE="mbedtls/esp_config.h" -DUNITY_INCLUDE_CONFIG_H -DWITH_POSIX -D_GNU_SOURCE -DIDF_VER="v4.4-dev-3235-g3e370c4296" -DESP_PLATFORM   "-I{compiler.sdk.path}/include/config" "-I{compiler.sdk.path}/include/newlib/platform_include" "-I{compiler.sdk.path}/include/freertos/include" "-I{compiler.sdk.path}/include/freertos/include/esp_additions/freertos" "-I{compiler.sdk.path}/include/freertos/port/xtensa/include" "-I{compiler.sdk.path}/include/freertos/include/esp_additions" "-I{compiler.sdk.path}/include/esp_hw_support/include" "-I{compiler.sdk.path}/include/esp_hw_support/include/soc" "-I{compiler.sdk.path}/include/esp_hw_support/include/soc/esp32" "-I{compiler.sdk.path}/include/esp_hw_support/port/esp32" "-I{compiler.sdk.path}/include/heap/include" "-I{compiler.sdk.path}/include/log/include" "-I{compiler.sdk.path}/include/lwip/include/apps" "-I{compiler.sdk.path}/include/lwip/include/apps/sntp" "-I{compiler.sdk.path}/include/lwip/lwip/src/include" "-I{compiler.sdk.path}/include/lwip/port/esp32/include" "-I{compiler.sdk.path}/include/lwip/port/esp32/include/arch" "-I{compiler.sdk.path}/include/soc/include" "-I{compiler.sdk.path}/include/soc/esp32" "-I{compiler.sdk.path}/include/soc/esp32/include" "-I{compiler.sdk.path}/include/hal/esp32/include" "-I{compiler.sdk.path}/include/hal/include" "-I{compiler.sdk.path}/include/hal/platform_port/include" "-I{compiler.sdk.path}/include/esp_rom/include" "-I{compiler.sdk.path}/include/esp_rom/include/esp32" "-I{compiler.sdk.path}/include/esp_rom/esp32" "-I{compiler.sdk.path}/include/esp_common/include" "-I{compiler.sdk.path}/include/esp_system/include" "-I{compiler.sdk.path}/include/esp_system/port/soc" "-I{compiler.sdk.path}/include/esp_system/port/public_compat" "-I{compiler.sdk.path}/include/esp32/include" "-I{compiler.sdk.path}/include/xtensa/include" "-I{compiler.sdk.path}/include/xtensa/esp32/include" "-I{compiler.sdk.path}/include/driver/include" "-I{compiler.sdk.path}/include/driver/esp32/include" "-I{compiler.sdk.path}/include/esp_pm/include" "-I{compiler.sdk.path}/include/esp_ringbuf/include" "-I{compiler.sdk.path}/include/efuse/include" "-I{compiler.sdk.path}/include/efuse/esp32/include" "-I{compiler.sdk.path}/include/vfs/include" "-I{compiler.sdk.path}/include/esp_wifi/include" "-I{compiler.sdk.path}/include/esp_event/include" "-I{compiler.sdk.path}/include/esp_netif/include" "-I{compiler.sdk.path}/include/esp_eth/include" "-I{compiler.sdk.path}/include/tcpip_adapter/include" "-I{compiler.sdk.path}/include/esp_phy/include" "-I{compiler.sdk.path}/include/esp_phy/esp32/include" "-I{compiler.sdk.path}/include/esp_ipc/include" "-I{compiler.sdk.path}/include/app_trace/include" "-I{compiler.sdk.path}/include/esp_timer/include" "-I{compiler.sdk.path}/include/mbedtls/port/include" "-I{compiler.sdk.path}/include/mbedtls/mbedtls/include" "-I{compiler.sdk.path}/include/mbedtls/esp_crt_bundle/include" "-I{compiler.sdk.path}/include/app_update/include" "-I{compiler.sdk.path}/include/spi_flash/include" "-I{compiler.sdk.path}/include/bootloader_support/include" "-I{compiler.sdk.path}/include/nvs_flash/include" "-I{compiler.sdk.path}/include/pthread/include" "-I{compiler.sdk.path}/include/esp_gdbstub/include" "-I{compiler.sdk.path}/include/esp_gdbstub/xtensa" "-I{compiler.sdk.path}/include/esp_gdbstub/esp32" "-I{compiler.sdk.path}/include/espcoredump/include" "-I{compiler.sdk.path}/include/espcoredump/include/port/xtensa" "-I{compiler.sdk.path}/include/wpa_supplicant/include" "-I{compiler.sdk.path}/include/wpa_supplicant/port/include" "-I{compiler.sdk.path}/include/wpa_supplicant/esp_supplicant/include" "-I{compiler.sdk.path}/include/ieee802154/include" "-I{compiler.sdk.path}/include/asio/asio/asio/include" "-I{compiler.sdk.path}/include/asio/port/include" "-I{compiler.sdk.path}/include/bt/common/osi/include" "-I{compiler.sdk.path}/include/bt/include/esp32/include" "-I{compiler.sdk.path}/include/bt/common/api/include/api" "-I{compiler.sdk.path}/include/bt/common/btc/profile/esp/blufi/include" "-I{compiler.sdk.path}/include/bt/common/btc/profile/esp/include" "-I{compiler.sdk.path}/include/bt/host/bluedroid/api/include/api" "-I{compiler.sdk.path}/include/cbor/port/include" "-I{compiler.sdk.path}/include/unity/include" "-I{compiler.sdk.path}/include/unity/unity/src" "-I{compiler.sdk.path}/include/cmock/CMock/src" "-I{compiler.sdk.path}/include/coap/port/include" "-I{compiler.sdk.path}/include/coap/libcoap/include" "-I{compiler.sdk.path}/include/console" "-I{compiler.sdk.path}/include/nghttp/port/include" "-I{compiler.sdk.path}/include/nghttp/nghttp2/lib/includes" "-I{compiler.sdk.path}/include/esp-tls" "-I{compiler.sdk.path}/include/esp-tls/esp-tls-crypto" "-I{compiler.sdk.path}/include/esp_adc_cal/include" "-I{compiler.sdk.path}/include/esp_hid/include" "-I{compiler.sdk.path}/include/tcp_transport/include" "-I{compiler.sdk.path}/include/esp_http_client/include" "-I{compiler.sdk.path}/include/esp_http_server/include" "-I{compiler.sdk.path}/include/esp_https_ota/include" "-I{compiler.sdk.path}/include/esp_lcd/include" "-I{compiler.sdk.path}/include/esp_lcd/interface" "-I{compiler.sdk.path}/include/protobuf-c/protobuf-c" "-I{compiler.sdk.path}/include/protocomm/include/common" "-I{compiler.sdk.path}/include/protocomm/include/security" "-I{compiler.sdk.path}/include/protocomm/include/transports" "-I{compiler.sdk.path}/include/mdns/include" "-I{compiler.sdk.path}/include/esp_local_ctrl/include" "-I{compiler.sdk.path}/include/sdmmc/include" "-I{compiler.sdk.path}/include/esp_serial_slave_link/include" "-I{compiler.sdk.path}/include/esp_websocket_client/include" "-I{compiler.sdk.path}/include/expat/expat/expat/lib" "-I{compiler.sdk.path}/include/expat/port/include" "-I{compiler.sdk.path}/include/wear_levelling/include" "-I{compiler.sdk.path}/include/fatfs/diskio" "-I{compiler.sdk.path}/include/fatfs/vfs" "-I{compiler.sdk.path}/include/fatfs/src" "-I{compiler.sdk.path}/include/freemodbus/common/include" "-I{compiler.sdk.path}/include/idf_test/include" "-I{compiler.sdk.path}/include/idf_test/include/esp32" "-I{compiler.sdk.path}/include/jsmn/include" "-I{compiler.sdk.path}/include/json/cJSON" "-I{compiler.sdk.path}/include/libsodium/libsodium/src/libsodium/include" "-I{compiler.sdk.path}/include/libsodium/port_include" "-I{compiler.sdk.path}/include/mqtt/esp-mqtt/include" "-I{compiler.sdk.path}/include/openssl/include" "-I{compiler.sdk.path}/include/perfmon/include" "-I{compiler.sdk.path}/include/spiffs/include" "-I{compiler.sdk.path}/include/ulp/include" "-I{compiler.sdk.path}/include/wifi_provisioning/include" "-I{compiler.sdk.path}/include/button/button/include" "-I{compiler.sdk.path}/include/json_parser" "-I{compiler.sdk.path}/include/json_parser/jsmn/include" "-I{compiler.sdk.path}/include/json_generator" "-I{compiler.sdk.path}/include/esp_schedule/include" "-I{compiler.sdk.path}/include/esp_rainmaker/include" "-I{compiler.sdk.path}/include/qrcode/include" "-I{compiler.sdk.path}/include/ws2812_led" "-I{compiler.sdk.path}/include/esp_littlefs/src" "-I{compiler.sdk.path}/include/esp_littlefs/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/dotprod/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/support/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/hann/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/blackman/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/blackman_harris/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/blackman_nuttall/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/nuttall/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/flat_top/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/iir/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/fir/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/add/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/sub/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/mul/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/addc/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/mulc/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/sqrt/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/matrix/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/fft/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/dct/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/conv/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/common/include" "-I{compiler.sdk.path}/include/esp-face/include" "-I{compiler.sdk.path}/include/esp-face/include/tool" "-I{compiler.sdk.path}/include/esp-face/include/typedef" "-I{compiler.sdk.path}/include/esp-face/include/image" "-I{compiler.sdk.path}/include/esp-face/include/math" "-I{compiler.sdk.path}/include/esp-face/include/nn" "-I{compiler.sdk.path}/include/esp-face/include/layer" "-I{compiler.sdk.path}/include/esp-face/include/detect" "-I{compiler.sdk.path}/include/esp-face/include/model_zoo" "-I{compiler.sdk.path}/include/esp32-camera/driver/include" "-I{compiler.sdk.path}/include/esp32-camera/conversions/include" "-I{compiler.sdk.path}/include/fb_gfx/include"
+compiler.cpreprocessor.flags.esp32=-DHAVE_CONFIG_H -DMBEDTLS_CONFIG_FILE="mbedtls/esp_config.h" -DUNITY_INCLUDE_CONFIG_H -DWITH_POSIX -D_GNU_SOURCE -DIDF_VER="v4.4-dev-3401-gb86fe0c66c" -DESP_PLATFORM   "-I{compiler.sdk.path}/include/config" "-I{compiler.sdk.path}/include/newlib/platform_include" "-I{compiler.sdk.path}/include/freertos/include" "-I{compiler.sdk.path}/include/freertos/include/esp_additions/freertos" "-I{compiler.sdk.path}/include/freertos/port/xtensa/include" "-I{compiler.sdk.path}/include/freertos/include/esp_additions" "-I{compiler.sdk.path}/include/esp_hw_support/include" "-I{compiler.sdk.path}/include/esp_hw_support/include/soc" "-I{compiler.sdk.path}/include/esp_hw_support/include/soc/esp32" "-I{compiler.sdk.path}/include/esp_hw_support/port/esp32" "-I{compiler.sdk.path}/include/heap/include" "-I{compiler.sdk.path}/include/log/include" "-I{compiler.sdk.path}/include/lwip/include/apps" "-I{compiler.sdk.path}/include/lwip/include/apps/sntp" "-I{compiler.sdk.path}/include/lwip/lwip/src/include" "-I{compiler.sdk.path}/include/lwip/port/esp32/include" "-I{compiler.sdk.path}/include/lwip/port/esp32/include/arch" "-I{compiler.sdk.path}/include/soc/include" "-I{compiler.sdk.path}/include/soc/esp32" "-I{compiler.sdk.path}/include/soc/esp32/include" "-I{compiler.sdk.path}/include/hal/esp32/include" "-I{compiler.sdk.path}/include/hal/include" "-I{compiler.sdk.path}/include/hal/platform_port/include" "-I{compiler.sdk.path}/include/esp_rom/include" "-I{compiler.sdk.path}/include/esp_rom/include/esp32" "-I{compiler.sdk.path}/include/esp_rom/esp32" "-I{compiler.sdk.path}/include/esp_common/include" "-I{compiler.sdk.path}/include/esp_system/include" "-I{compiler.sdk.path}/include/esp_system/port/soc" "-I{compiler.sdk.path}/include/esp_system/port/public_compat" "-I{compiler.sdk.path}/include/esp32/include" "-I{compiler.sdk.path}/include/xtensa/include" "-I{compiler.sdk.path}/include/xtensa/esp32/include" "-I{compiler.sdk.path}/include/driver/include" "-I{compiler.sdk.path}/include/driver/esp32/include" "-I{compiler.sdk.path}/include/esp_pm/include" "-I{compiler.sdk.path}/include/esp_ringbuf/include" "-I{compiler.sdk.path}/include/efuse/include" "-I{compiler.sdk.path}/include/efuse/esp32/include" "-I{compiler.sdk.path}/include/vfs/include" "-I{compiler.sdk.path}/include/esp_wifi/include" "-I{compiler.sdk.path}/include/esp_event/include" "-I{compiler.sdk.path}/include/esp_netif/include" "-I{compiler.sdk.path}/include/esp_eth/include" "-I{compiler.sdk.path}/include/tcpip_adapter/include" "-I{compiler.sdk.path}/include/esp_phy/include" "-I{compiler.sdk.path}/include/esp_phy/esp32/include" "-I{compiler.sdk.path}/include/esp_ipc/include" "-I{compiler.sdk.path}/include/app_trace/include" "-I{compiler.sdk.path}/include/esp_timer/include" "-I{compiler.sdk.path}/include/mbedtls/port/include" "-I{compiler.sdk.path}/include/mbedtls/mbedtls/include" "-I{compiler.sdk.path}/include/mbedtls/esp_crt_bundle/include" "-I{compiler.sdk.path}/include/app_update/include" "-I{compiler.sdk.path}/include/spi_flash/include" "-I{compiler.sdk.path}/include/bootloader_support/include" "-I{compiler.sdk.path}/include/nvs_flash/include" "-I{compiler.sdk.path}/include/pthread/include" "-I{compiler.sdk.path}/include/esp_gdbstub/include" "-I{compiler.sdk.path}/include/esp_gdbstub/xtensa" "-I{compiler.sdk.path}/include/esp_gdbstub/esp32" "-I{compiler.sdk.path}/include/espcoredump/include" "-I{compiler.sdk.path}/include/espcoredump/include/port/xtensa" "-I{compiler.sdk.path}/include/wpa_supplicant/include" "-I{compiler.sdk.path}/include/wpa_supplicant/port/include" "-I{compiler.sdk.path}/include/wpa_supplicant/esp_supplicant/include" "-I{compiler.sdk.path}/include/ieee802154/include" "-I{compiler.sdk.path}/include/asio/asio/asio/include" "-I{compiler.sdk.path}/include/asio/port/include" "-I{compiler.sdk.path}/include/bt/common/osi/include" "-I{compiler.sdk.path}/include/bt/include/esp32/include" "-I{compiler.sdk.path}/include/bt/common/api/include/api" "-I{compiler.sdk.path}/include/bt/common/btc/profile/esp/blufi/include" "-I{compiler.sdk.path}/include/bt/common/btc/profile/esp/include" "-I{compiler.sdk.path}/include/bt/host/bluedroid/api/include/api" "-I{compiler.sdk.path}/include/cbor/port/include" "-I{compiler.sdk.path}/include/unity/include" "-I{compiler.sdk.path}/include/unity/unity/src" "-I{compiler.sdk.path}/include/cmock/CMock/src" "-I{compiler.sdk.path}/include/coap/port/include" "-I{compiler.sdk.path}/include/coap/libcoap/include" "-I{compiler.sdk.path}/include/console" "-I{compiler.sdk.path}/include/nghttp/port/include" "-I{compiler.sdk.path}/include/nghttp/nghttp2/lib/includes" "-I{compiler.sdk.path}/include/esp-tls" "-I{compiler.sdk.path}/include/esp-tls/esp-tls-crypto" "-I{compiler.sdk.path}/include/esp_adc_cal/include" "-I{compiler.sdk.path}/include/esp_hid/include" "-I{compiler.sdk.path}/include/tcp_transport/include" "-I{compiler.sdk.path}/include/esp_http_client/include" "-I{compiler.sdk.path}/include/esp_http_server/include" "-I{compiler.sdk.path}/include/esp_https_ota/include" "-I{compiler.sdk.path}/include/esp_lcd/include" "-I{compiler.sdk.path}/include/esp_lcd/interface" "-I{compiler.sdk.path}/include/protobuf-c/protobuf-c" "-I{compiler.sdk.path}/include/protocomm/include/common" "-I{compiler.sdk.path}/include/protocomm/include/security" "-I{compiler.sdk.path}/include/protocomm/include/transports" "-I{compiler.sdk.path}/include/mdns/include" "-I{compiler.sdk.path}/include/esp_local_ctrl/include" "-I{compiler.sdk.path}/include/sdmmc/include" "-I{compiler.sdk.path}/include/esp_serial_slave_link/include" "-I{compiler.sdk.path}/include/esp_websocket_client/include" "-I{compiler.sdk.path}/include/expat/expat/expat/lib" "-I{compiler.sdk.path}/include/expat/port/include" "-I{compiler.sdk.path}/include/wear_levelling/include" "-I{compiler.sdk.path}/include/fatfs/diskio" "-I{compiler.sdk.path}/include/fatfs/vfs" "-I{compiler.sdk.path}/include/fatfs/src" "-I{compiler.sdk.path}/include/freemodbus/common/include" "-I{compiler.sdk.path}/include/idf_test/include" "-I{compiler.sdk.path}/include/idf_test/include/esp32" "-I{compiler.sdk.path}/include/jsmn/include" "-I{compiler.sdk.path}/include/json/cJSON" "-I{compiler.sdk.path}/include/libsodium/libsodium/src/libsodium/include" "-I{compiler.sdk.path}/include/libsodium/port_include" "-I{compiler.sdk.path}/include/mqtt/esp-mqtt/include" "-I{compiler.sdk.path}/include/openssl/include" "-I{compiler.sdk.path}/include/perfmon/include" "-I{compiler.sdk.path}/include/spiffs/include" "-I{compiler.sdk.path}/include/ulp/include" "-I{compiler.sdk.path}/include/wifi_provisioning/include" "-I{compiler.sdk.path}/include/button/button/include" "-I{compiler.sdk.path}/include/json_parser" "-I{compiler.sdk.path}/include/json_parser/jsmn/include" "-I{compiler.sdk.path}/include/json_generator" "-I{compiler.sdk.path}/include/esp_schedule/include" "-I{compiler.sdk.path}/include/esp_rainmaker/include" "-I{compiler.sdk.path}/include/qrcode/include" "-I{compiler.sdk.path}/include/ws2812_led" "-I{compiler.sdk.path}/include/esp_littlefs/src" "-I{compiler.sdk.path}/include/esp_littlefs/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/dotprod/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/support/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/hann/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/blackman/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/blackman_harris/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/blackman_nuttall/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/nuttall/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/flat_top/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/iir/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/fir/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/add/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/sub/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/mul/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/addc/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/mulc/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/sqrt/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/matrix/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/fft/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/dct/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/conv/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/common/include" "-I{compiler.sdk.path}/include/esp-face/include" "-I{compiler.sdk.path}/include/esp-face/include/tool" "-I{compiler.sdk.path}/include/esp-face/include/typedef" "-I{compiler.sdk.path}/include/esp-face/include/image" "-I{compiler.sdk.path}/include/esp-face/include/math" "-I{compiler.sdk.path}/include/esp-face/include/nn" "-I{compiler.sdk.path}/include/esp-face/include/layer" "-I{compiler.sdk.path}/include/esp-face/include/detect" "-I{compiler.sdk.path}/include/esp-face/include/model_zoo" "-I{compiler.sdk.path}/include/esp32-camera/driver/include" "-I{compiler.sdk.path}/include/esp32-camera/conversions/include" "-I{compiler.sdk.path}/include/fb_gfx/include"
 compiler.c.elf.libs.esp32=-lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lasio -lbt -lcbor -lunity -lcmock -lcoap -lconsole -lnghttp -lesp-tls -lesp_adc_cal -lesp_hid -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lesp_lcd -lprotobuf-c -lprotocomm -lmdns -lesp_local_ctrl -lsdmmc -lesp_serial_slave_link -lesp_websocket_client -lexpat -lwear_levelling -lfatfs -lfreemodbus -ljsmn -ljson -llibsodium -lmqtt -lopenssl -lperfmon -lspiffs -lulp -lwifi_provisioning -lbutton -ljson_parser -ljson_generator -lesp_schedule -lesp_rainmaker -lqrcode -lws2812_led -lesp-dsp -lesp32-camera -lesp_littlefs -lfb_gfx -lasio -lcbor -lcmock -lunity -lcoap -lesp_lcd -lesp_local_ctrl -lesp_websocket_client -lexpat -lfreemodbus -ljsmn -llibsodium -lperfmon -lesp_adc_cal -lesp_hid -lfatfs -lwear_levelling -lopenssl -lspiffs -lesp_rainmaker -lmqtt -lwifi_provisioning -lprotocomm -lbt -lbtdm_app -lprotobuf-c -lmdns -lconsole -ljson -ljson_parser -ljson_generator -lesp_schedule -lqrcode -lcat_face_detect -lhuman_face_detect -ldl -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lulp -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lulp -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lulp -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lulp -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lulp -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lphy -lrtc -lesp_phy -lphy -lrtc -lesp_phy -lphy -lrtc -lxt_hal -lm -lnewlib -lstdc++ -lpthread -lgcc -lcxx -lapp_trace -lgcov -lapp_trace -lgcov -lc 
 compiler.c.flags.esp32=-mlongcalls -Wno-frame-address -ffunction-sections -fdata-sections -Wno-error=unused-function -Wno-error=unused-variable -Wno-error=deprecated-declarations -Wno-unused-parameter -Wno-sign-compare -ggdb -O2 -Wwrite-strings -fstack-protector -fstrict-volatile-bitfields -Wno-error=unused-but-set-variable -fno-jump-tables -fno-tree-switch-conversion -std=gnu99 -Wno-old-style-declaration  -MMD -c
 compiler.cpp.flags.esp32=-mlongcalls -Wno-frame-address -ffunction-sections -fdata-sections -Wno-error=unused-function -Wno-error=unused-variable -Wno-error=deprecated-declarations -Wno-unused-parameter -Wno-sign-compare -ggdb -O2 -Wwrite-strings -fstack-protector -fstrict-volatile-bitfields -Wno-error=unused-but-set-variable -fno-jump-tables -fno-tree-switch-conversion -std=gnu++11 -fexceptions -fno-rtti  -MMD -c
@@ -38,7 +38,7 @@ build.extra_flags.esp32=-DARDUINO_USB_CDC_ON_BOOT=0
 #
 # ESP32S2 Support Start
 #
-compiler.cpreprocessor.flags.esp32s2=-DHAVE_CONFIG_H -DMBEDTLS_CONFIG_FILE="mbedtls/esp_config.h" -DUNITY_INCLUDE_CONFIG_H -DWITH_POSIX -D_GNU_SOURCE -DIDF_VER="v4.4-dev-3235-g3e370c4296" -DESP_PLATFORM   "-I{compiler.sdk.path}/include/config" "-I{compiler.sdk.path}/include/newlib/platform_include" "-I{compiler.sdk.path}/include/freertos/include" "-I{compiler.sdk.path}/include/freertos/include/esp_additions/freertos" "-I{compiler.sdk.path}/include/freertos/port/xtensa/include" "-I{compiler.sdk.path}/include/freertos/include/esp_additions" "-I{compiler.sdk.path}/include/esp_hw_support/include" "-I{compiler.sdk.path}/include/esp_hw_support/include/soc" "-I{compiler.sdk.path}/include/esp_hw_support/include/soc/esp32s2" "-I{compiler.sdk.path}/include/esp_hw_support/port/esp32s2" "-I{compiler.sdk.path}/include/esp_hw_support/port/esp32s2/private_include" "-I{compiler.sdk.path}/include/heap/include" "-I{compiler.sdk.path}/include/log/include" "-I{compiler.sdk.path}/include/lwip/include/apps" "-I{compiler.sdk.path}/include/lwip/include/apps/sntp" "-I{compiler.sdk.path}/include/lwip/lwip/src/include" "-I{compiler.sdk.path}/include/lwip/port/esp32/include" "-I{compiler.sdk.path}/include/lwip/port/esp32/include/arch" "-I{compiler.sdk.path}/include/soc/include" "-I{compiler.sdk.path}/include/soc/esp32s2" "-I{compiler.sdk.path}/include/soc/esp32s2/include" "-I{compiler.sdk.path}/include/hal/esp32s2/include" "-I{compiler.sdk.path}/include/hal/include" "-I{compiler.sdk.path}/include/hal/platform_port/include" "-I{compiler.sdk.path}/include/esp_rom/include" "-I{compiler.sdk.path}/include/esp_rom/include/esp32s2" "-I{compiler.sdk.path}/include/esp_rom/esp32s2" "-I{compiler.sdk.path}/include/esp_common/include" "-I{compiler.sdk.path}/include/esp_system/include" "-I{compiler.sdk.path}/include/esp_system/port/soc" "-I{compiler.sdk.path}/include/esp_system/port/public_compat" "-I{compiler.sdk.path}/include/xtensa/include" "-I{compiler.sdk.path}/include/xtensa/esp32s2/include" "-I{compiler.sdk.path}/include/driver/include" "-I{compiler.sdk.path}/include/driver/esp32s2/include" "-I{compiler.sdk.path}/include/esp_pm/include" "-I{compiler.sdk.path}/include/esp_ringbuf/include" "-I{compiler.sdk.path}/include/efuse/include" "-I{compiler.sdk.path}/include/efuse/esp32s2/include" "-I{compiler.sdk.path}/include/vfs/include" "-I{compiler.sdk.path}/include/esp_wifi/include" "-I{compiler.sdk.path}/include/esp_event/include" "-I{compiler.sdk.path}/include/esp_netif/include" "-I{compiler.sdk.path}/include/esp_eth/include" "-I{compiler.sdk.path}/include/tcpip_adapter/include" "-I{compiler.sdk.path}/include/esp_phy/include" "-I{compiler.sdk.path}/include/esp_phy/esp32s2/include" "-I{compiler.sdk.path}/include/esp_ipc/include" "-I{compiler.sdk.path}/include/app_trace/include" "-I{compiler.sdk.path}/include/esp_timer/include" "-I{compiler.sdk.path}/include/mbedtls/port/include" "-I{compiler.sdk.path}/include/mbedtls/mbedtls/include" "-I{compiler.sdk.path}/include/mbedtls/esp_crt_bundle/include" "-I{compiler.sdk.path}/include/app_update/include" "-I{compiler.sdk.path}/include/spi_flash/include" "-I{compiler.sdk.path}/include/bootloader_support/include" "-I{compiler.sdk.path}/include/nvs_flash/include" "-I{compiler.sdk.path}/include/pthread/include" "-I{compiler.sdk.path}/include/esp_gdbstub/include" "-I{compiler.sdk.path}/include/esp_gdbstub/xtensa" "-I{compiler.sdk.path}/include/esp_gdbstub/esp32s2" "-I{compiler.sdk.path}/include/espcoredump/include" "-I{compiler.sdk.path}/include/espcoredump/include/port/xtensa" "-I{compiler.sdk.path}/include/wpa_supplicant/include" "-I{compiler.sdk.path}/include/wpa_supplicant/port/include" "-I{compiler.sdk.path}/include/wpa_supplicant/esp_supplicant/include" "-I{compiler.sdk.path}/include/ieee802154/include" "-I{compiler.sdk.path}/include/asio/asio/asio/include" "-I{compiler.sdk.path}/include/asio/port/include" "-I{compiler.sdk.path}/include/cbor/port/include" "-I{compiler.sdk.path}/include/unity/include" "-I{compiler.sdk.path}/include/unity/unity/src" "-I{compiler.sdk.path}/include/cmock/CMock/src" "-I{compiler.sdk.path}/include/coap/port/include" "-I{compiler.sdk.path}/include/coap/libcoap/include" "-I{compiler.sdk.path}/include/console" "-I{compiler.sdk.path}/include/nghttp/port/include" "-I{compiler.sdk.path}/include/nghttp/nghttp2/lib/includes" "-I{compiler.sdk.path}/include/esp-tls" "-I{compiler.sdk.path}/include/esp-tls/esp-tls-crypto" "-I{compiler.sdk.path}/include/esp_adc_cal/include" "-I{compiler.sdk.path}/include/esp_hid/include" "-I{compiler.sdk.path}/include/tcp_transport/include" "-I{compiler.sdk.path}/include/esp_http_client/include" "-I{compiler.sdk.path}/include/esp_http_server/include" "-I{compiler.sdk.path}/include/esp_https_ota/include" "-I{compiler.sdk.path}/include/esp_https_server/include" "-I{compiler.sdk.path}/include/esp_lcd/include" "-I{compiler.sdk.path}/include/esp_lcd/interface" "-I{compiler.sdk.path}/include/protobuf-c/protobuf-c" "-I{compiler.sdk.path}/include/protocomm/include/common" "-I{compiler.sdk.path}/include/protocomm/include/security" "-I{compiler.sdk.path}/include/protocomm/include/transports" "-I{compiler.sdk.path}/include/mdns/include" "-I{compiler.sdk.path}/include/esp_local_ctrl/include" "-I{compiler.sdk.path}/include/sdmmc/include" "-I{compiler.sdk.path}/include/esp_serial_slave_link/include" "-I{compiler.sdk.path}/include/esp_websocket_client/include" "-I{compiler.sdk.path}/include/expat/expat/expat/lib" "-I{compiler.sdk.path}/include/expat/port/include" "-I{compiler.sdk.path}/include/wear_levelling/include" "-I{compiler.sdk.path}/include/fatfs/diskio" "-I{compiler.sdk.path}/include/fatfs/vfs" "-I{compiler.sdk.path}/include/fatfs/src" "-I{compiler.sdk.path}/include/freemodbus/common/include" "-I{compiler.sdk.path}/include/idf_test/include" "-I{compiler.sdk.path}/include/idf_test/include/esp32s2" "-I{compiler.sdk.path}/include/jsmn/include" "-I{compiler.sdk.path}/include/json/cJSON" "-I{compiler.sdk.path}/include/libsodium/libsodium/src/libsodium/include" "-I{compiler.sdk.path}/include/libsodium/port_include" "-I{compiler.sdk.path}/include/mqtt/esp-mqtt/include" "-I{compiler.sdk.path}/include/openssl/include" "-I{compiler.sdk.path}/include/perfmon/include" "-I{compiler.sdk.path}/include/spiffs/include" "-I{compiler.sdk.path}/include/usb/include" "-I{compiler.sdk.path}/include/touch_element/include" "-I{compiler.sdk.path}/include/ulp/include" "-I{compiler.sdk.path}/include/wifi_provisioning/include" "-I{compiler.sdk.path}/include/freertos/include/freertos" "-I{compiler.sdk.path}/include/arduino_tinyusb/tinyusb/src" "-I{compiler.sdk.path}/include/arduino_tinyusb/include" "-I{compiler.sdk.path}/include/esp_littlefs/src" "-I{compiler.sdk.path}/include/esp_littlefs/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/dotprod/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/support/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/hann/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/blackman/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/blackman_harris/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/blackman_nuttall/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/nuttall/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/flat_top/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/iir/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/fir/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/add/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/sub/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/mul/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/addc/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/mulc/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/sqrt/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/matrix/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/fft/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/dct/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/conv/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/common/include" "-I{compiler.sdk.path}/include/esp-face/include" "-I{compiler.sdk.path}/include/esp-face/include/tool" "-I{compiler.sdk.path}/include/esp-face/include/typedef" "-I{compiler.sdk.path}/include/esp-face/include/image" "-I{compiler.sdk.path}/include/esp-face/include/math" "-I{compiler.sdk.path}/include/esp-face/include/nn" "-I{compiler.sdk.path}/include/esp-face/include/layer" "-I{compiler.sdk.path}/include/esp-face/include/detect" "-I{compiler.sdk.path}/include/esp-face/include/model_zoo" "-I{compiler.sdk.path}/include/esp32-camera/driver/include" "-I{compiler.sdk.path}/include/esp32-camera/conversions/include" "-I{compiler.sdk.path}/include/fb_gfx/include"
+compiler.cpreprocessor.flags.esp32s2=-DHAVE_CONFIG_H -DMBEDTLS_CONFIG_FILE="mbedtls/esp_config.h" -DUNITY_INCLUDE_CONFIG_H -DWITH_POSIX -D_GNU_SOURCE -DIDF_VER="v4.4-dev-3401-gb86fe0c66c" -DESP_PLATFORM   "-I{compiler.sdk.path}/include/config" "-I{compiler.sdk.path}/include/newlib/platform_include" "-I{compiler.sdk.path}/include/freertos/include" "-I{compiler.sdk.path}/include/freertos/include/esp_additions/freertos" "-I{compiler.sdk.path}/include/freertos/port/xtensa/include" "-I{compiler.sdk.path}/include/freertos/include/esp_additions" "-I{compiler.sdk.path}/include/esp_hw_support/include" "-I{compiler.sdk.path}/include/esp_hw_support/include/soc" "-I{compiler.sdk.path}/include/esp_hw_support/include/soc/esp32s2" "-I{compiler.sdk.path}/include/esp_hw_support/port/esp32s2" "-I{compiler.sdk.path}/include/esp_hw_support/port/esp32s2/private_include" "-I{compiler.sdk.path}/include/heap/include" "-I{compiler.sdk.path}/include/log/include" "-I{compiler.sdk.path}/include/lwip/include/apps" "-I{compiler.sdk.path}/include/lwip/include/apps/sntp" "-I{compiler.sdk.path}/include/lwip/lwip/src/include" "-I{compiler.sdk.path}/include/lwip/port/esp32/include" "-I{compiler.sdk.path}/include/lwip/port/esp32/include/arch" "-I{compiler.sdk.path}/include/soc/include" "-I{compiler.sdk.path}/include/soc/esp32s2" "-I{compiler.sdk.path}/include/soc/esp32s2/include" "-I{compiler.sdk.path}/include/hal/esp32s2/include" "-I{compiler.sdk.path}/include/hal/include" "-I{compiler.sdk.path}/include/hal/platform_port/include" "-I{compiler.sdk.path}/include/esp_rom/include" "-I{compiler.sdk.path}/include/esp_rom/include/esp32s2" "-I{compiler.sdk.path}/include/esp_rom/esp32s2" "-I{compiler.sdk.path}/include/esp_common/include" "-I{compiler.sdk.path}/include/esp_system/include" "-I{compiler.sdk.path}/include/esp_system/port/soc" "-I{compiler.sdk.path}/include/esp_system/port/public_compat" "-I{compiler.sdk.path}/include/xtensa/include" "-I{compiler.sdk.path}/include/xtensa/esp32s2/include" "-I{compiler.sdk.path}/include/driver/include" "-I{compiler.sdk.path}/include/driver/esp32s2/include" "-I{compiler.sdk.path}/include/esp_pm/include" "-I{compiler.sdk.path}/include/esp_ringbuf/include" "-I{compiler.sdk.path}/include/efuse/include" "-I{compiler.sdk.path}/include/efuse/esp32s2/include" "-I{compiler.sdk.path}/include/vfs/include" "-I{compiler.sdk.path}/include/esp_wifi/include" "-I{compiler.sdk.path}/include/esp_event/include" "-I{compiler.sdk.path}/include/esp_netif/include" "-I{compiler.sdk.path}/include/esp_eth/include" "-I{compiler.sdk.path}/include/tcpip_adapter/include" "-I{compiler.sdk.path}/include/esp_phy/include" "-I{compiler.sdk.path}/include/esp_phy/esp32s2/include" "-I{compiler.sdk.path}/include/esp_ipc/include" "-I{compiler.sdk.path}/include/app_trace/include" "-I{compiler.sdk.path}/include/esp_timer/include" "-I{compiler.sdk.path}/include/mbedtls/port/include" "-I{compiler.sdk.path}/include/mbedtls/mbedtls/include" "-I{compiler.sdk.path}/include/mbedtls/esp_crt_bundle/include" "-I{compiler.sdk.path}/include/app_update/include" "-I{compiler.sdk.path}/include/spi_flash/include" "-I{compiler.sdk.path}/include/bootloader_support/include" "-I{compiler.sdk.path}/include/nvs_flash/include" "-I{compiler.sdk.path}/include/pthread/include" "-I{compiler.sdk.path}/include/esp_gdbstub/include" "-I{compiler.sdk.path}/include/esp_gdbstub/xtensa" "-I{compiler.sdk.path}/include/esp_gdbstub/esp32s2" "-I{compiler.sdk.path}/include/espcoredump/include" "-I{compiler.sdk.path}/include/espcoredump/include/port/xtensa" "-I{compiler.sdk.path}/include/wpa_supplicant/include" "-I{compiler.sdk.path}/include/wpa_supplicant/port/include" "-I{compiler.sdk.path}/include/wpa_supplicant/esp_supplicant/include" "-I{compiler.sdk.path}/include/ieee802154/include" "-I{compiler.sdk.path}/include/asio/asio/asio/include" "-I{compiler.sdk.path}/include/asio/port/include" "-I{compiler.sdk.path}/include/cbor/port/include" "-I{compiler.sdk.path}/include/unity/include" "-I{compiler.sdk.path}/include/unity/unity/src" "-I{compiler.sdk.path}/include/cmock/CMock/src" "-I{compiler.sdk.path}/include/coap/port/include" "-I{compiler.sdk.path}/include/coap/libcoap/include" "-I{compiler.sdk.path}/include/console" "-I{compiler.sdk.path}/include/nghttp/port/include" "-I{compiler.sdk.path}/include/nghttp/nghttp2/lib/includes" "-I{compiler.sdk.path}/include/esp-tls" "-I{compiler.sdk.path}/include/esp-tls/esp-tls-crypto" "-I{compiler.sdk.path}/include/esp_adc_cal/include" "-I{compiler.sdk.path}/include/esp_hid/include" "-I{compiler.sdk.path}/include/tcp_transport/include" "-I{compiler.sdk.path}/include/esp_http_client/include" "-I{compiler.sdk.path}/include/esp_http_server/include" "-I{compiler.sdk.path}/include/esp_https_ota/include" "-I{compiler.sdk.path}/include/esp_https_server/include" "-I{compiler.sdk.path}/include/esp_lcd/include" "-I{compiler.sdk.path}/include/esp_lcd/interface" "-I{compiler.sdk.path}/include/protobuf-c/protobuf-c" "-I{compiler.sdk.path}/include/protocomm/include/common" "-I{compiler.sdk.path}/include/protocomm/include/security" "-I{compiler.sdk.path}/include/protocomm/include/transports" "-I{compiler.sdk.path}/include/mdns/include" "-I{compiler.sdk.path}/include/esp_local_ctrl/include" "-I{compiler.sdk.path}/include/sdmmc/include" "-I{compiler.sdk.path}/include/esp_serial_slave_link/include" "-I{compiler.sdk.path}/include/esp_websocket_client/include" "-I{compiler.sdk.path}/include/expat/expat/expat/lib" "-I{compiler.sdk.path}/include/expat/port/include" "-I{compiler.sdk.path}/include/wear_levelling/include" "-I{compiler.sdk.path}/include/fatfs/diskio" "-I{compiler.sdk.path}/include/fatfs/vfs" "-I{compiler.sdk.path}/include/fatfs/src" "-I{compiler.sdk.path}/include/freemodbus/common/include" "-I{compiler.sdk.path}/include/idf_test/include" "-I{compiler.sdk.path}/include/idf_test/include/esp32s2" "-I{compiler.sdk.path}/include/jsmn/include" "-I{compiler.sdk.path}/include/json/cJSON" "-I{compiler.sdk.path}/include/libsodium/libsodium/src/libsodium/include" "-I{compiler.sdk.path}/include/libsodium/port_include" "-I{compiler.sdk.path}/include/mqtt/esp-mqtt/include" "-I{compiler.sdk.path}/include/openssl/include" "-I{compiler.sdk.path}/include/perfmon/include" "-I{compiler.sdk.path}/include/spiffs/include" "-I{compiler.sdk.path}/include/usb/include" "-I{compiler.sdk.path}/include/touch_element/include" "-I{compiler.sdk.path}/include/ulp/include" "-I{compiler.sdk.path}/include/wifi_provisioning/include" "-I{compiler.sdk.path}/include/freertos/include/freertos" "-I{compiler.sdk.path}/include/arduino_tinyusb/tinyusb/src" "-I{compiler.sdk.path}/include/arduino_tinyusb/include" "-I{compiler.sdk.path}/include/esp_littlefs/src" "-I{compiler.sdk.path}/include/esp_littlefs/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/dotprod/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/support/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/hann/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/blackman/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/blackman_harris/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/blackman_nuttall/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/nuttall/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/windows/flat_top/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/iir/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/fir/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/add/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/sub/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/mul/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/addc/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/mulc/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/math/sqrt/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/matrix/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/fft/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/dct/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/conv/include" "-I{compiler.sdk.path}/include/esp-dsp/modules/common/include" "-I{compiler.sdk.path}/include/esp-face/include" "-I{compiler.sdk.path}/include/esp-face/include/tool" "-I{compiler.sdk.path}/include/esp-face/include/typedef" "-I{compiler.sdk.path}/include/esp-face/include/image" "-I{compiler.sdk.path}/include/esp-face/include/math" "-I{compiler.sdk.path}/include/esp-face/include/nn" "-I{compiler.sdk.path}/include/esp-face/include/layer" "-I{compiler.sdk.path}/include/esp-face/include/detect" "-I{compiler.sdk.path}/include/esp-face/include/model_zoo" "-I{compiler.sdk.path}/include/esp32-camera/driver/include" "-I{compiler.sdk.path}/include/esp32-camera/conversions/include" "-I{compiler.sdk.path}/include/fb_gfx/include"
 compiler.c.elf.libs.esp32s2=-lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lasio -lcbor -lunity -lcmock -lcoap -lconsole -lnghttp -lesp-tls -lesp_adc_cal -lesp_hid -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lesp_https_server -lesp_lcd -lprotobuf-c -lprotocomm -lmdns -lesp_local_ctrl -lsdmmc -lesp_serial_slave_link -lesp_websocket_client -lexpat -lwear_levelling -lfatfs -lfreemodbus -ljsmn -ljson -llibsodium -lmqtt -lopenssl -lperfmon -lspiffs -lusb -ltouch_element -lulp -lwifi_provisioning -lesp-dsp -lesp32-camera -lesp_littlefs -lfb_gfx -lasio -lcbor -lcmock -lunity -lcoap -lesp_lcd -lesp_local_ctrl -lesp_https_server -lesp_websocket_client -lexpat -lfreemodbus -ljsmn -llibsodium -lmqtt -lperfmon -lusb -ltouch_element -lesp_adc_cal -lesp_hid -lfatfs -lwear_levelling -lopenssl -lspiffs -lwifi_provisioning -lprotocomm -lprotobuf-c -lmdns -lconsole -ljson -larduino_tinyusb -lcat_face_detect -lhuman_face_detect -ldl -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lulp -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lulp -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lulp -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lulp -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lulp -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lxtensa -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lulp -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lphy -lesp_phy -lphy -lesp_phy -lphy -lxt_hal -lm -lnewlib -lstdc++ -lpthread -lgcc -lcxx -lapp_trace -lgcov -lapp_trace -lgcov -lc 
 compiler.c.flags.esp32s2=-mlongcalls -ffunction-sections -fdata-sections -Wno-error=unused-function -Wno-error=unused-variable -Wno-error=deprecated-declarations -Wno-unused-parameter -Wno-sign-compare -ggdb -O2 -Wwrite-strings -fstack-protector -fstrict-volatile-bitfields -Wno-error=unused-but-set-variable -fno-jump-tables -fno-tree-switch-conversion -std=gnu99 -Wno-old-style-declaration  -MMD -c
 compiler.cpp.flags.esp32s2=-mlongcalls -ffunction-sections -fdata-sections -Wno-error=unused-function -Wno-error=unused-variable -Wno-error=deprecated-declarations -Wno-unused-parameter -Wno-sign-compare -ggdb -O2 -Wwrite-strings -fstack-protector -fstrict-volatile-bitfields -Wno-error=unused-but-set-variable -fno-jump-tables -fno-tree-switch-conversion -std=gnu++11 -fexceptions -fno-rtti  -MMD -c
@@ -53,8 +53,8 @@ build.extra_flags.esp32s2=-DARDUINO_USB_CDC_ON_BOOT={build.cdc_on_boot} -DARDUIN
 #
 # ESP32C3 Support Start
 #
-compiler.cpreprocessor.flags.esp32c3=-DHAVE_CONFIG_H -DMBEDTLS_CONFIG_FILE="mbedtls/esp_config.h" -DUNITY_INCLUDE_CONFIG_H -DWITH_POSIX -D_GNU_SOURCE -DIDF_VER="v4.4-dev-3235-g3e370c4296" -DESP_PLATFORM   "-I{compiler.sdk.path}/include/config" "-I{compiler.sdk.path}/include/newlib/platform_include" "-I{compiler.sdk.path}/include/freertos/include" "-I{compiler.sdk.path}/include/freertos/include/esp_additions/freertos" "-I{compiler.sdk.path}/include/freertos/port/riscv/include" "-I{compiler.sdk.path}/include/freertos/include/esp_additions" "-I{compiler.sdk.path}/include/esp_hw_support/include" "-I{compiler.sdk.path}/include/esp_hw_support/include/soc" "-I{compiler.sdk.path}/include/esp_hw_support/include/soc/esp32c3" "-I{compiler.sdk.path}/include/esp_hw_support/port/esp32c3" "-I{compiler.sdk.path}/include/esp_hw_support/port/esp32c3/private_include" "-I{compiler.sdk.path}/include/heap/include" "-I{compiler.sdk.path}/include/log/include" "-I{compiler.sdk.path}/include/lwip/include/apps" "-I{compiler.sdk.path}/include/lwip/include/apps/sntp" "-I{compiler.sdk.path}/include/lwip/lwip/src/include" "-I{compiler.sdk.path}/include/lwip/port/esp32/include" "-I{compiler.sdk.path}/include/lwip/port/esp32/include/arch" "-I{compiler.sdk.path}/include/soc/include" "-I{compiler.sdk.path}/include/soc/esp32c3" "-I{compiler.sdk.path}/include/soc/esp32c3/include" "-I{compiler.sdk.path}/include/hal/esp32c3/include" "-I{compiler.sdk.path}/include/hal/include" "-I{compiler.sdk.path}/include/hal/platform_port/include" "-I{compiler.sdk.path}/include/esp_rom/include" "-I{compiler.sdk.path}/include/esp_rom/include/esp32c3" "-I{compiler.sdk.path}/include/esp_rom/esp32c3" "-I{compiler.sdk.path}/include/esp_common/include" "-I{compiler.sdk.path}/include/esp_system/include" "-I{compiler.sdk.path}/include/esp_system/port/soc" "-I{compiler.sdk.path}/include/esp_system/port/include/riscv" "-I{compiler.sdk.path}/include/esp_system/port/public_compat" "-I{compiler.sdk.path}/include/riscv/include" "-I{compiler.sdk.path}/include/driver/include" "-I{compiler.sdk.path}/include/driver/esp32c3/include" "-I{compiler.sdk.path}/include/esp_pm/include" "-I{compiler.sdk.path}/include/esp_ringbuf/include" "-I{compiler.sdk.path}/include/efuse/include" "-I{compiler.sdk.path}/include/efuse/esp32c3/include" "-I{compiler.sdk.path}/include/vfs/include" "-I{compiler.sdk.path}/include/esp_wifi/include" "-I{compiler.sdk.path}/include/esp_event/include" "-I{compiler.sdk.path}/include/esp_netif/include" "-I{compiler.sdk.path}/include/esp_eth/include" "-I{compiler.sdk.path}/include/tcpip_adapter/include" "-I{compiler.sdk.path}/include/esp_phy/include" "-I{compiler.sdk.path}/include/esp_phy/esp32c3/include" "-I{compiler.sdk.path}/include/esp_ipc/include" "-I{compiler.sdk.path}/include/app_trace/include" "-I{compiler.sdk.path}/include/esp_timer/include" "-I{compiler.sdk.path}/include/mbedtls/port/include" "-I{compiler.sdk.path}/include/mbedtls/mbedtls/include" "-I{compiler.sdk.path}/include/mbedtls/esp_crt_bundle/include" "-I{compiler.sdk.path}/include/app_update/include" "-I{compiler.sdk.path}/include/spi_flash/include" "-I{compiler.sdk.path}/include/bootloader_support/include" "-I{compiler.sdk.path}/include/nvs_flash/include" "-I{compiler.sdk.path}/include/pthread/include" "-I{compiler.sdk.path}/include/esp_gdbstub/include" "-I{compiler.sdk.path}/include/esp_gdbstub/riscv" "-I{compiler.sdk.path}/include/esp_gdbstub/esp32c3" "-I{compiler.sdk.path}/include/espcoredump/include" "-I{compiler.sdk.path}/include/espcoredump/include/port/riscv" "-I{compiler.sdk.path}/include/wpa_supplicant/include" "-I{compiler.sdk.path}/include/wpa_supplicant/port/include" "-I{compiler.sdk.path}/include/wpa_supplicant/esp_supplicant/include" "-I{compiler.sdk.path}/include/ieee802154/include" "-I{compiler.sdk.path}/include/asio/asio/asio/include" "-I{compiler.sdk.path}/include/asio/port/include" "-I{compiler.sdk.path}/include/bt/common/osi/include" "-I{compiler.sdk.path}/include/bt/include/esp32c3/include" "-I{compiler.sdk.path}/include/bt/common/api/include/api" "-I{compiler.sdk.path}/include/bt/common/btc/profile/esp/blufi/include" "-I{compiler.sdk.path}/include/bt/common/btc/profile/esp/include" "-I{compiler.sdk.path}/include/bt/host/bluedroid/api/include/api" "-I{compiler.sdk.path}/include/cbor/port/include" "-I{compiler.sdk.path}/include/unity/include" "-I{compiler.sdk.path}/include/unity/unity/src" "-I{compiler.sdk.path}/include/cmock/CMock/src" "-I{compiler.sdk.path}/include/coap/port/include" "-I{compiler.sdk.path}/include/coap/libcoap/include" "-I{compiler.sdk.path}/include/console" "-I{compiler.sdk.path}/include/nghttp/port/include" "-I{compiler.sdk.path}/include/nghttp/nghttp2/lib/includes" "-I{compiler.sdk.path}/include/esp-tls" "-I{compiler.sdk.path}/include/esp-tls/esp-tls-crypto" "-I{compiler.sdk.path}/include/esp_adc_cal/include" "-I{compiler.sdk.path}/include/esp_hid/include" "-I{compiler.sdk.path}/include/tcp_transport/include" "-I{compiler.sdk.path}/include/esp_http_client/include" "-I{compiler.sdk.path}/include/esp_http_server/include" "-I{compiler.sdk.path}/include/esp_https_ota/include" "-I{compiler.sdk.path}/include/esp_https_server/include" "-I{compiler.sdk.path}/include/esp_lcd/include" "-I{compiler.sdk.path}/include/esp_lcd/interface" "-I{compiler.sdk.path}/include/protobuf-c/protobuf-c" "-I{compiler.sdk.path}/include/protocomm/include/common" "-I{compiler.sdk.path}/include/protocomm/include/security" "-I{compiler.sdk.path}/include/protocomm/include/transports" "-I{compiler.sdk.path}/include/mdns/include" "-I{compiler.sdk.path}/include/esp_local_ctrl/include" "-I{compiler.sdk.path}/include/sdmmc/include" "-I{compiler.sdk.path}/include/esp_serial_slave_link/include" "-I{compiler.sdk.path}/include/esp_websocket_client/include" "-I{compiler.sdk.path}/include/expat/expat/expat/lib" "-I{compiler.sdk.path}/include/expat/port/include" "-I{compiler.sdk.path}/include/wear_levelling/include" "-I{compiler.sdk.path}/include/fatfs/diskio" "-I{compiler.sdk.path}/include/fatfs/vfs" "-I{compiler.sdk.path}/include/fatfs/src" "-I{compiler.sdk.path}/include/freemodbus/common/include" "-I{compiler.sdk.path}/include/idf_test/include" "-I{compiler.sdk.path}/include/idf_test/include/esp32c3" "-I{compiler.sdk.path}/include/jsmn/include" "-I{compiler.sdk.path}/include/json/cJSON" "-I{compiler.sdk.path}/include/libsodium/libsodium/src/libsodium/include" "-I{compiler.sdk.path}/include/libsodium/port_include" "-I{compiler.sdk.path}/include/mqtt/esp-mqtt/include" "-I{compiler.sdk.path}/include/openssl/include" "-I{compiler.sdk.path}/include/spiffs/include" "-I{compiler.sdk.path}/include/wifi_provisioning/include" "-I{compiler.sdk.path}/include/esp_littlefs/src" "-I{compiler.sdk.path}/include/esp_littlefs/include" "-I{compiler.sdk.path}/include/fb_gfx/include"
-compiler.c.elf.libs.esp32c3=-lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lasio -lbt -lcbor -lunity -lcmock -lcoap -lconsole -lnghttp -lesp-tls -lesp_adc_cal -lesp_hid -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lesp_https_server -lesp_lcd -lprotobuf-c -lprotocomm -lmdns -lesp_local_ctrl -lsdmmc -lesp_serial_slave_link -lesp_websocket_client -lexpat -lwear_levelling -lfatfs -lfreemodbus -ljsmn -ljson -llibsodium -lmqtt -lopenssl -lspiffs -lwifi_provisioning -lesp_littlefs -lfb_gfx -lasio -lcbor -lcmock -lunity -lcoap -lesp_lcd -lesp_local_ctrl -lesp_https_server -lesp_websocket_client -lexpat -lfreemodbus -ljsmn -llibsodium -lmqtt -lesp_adc_cal -lesp_hid -lfatfs -lwear_levelling -lopenssl -lspiffs -lwifi_provisioning -lprotocomm -lbt -lbtdm_app -lprotobuf-c -lmdns -lconsole -ljson -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lphy -lbtbb -lesp_phy -lphy -lbtbb -lesp_phy -lphy -lbtbb -lm -lnewlib -lstdc++ -lpthread -lgcc -lcxx -lapp_trace -lgcov -lapp_trace -lgcov -lc 
+compiler.cpreprocessor.flags.esp32c3=-DHAVE_CONFIG_H -DMBEDTLS_CONFIG_FILE="mbedtls/esp_config.h" -DUNITY_INCLUDE_CONFIG_H -DWITH_POSIX -D_GNU_SOURCE -DIDF_VER="v4.4-dev-3401-gb86fe0c66c" -DESP_PLATFORM   "-I{compiler.sdk.path}/include/config" "-I{compiler.sdk.path}/include/newlib/platform_include" "-I{compiler.sdk.path}/include/freertos/include" "-I{compiler.sdk.path}/include/freertos/include/esp_additions/freertos" "-I{compiler.sdk.path}/include/freertos/port/riscv/include" "-I{compiler.sdk.path}/include/freertos/include/esp_additions" "-I{compiler.sdk.path}/include/esp_hw_support/include" "-I{compiler.sdk.path}/include/esp_hw_support/include/soc" "-I{compiler.sdk.path}/include/esp_hw_support/include/soc/esp32c3" "-I{compiler.sdk.path}/include/esp_hw_support/port/esp32c3" "-I{compiler.sdk.path}/include/esp_hw_support/port/esp32c3/private_include" "-I{compiler.sdk.path}/include/heap/include" "-I{compiler.sdk.path}/include/log/include" "-I{compiler.sdk.path}/include/lwip/include/apps" "-I{compiler.sdk.path}/include/lwip/include/apps/sntp" "-I{compiler.sdk.path}/include/lwip/lwip/src/include" "-I{compiler.sdk.path}/include/lwip/port/esp32/include" "-I{compiler.sdk.path}/include/lwip/port/esp32/include/arch" "-I{compiler.sdk.path}/include/soc/include" "-I{compiler.sdk.path}/include/soc/esp32c3" "-I{compiler.sdk.path}/include/soc/esp32c3/include" "-I{compiler.sdk.path}/include/hal/esp32c3/include" "-I{compiler.sdk.path}/include/hal/include" "-I{compiler.sdk.path}/include/hal/platform_port/include" "-I{compiler.sdk.path}/include/esp_rom/include" "-I{compiler.sdk.path}/include/esp_rom/include/esp32c3" "-I{compiler.sdk.path}/include/esp_rom/esp32c3" "-I{compiler.sdk.path}/include/esp_common/include" "-I{compiler.sdk.path}/include/esp_system/include" "-I{compiler.sdk.path}/include/esp_system/port/soc" "-I{compiler.sdk.path}/include/esp_system/port/include/riscv" "-I{compiler.sdk.path}/include/esp_system/port/public_compat" "-I{compiler.sdk.path}/include/riscv/include" "-I{compiler.sdk.path}/include/driver/include" "-I{compiler.sdk.path}/include/driver/esp32c3/include" "-I{compiler.sdk.path}/include/esp_pm/include" "-I{compiler.sdk.path}/include/esp_ringbuf/include" "-I{compiler.sdk.path}/include/efuse/include" "-I{compiler.sdk.path}/include/efuse/esp32c3/include" "-I{compiler.sdk.path}/include/vfs/include" "-I{compiler.sdk.path}/include/esp_wifi/include" "-I{compiler.sdk.path}/include/esp_event/include" "-I{compiler.sdk.path}/include/esp_netif/include" "-I{compiler.sdk.path}/include/esp_eth/include" "-I{compiler.sdk.path}/include/tcpip_adapter/include" "-I{compiler.sdk.path}/include/esp_phy/include" "-I{compiler.sdk.path}/include/esp_phy/esp32c3/include" "-I{compiler.sdk.path}/include/esp_ipc/include" "-I{compiler.sdk.path}/include/app_trace/include" "-I{compiler.sdk.path}/include/esp_timer/include" "-I{compiler.sdk.path}/include/mbedtls/port/include" "-I{compiler.sdk.path}/include/mbedtls/mbedtls/include" "-I{compiler.sdk.path}/include/mbedtls/esp_crt_bundle/include" "-I{compiler.sdk.path}/include/app_update/include" "-I{compiler.sdk.path}/include/spi_flash/include" "-I{compiler.sdk.path}/include/bootloader_support/include" "-I{compiler.sdk.path}/include/nvs_flash/include" "-I{compiler.sdk.path}/include/pthread/include" "-I{compiler.sdk.path}/include/esp_gdbstub/include" "-I{compiler.sdk.path}/include/esp_gdbstub/riscv" "-I{compiler.sdk.path}/include/esp_gdbstub/esp32c3" "-I{compiler.sdk.path}/include/espcoredump/include" "-I{compiler.sdk.path}/include/espcoredump/include/port/riscv" "-I{compiler.sdk.path}/include/wpa_supplicant/include" "-I{compiler.sdk.path}/include/wpa_supplicant/port/include" "-I{compiler.sdk.path}/include/wpa_supplicant/esp_supplicant/include" "-I{compiler.sdk.path}/include/ieee802154/include" "-I{compiler.sdk.path}/include/asio/asio/asio/include" "-I{compiler.sdk.path}/include/asio/port/include" "-I{compiler.sdk.path}/include/bt/common/osi/include" "-I{compiler.sdk.path}/include/bt/include/esp32c3/include" "-I{compiler.sdk.path}/include/bt/common/api/include/api" "-I{compiler.sdk.path}/include/bt/common/btc/profile/esp/blufi/include" "-I{compiler.sdk.path}/include/bt/common/btc/profile/esp/include" "-I{compiler.sdk.path}/include/bt/host/bluedroid/api/include/api" "-I{compiler.sdk.path}/include/cbor/port/include" "-I{compiler.sdk.path}/include/unity/include" "-I{compiler.sdk.path}/include/unity/unity/src" "-I{compiler.sdk.path}/include/cmock/CMock/src" "-I{compiler.sdk.path}/include/coap/port/include" "-I{compiler.sdk.path}/include/coap/libcoap/include" "-I{compiler.sdk.path}/include/console" "-I{compiler.sdk.path}/include/nghttp/port/include" "-I{compiler.sdk.path}/include/nghttp/nghttp2/lib/includes" "-I{compiler.sdk.path}/include/esp-tls" "-I{compiler.sdk.path}/include/esp-tls/esp-tls-crypto" "-I{compiler.sdk.path}/include/esp_adc_cal/include" "-I{compiler.sdk.path}/include/esp_hid/include" "-I{compiler.sdk.path}/include/tcp_transport/include" "-I{compiler.sdk.path}/include/esp_http_client/include" "-I{compiler.sdk.path}/include/esp_http_server/include" "-I{compiler.sdk.path}/include/esp_https_ota/include" "-I{compiler.sdk.path}/include/esp_https_server/include" "-I{compiler.sdk.path}/include/esp_lcd/include" "-I{compiler.sdk.path}/include/esp_lcd/interface" "-I{compiler.sdk.path}/include/protobuf-c/protobuf-c" "-I{compiler.sdk.path}/include/protocomm/include/common" "-I{compiler.sdk.path}/include/protocomm/include/security" "-I{compiler.sdk.path}/include/protocomm/include/transports" "-I{compiler.sdk.path}/include/mdns/include" "-I{compiler.sdk.path}/include/esp_local_ctrl/include" "-I{compiler.sdk.path}/include/sdmmc/include" "-I{compiler.sdk.path}/include/esp_serial_slave_link/include" "-I{compiler.sdk.path}/include/esp_websocket_client/include" "-I{compiler.sdk.path}/include/expat/expat/expat/lib" "-I{compiler.sdk.path}/include/expat/port/include" "-I{compiler.sdk.path}/include/wear_levelling/include" "-I{compiler.sdk.path}/include/fatfs/diskio" "-I{compiler.sdk.path}/include/fatfs/vfs" "-I{compiler.sdk.path}/include/fatfs/src" "-I{compiler.sdk.path}/include/freemodbus/common/include" "-I{compiler.sdk.path}/include/idf_test/include" "-I{compiler.sdk.path}/include/idf_test/include/esp32c3" "-I{compiler.sdk.path}/include/jsmn/include" "-I{compiler.sdk.path}/include/json/cJSON" "-I{compiler.sdk.path}/include/libsodium/libsodium/src/libsodium/include" "-I{compiler.sdk.path}/include/libsodium/port_include" "-I{compiler.sdk.path}/include/mqtt/esp-mqtt/include" "-I{compiler.sdk.path}/include/openssl/include" "-I{compiler.sdk.path}/include/spiffs/include" "-I{compiler.sdk.path}/include/wifi_provisioning/include" "-I{compiler.sdk.path}/include/esp_littlefs/src" "-I{compiler.sdk.path}/include/esp_littlefs/include" "-I{compiler.sdk.path}/include/esp-face/include" "-I{compiler.sdk.path}/include/esp-face/include/tool" "-I{compiler.sdk.path}/include/esp-face/include/typedef" "-I{compiler.sdk.path}/include/esp-face/include/image" "-I{compiler.sdk.path}/include/esp-face/include/math" "-I{compiler.sdk.path}/include/esp-face/include/nn" "-I{compiler.sdk.path}/include/esp-face/include/layer" "-I{compiler.sdk.path}/include/esp-face/include/detect" "-I{compiler.sdk.path}/include/esp-face/include/model_zoo" "-I{compiler.sdk.path}/include/fb_gfx/include"
+compiler.c.elf.libs.esp32c3=-lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lasio -lbt -lcbor -lunity -lcmock -lcoap -lconsole -lnghttp -lesp-tls -lesp_adc_cal -lesp_hid -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lesp_https_server -lesp_lcd -lprotobuf-c -lprotocomm -lmdns -lesp_local_ctrl -lsdmmc -lesp_serial_slave_link -lesp_websocket_client -lexpat -lwear_levelling -lfatfs -lfreemodbus -ljsmn -ljson -llibsodium -lmqtt -lopenssl -lspiffs -lwifi_provisioning -lesp_littlefs -lfb_gfx -lasio -lcbor -lcmock -lunity -lcoap -lesp_lcd -lesp_local_ctrl -lesp_https_server -lesp_websocket_client -lexpat -lfreemodbus -ljsmn -llibsodium -lmqtt -lesp_adc_cal -lesp_hid -lfatfs -lwear_levelling -lopenssl -lspiffs -lwifi_provisioning -lprotocomm -lbt -lbtdm_app -lprotobuf-c -lmdns -lconsole -ljson -lcat_face_detect -lhuman_face_detect -ldl -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lesp_ringbuf -lefuse -lesp_ipc -ldriver -lesp_pm -lmbedtls -lapp_update -lbootloader_support -lspi_flash -lnvs_flash -lpthread -lesp_gdbstub -lespcoredump -lesp_phy -lesp_system -lesp_rom -lhal -lvfs -lesp_eth -ltcpip_adapter -lesp_netif -lesp_event -lwpa_supplicant -lesp_wifi -llwip -llog -lheap -lsoc -lesp_hw_support -lriscv -lesp_common -lesp_timer -lfreertos -lnewlib -lcxx -lapp_trace -lnghttp -lesp-tls -ltcp_transport -lesp_http_client -lesp_http_server -lesp_https_ota -lsdmmc -lesp_serial_slave_link -lmbedtls -lmbedcrypto -lmbedx509 -lcoexist -lcore -lespnow -lmesh -lnet80211 -lpp -lsmartconfig -lwapi -lphy -lbtbb -lesp_phy -lphy -lbtbb -lesp_phy -lphy -lbtbb -lm -lnewlib -lstdc++ -lpthread -lgcc -lcxx -lapp_trace -lgcov -lapp_trace -lgcov -lc 
 compiler.c.flags.esp32c3=-march=rv32imc -ffunction-sections -fdata-sections -Wno-error=unused-function -Wno-error=unused-variable -Wno-error=deprecated-declarations -Wno-unused-parameter -Wno-sign-compare -ggdb -Wno-error=format= -nostartfiles -Wno-format -Og -fstrict-volatile-bitfields -Wno-error=unused-but-set-variable -fno-jump-tables -fno-tree-switch-conversion -std=gnu99 -Wno-old-style-declaration  -MMD -c
 compiler.cpp.flags.esp32c3=-march=rv32imc -ffunction-sections -fdata-sections -Wno-error=unused-function -Wno-error=unused-variable -Wno-error=deprecated-declarations -Wno-unused-parameter -Wno-sign-compare -ggdb -Wno-error=format= -nostartfiles -Wno-format -Og -fstrict-volatile-bitfields -Wno-error=unused-but-set-variable -fno-jump-tables -fno-tree-switch-conversion -std=gnu++11 -fno-exceptions -fno-rtti  -MMD -c
 compiler.S.flags.esp32c3=-ffunction-sections -fdata-sections -Wno-error=unused-function -Wno-error=unused-variable -Wno-error=deprecated-declarations -Wno-unused-parameter -Wno-sign-compare -ggdb -Wno-error=format= -nostartfiles -Wno-format -Og -fstrict-volatile-bitfields -Wno-error=unused-but-set-variable -fno-jump-tables -fno-tree-switch-conversion  -x assembler-with-cpp -MMD -c
diff --git a/tools/platformio-build-esp32.py b/tools/platformio-build-esp32.py
index 083169f60d9..206b17a71ad 100644
--- a/tools/platformio-build-esp32.py
+++ b/tools/platformio-build-esp32.py
@@ -300,7 +300,7 @@
         "UNITY_INCLUDE_CONFIG_H",
         "WITH_POSIX",
         "_GNU_SOURCE",
-        ("IDF_VER", '\\"v4.4-dev-3235-g3e370c4296\\"'),
+        ("IDF_VER", '\\"v4.4-dev-3401-gb86fe0c66c\\"'),
         "ESP_PLATFORM",
         "ARDUINO_ARCH_ESP32",
         "ESP32",
diff --git a/tools/platformio-build-esp32c3.py b/tools/platformio-build-esp32c3.py
index 63680ad42a9..f53295e3267 100644
--- a/tools/platformio-build-esp32c3.py
+++ b/tools/platformio-build-esp32c3.py
@@ -263,6 +263,15 @@
         join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "wifi_provisioning", "include"),
         join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "esp_littlefs", "src"),
         join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "esp_littlefs", "include"),
+        join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "esp-face", "include"),
+        join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "esp-face", "include", "tool"),
+        join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "esp-face", "include", "typedef"),
+        join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "esp-face", "include", "image"),
+        join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "esp-face", "include", "math"),
+        join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "esp-face", "include", "nn"),
+        join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "esp-face", "include", "layer"),
+        join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "esp-face", "include", "detect"),
+        join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "esp-face", "include", "model_zoo"),
         join(FRAMEWORK_DIR, "tools", "sdk", "esp32c3", "include", "fb_gfx", "include"),
         join(FRAMEWORK_DIR, "cores", env.BoardConfig().get("build.core"))
     ],
@@ -273,7 +282,7 @@
     ],
 
     LIBS=[
-        "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lasio", "-lbt", "-lcbor", "-lunity", "-lcmock", "-lcoap", "-lconsole", "-lnghttp", "-lesp-tls", "-lesp_adc_cal", "-lesp_hid", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lesp_https_server", "-lesp_lcd", "-lprotobuf-c", "-lprotocomm", "-lmdns", "-lesp_local_ctrl", "-lsdmmc", "-lesp_serial_slave_link", "-lesp_websocket_client", "-lexpat", "-lwear_levelling", "-lfatfs", "-lfreemodbus", "-ljsmn", "-ljson", "-llibsodium", "-lmqtt", "-lopenssl", "-lspiffs", "-lwifi_provisioning", "-lesp_littlefs", "-lfb_gfx", "-lasio", "-lcbor", "-lcmock", "-lunity", "-lcoap", "-lesp_lcd", "-lesp_local_ctrl", "-lesp_https_server", "-lesp_websocket_client", "-lexpat", "-lfreemodbus", "-ljsmn", "-llibsodium", "-lmqtt", "-lesp_adc_cal", "-lesp_hid", "-lfatfs", "-lwear_levelling", "-lopenssl", "-lspiffs", "-lwifi_provisioning", "-lprotocomm", "-lbt", "-lbtdm_app", "-lprotobuf-c", "-lmdns", "-lconsole", "-ljson", "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lnghttp", "-lesp-tls", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lsdmmc", "-lesp_serial_slave_link", "-lmbedtls", "-lmbedcrypto", "-lmbedx509", "-lcoexist", "-lcore", "-lespnow", "-lmesh", "-lnet80211", "-lpp", "-lsmartconfig", "-lwapi", "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lnghttp", "-lesp-tls", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lsdmmc", "-lesp_serial_slave_link", "-lmbedtls", "-lmbedcrypto", "-lmbedx509", "-lcoexist", "-lcore", "-lespnow", "-lmesh", "-lnet80211", "-lpp", "-lsmartconfig", "-lwapi", "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lnghttp", "-lesp-tls", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lsdmmc", "-lesp_serial_slave_link", "-lmbedtls", "-lmbedcrypto", "-lmbedx509", "-lcoexist", "-lcore", "-lespnow", "-lmesh", "-lnet80211", "-lpp", "-lsmartconfig", "-lwapi", "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lnghttp", "-lesp-tls", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lsdmmc", "-lesp_serial_slave_link", "-lmbedtls", "-lmbedcrypto", "-lmbedx509", "-lcoexist", "-lcore", "-lespnow", "-lmesh", "-lnet80211", "-lpp", "-lsmartconfig", "-lwapi", "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lnghttp", "-lesp-tls", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lsdmmc", "-lesp_serial_slave_link", "-lmbedtls", "-lmbedcrypto", "-lmbedx509", "-lcoexist", "-lcore", "-lespnow", "-lmesh", "-lnet80211", "-lpp", "-lsmartconfig", "-lwapi", "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lnghttp", "-lesp-tls", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lsdmmc", "-lesp_serial_slave_link", "-lmbedtls", "-lmbedcrypto", "-lmbedx509", "-lcoexist", "-lcore", "-lespnow", "-lmesh", "-lnet80211", "-lpp", "-lsmartconfig", "-lwapi", "-lphy", "-lbtbb", "-lesp_phy", "-lphy", "-lbtbb", "-lesp_phy", "-lphy", "-lbtbb", "-lm", "-lnewlib", "-lstdc++", "-lpthread", "-lgcc", "-lcxx", "-lapp_trace", "-lgcov", "-lapp_trace", "-lgcov", "-lc"
+        "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lasio", "-lbt", "-lcbor", "-lunity", "-lcmock", "-lcoap", "-lconsole", "-lnghttp", "-lesp-tls", "-lesp_adc_cal", "-lesp_hid", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lesp_https_server", "-lesp_lcd", "-lprotobuf-c", "-lprotocomm", "-lmdns", "-lesp_local_ctrl", "-lsdmmc", "-lesp_serial_slave_link", "-lesp_websocket_client", "-lexpat", "-lwear_levelling", "-lfatfs", "-lfreemodbus", "-ljsmn", "-ljson", "-llibsodium", "-lmqtt", "-lopenssl", "-lspiffs", "-lwifi_provisioning", "-lesp_littlefs", "-lfb_gfx", "-lasio", "-lcbor", "-lcmock", "-lunity", "-lcoap", "-lesp_lcd", "-lesp_local_ctrl", "-lesp_https_server", "-lesp_websocket_client", "-lexpat", "-lfreemodbus", "-ljsmn", "-llibsodium", "-lmqtt", "-lesp_adc_cal", "-lesp_hid", "-lfatfs", "-lwear_levelling", "-lopenssl", "-lspiffs", "-lwifi_provisioning", "-lprotocomm", "-lbt", "-lbtdm_app", "-lprotobuf-c", "-lmdns", "-lconsole", "-ljson", "-lcat_face_detect", "-lhuman_face_detect", "-ldl", "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lnghttp", "-lesp-tls", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lsdmmc", "-lesp_serial_slave_link", "-lmbedtls", "-lmbedcrypto", "-lmbedx509", "-lcoexist", "-lcore", "-lespnow", "-lmesh", "-lnet80211", "-lpp", "-lsmartconfig", "-lwapi", "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lnghttp", "-lesp-tls", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lsdmmc", "-lesp_serial_slave_link", "-lmbedtls", "-lmbedcrypto", "-lmbedx509", "-lcoexist", "-lcore", "-lespnow", "-lmesh", "-lnet80211", "-lpp", "-lsmartconfig", "-lwapi", "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lnghttp", "-lesp-tls", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lsdmmc", "-lesp_serial_slave_link", "-lmbedtls", "-lmbedcrypto", "-lmbedx509", "-lcoexist", "-lcore", "-lespnow", "-lmesh", "-lnet80211", "-lpp", "-lsmartconfig", "-lwapi", "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lnghttp", "-lesp-tls", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lsdmmc", "-lesp_serial_slave_link", "-lmbedtls", "-lmbedcrypto", "-lmbedx509", "-lcoexist", "-lcore", "-lespnow", "-lmesh", "-lnet80211", "-lpp", "-lsmartconfig", "-lwapi", "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lnghttp", "-lesp-tls", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lsdmmc", "-lesp_serial_slave_link", "-lmbedtls", "-lmbedcrypto", "-lmbedx509", "-lcoexist", "-lcore", "-lespnow", "-lmesh", "-lnet80211", "-lpp", "-lsmartconfig", "-lwapi", "-lesp_ringbuf", "-lefuse", "-lesp_ipc", "-ldriver", "-lesp_pm", "-lmbedtls", "-lapp_update", "-lbootloader_support", "-lspi_flash", "-lnvs_flash", "-lpthread", "-lesp_gdbstub", "-lespcoredump", "-lesp_phy", "-lesp_system", "-lesp_rom", "-lhal", "-lvfs", "-lesp_eth", "-ltcpip_adapter", "-lesp_netif", "-lesp_event", "-lwpa_supplicant", "-lesp_wifi", "-llwip", "-llog", "-lheap", "-lsoc", "-lesp_hw_support", "-lriscv", "-lesp_common", "-lesp_timer", "-lfreertos", "-lnewlib", "-lcxx", "-lapp_trace", "-lnghttp", "-lesp-tls", "-ltcp_transport", "-lesp_http_client", "-lesp_http_server", "-lesp_https_ota", "-lsdmmc", "-lesp_serial_slave_link", "-lmbedtls", "-lmbedcrypto", "-lmbedx509", "-lcoexist", "-lcore", "-lespnow", "-lmesh", "-lnet80211", "-lpp", "-lsmartconfig", "-lwapi", "-lphy", "-lbtbb", "-lesp_phy", "-lphy", "-lbtbb", "-lesp_phy", "-lphy", "-lbtbb", "-lm", "-lnewlib", "-lstdc++", "-lpthread", "-lgcc", "-lcxx", "-lapp_trace", "-lgcov", "-lapp_trace", "-lgcov", "-lc"
     ],
 
     CPPDEFINES=[
@@ -282,7 +291,7 @@
         "UNITY_INCLUDE_CONFIG_H",
         "WITH_POSIX",
         "_GNU_SOURCE",
-        ("IDF_VER", '\\"v4.4-dev-3235-g3e370c4296\\"'),
+        ("IDF_VER", '\\"v4.4-dev-3401-gb86fe0c66c\\"'),
         "ESP_PLATFORM",
         "ARDUINO_ARCH_ESP32",
         "ESP32",
diff --git a/tools/platformio-build-esp32s2.py b/tools/platformio-build-esp32s2.py
index dac966a52b0..393cc493630 100644
--- a/tools/platformio-build-esp32s2.py
+++ b/tools/platformio-build-esp32s2.py
@@ -286,7 +286,7 @@
         "UNITY_INCLUDE_CONFIG_H",
         "WITH_POSIX",
         "_GNU_SOURCE",
-        ("IDF_VER", '\\"v4.4-dev-3235-g3e370c4296\\"'),
+        ("IDF_VER", '\\"v4.4-dev-3401-gb86fe0c66c\\"'),
         "ESP_PLATFORM",
         "ARDUINO_ARCH_ESP32",
         "ESP32",
diff --git a/tools/sdk/esp32/include/asio/port/include/esp_asio_config.h b/tools/sdk/esp32/include/asio/port/include/esp_asio_config.h
index cba316527e6..3f3a9b03ed4 100644
--- a/tools/sdk/esp32/include/asio/port/include/esp_asio_config.h
+++ b/tools/sdk/esp32/include/asio/port/include/esp_asio_config.h
@@ -18,6 +18,11 @@
 #  define ASIO_NO_TYPEID
 # endif   // CONFIG_COMPILER_RTTI
 
+//
+// Supress OpenSSL deprecation warning, when building ASIO
+//
+#define ESP_OPENSSL_SUPPRESS_LEGACY_WARNING
+
 //
 // LWIP compatibility inet and address macros/functions
 //
diff --git a/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_block_internal.h b/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_block_internal.h
index 9abe81557fa..b7ad0a554cc 100644
--- a/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_block_internal.h
+++ b/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_block_internal.h
@@ -191,6 +191,9 @@ int coap_handle_response_get_block(coap_context_t *context,
 void coap_block_delete_lg_xmit(coap_session_t *session,
                                coap_lg_xmit_t *lg_xmit);
 
+coap_tick_t coap_block_check_lg_xmit_timeouts(coap_session_t *session,
+                                              coap_tick_t now);
+
 /**
  * The function that does all the work for the coap_add_data_large*()
  * functions.
diff --git a/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_dtls.h b/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_dtls.h
index cbd369dfce6..fc30445431d 100644
--- a/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_dtls.h
+++ b/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_dtls.h
@@ -27,6 +27,12 @@ typedef struct coap_dtls_pki_t coap_dtls_pki_t;
 #ifndef COAP_DTLS_HINT_LENGTH
 #define COAP_DTLS_HINT_LENGTH 128
 #endif
+#ifndef COAP_DTLS_MAX_PSK_IDENTITY
+#define COAP_DTLS_MAX_PSK_IDENTITY 64
+#endif
+#ifndef COAP_DTLS_MAX_PSK
+#define COAP_DTLS_MAX_PSK 64
+#endif
 
 typedef enum coap_dtls_role_t {
   COAP_DTLS_ROLE_CLIENT, /**< Internal function invoked for client */
diff --git a/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_event.h b/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_event.h
index 89b2a63967c..b6ea60524a2 100644
--- a/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_event.h
+++ b/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_event.h
@@ -24,34 +24,34 @@
  * Scalar type to represent different events, e.g. DTLS events or
  * retransmission timeouts.
  */
- typedef unsigned int coap_event_t;
-
+typedef enum coap_event_t {
 /**
  * (D)TLS events for COAP_PROTO_DTLS and COAP_PROTO_TLS
  */
-#define COAP_EVENT_DTLS_CLOSED        0x0000
-#define COAP_EVENT_DTLS_CONNECTED     0x01DE
-#define COAP_EVENT_DTLS_RENEGOTIATE   0x01DF
-#define COAP_EVENT_DTLS_ERROR         0x0200
+  COAP_EVENT_DTLS_CLOSED       = 0x0000,
+  COAP_EVENT_DTLS_CONNECTED    = 0x01DE,
+  COAP_EVENT_DTLS_RENEGOTIATE  = 0x01DF,
+  COAP_EVENT_DTLS_ERROR        = 0x0200,
 
 /**
  * TCP events for COAP_PROTO_TCP and COAP_PROTO_TLS
  */
-#define COAP_EVENT_TCP_CONNECTED      0x1001
-#define COAP_EVENT_TCP_CLOSED         0x1002
-#define COAP_EVENT_TCP_FAILED         0x1003
+  COAP_EVENT_TCP_CONNECTED     = 0x1001,
+  COAP_EVENT_TCP_CLOSED        = 0x1002,
+  COAP_EVENT_TCP_FAILED        = 0x1003,
 
 /**
  * CSM exchange events for reliable protocols only
  */
-#define COAP_EVENT_SESSION_CONNECTED  0x2001
-#define COAP_EVENT_SESSION_CLOSED     0x2002
-#define COAP_EVENT_SESSION_FAILED     0x2003
+  COAP_EVENT_SESSION_CONNECTED = 0x2001,
+  COAP_EVENT_SESSION_CLOSED    = 0x2002,
+  COAP_EVENT_SESSION_FAILED    = 0x2003,
 
 /**
- * BLOCK2 receive errors
+ * (Q-)BLOCK receive errors
  */
-#define COAP_EVENT_PARTIAL_BLOCK      0x3001
+  COAP_EVENT_PARTIAL_BLOCK     = 0x3001
+} coap_event_t;
 
 /**
  * Type for event handler functions that can be registered with a CoAP
diff --git a/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_time.h b/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_time.h
index baa8650eaff..99d117a4eb7 100644
--- a/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_time.h
+++ b/tools/sdk/esp32/include/coap/libcoap/include/coap3/coap_time.h
@@ -88,7 +88,11 @@ COAP_STATIC_INLINE uint64_t coap_ticks_to_rt_us(coap_tick_t t) {
 #elif defined(RIOT_VERSION)
 #include <xtimer.h>
 
+#ifdef XTIMER_HZ
 #define COAP_TICKS_PER_SECOND (XTIMER_HZ)
+#else /* XTIMER_HZ */
+#define COAP_TICKS_PER_SECOND (XTIMER_HZ_BASE)
+#endif /* XTIMER_HZ */
 
 typedef uint64_t coap_tick_t;
 typedef int64_t coap_tick_diff_t;
diff --git a/tools/sdk/esp32/include/coap/libcoap/include/coap3/net.h b/tools/sdk/esp32/include/coap/libcoap/include/coap3/net.h
index 577a0b5d5a5..ea5a2cba1bb 100644
--- a/tools/sdk/esp32/include/coap/libcoap/include/coap3/net.h
+++ b/tools/sdk/esp32/include/coap/libcoap/include/coap3/net.h
@@ -15,6 +15,7 @@
 #include <stdlib.h>
 #include <string.h>
 #ifndef _WIN32
+#include <sys/select.h>
 #include <sys/time.h>
 #endif
 #include <time.h>
diff --git a/tools/sdk/esp32/include/coap/libcoap/include/coap3/pdu.h b/tools/sdk/esp32/include/coap/libcoap/include/coap3/pdu.h
index a22869b69ed..8031a1c2c40 100644
--- a/tools/sdk/esp32/include/coap/libcoap/include/coap3/pdu.h
+++ b/tools/sdk/esp32/include/coap/libcoap/include/coap3/pdu.h
@@ -299,7 +299,6 @@ typedef enum coap_pdu_code_t {
   COAP_REQUEST_CODE_PATCH  = COAP_REQUEST_PATCH,
   COAP_REQUEST_CODE_IPATCH = COAP_REQUEST_IPATCH,
 
-  COAP_RESPONSE_CODE_OK                         = COAP_RESPONSE_CODE(200),
   COAP_RESPONSE_CODE_CREATED                    = COAP_RESPONSE_CODE(201),
   COAP_RESPONSE_CODE_DELETED                    = COAP_RESPONSE_CODE(202),
   COAP_RESPONSE_CODE_VALID                      = COAP_RESPONSE_CODE(203),
diff --git a/tools/sdk/esp32/include/coap/libcoap/include/coap3/resource.h b/tools/sdk/esp32/include/coap/libcoap/include/coap3/resource.h
index 2cd9aea48fa..5cf7751f67e 100644
--- a/tools/sdk/esp32/include/coap/libcoap/include/coap3/resource.h
+++ b/tools/sdk/esp32/include/coap/libcoap/include/coap3/resource.h
@@ -83,7 +83,8 @@ typedef void (*coap_method_handler_t)
  * variable of coap_str_const_t has to point to constant text, or point to data
  * within the allocated coap_str_const_t parameter.
  *
- * @param uri_path The string URI path of the new resource.
+ * @param uri_path The string URI path of the new resource. The leading '/' is
+ *                 not normally required - e.g. just "full/path/for/resource".
  * @param flags    Flags for memory management (in particular release of
  *                 memory). Possible values:@n
  *
diff --git a/tools/sdk/esp32/include/config/sdkconfig.h b/tools/sdk/esp32/include/config/sdkconfig.h
index b57d1022057..084b3a554b5 100644
--- a/tools/sdk/esp32/include/config/sdkconfig.h
+++ b/tools/sdk/esp32/include/config/sdkconfig.h
@@ -26,6 +26,7 @@
 #define CONFIG_BOOTLOADER_FLASH_XMC_SUPPORT 1
 #define CONFIG_ESPTOOLPY_BAUD_OTHER_VAL 115200
 #define CONFIG_ESPTOOLPY_FLASHMODE_DIO 1
+#define CONFIG_ESPTOOLPY_FLASH_SAMPLE_MODE_STR 1
 #define CONFIG_ESPTOOLPY_FLASHMODE "dio"
 #define CONFIG_ESPTOOLPY_FLASHFREQ_40M 1
 #define CONFIG_ESPTOOLPY_FLASHFREQ "40m"
@@ -369,6 +370,7 @@
 #define CONFIG_LWIP_GARP_TMR_INTERVAL 60
 #define CONFIG_LWIP_TCPIP_RECVMBOX_SIZE 32
 #define CONFIG_LWIP_DHCP_RESTORE_LAST_IP 1
+#define CONFIG_LWIP_DHCP_OPTIONS_LEN 68
 #define CONFIG_LWIP_DHCPS 1
 #define CONFIG_LWIP_DHCPS_LEASE_UNIT 60
 #define CONFIG_LWIP_DHCPS_MAX_STATION_NUM 8
@@ -480,6 +482,7 @@
 #define CONFIG_MDNS_TASK_AFFINITY 0x0
 #define CONFIG_MDNS_SERVICE_ADD_TIMEOUT_MS 2000
 #define CONFIG_MDNS_TIMER_PERIOD_MS 100
+#define CONFIG_MDNS_MULTIPLE_INSTANCE 1
 #define CONFIG_MQTT_PROTOCOL_311 1
 #define CONFIG_MQTT_TRANSPORT_SSL 1
 #define CONFIG_MQTT_TRANSPORT_WEBSOCKET 1
@@ -675,5 +678,5 @@
 #define CONFIG_ULP_COPROC_ENABLED CONFIG_ESP32_ULP_COPROC_ENABLED
 #define CONFIG_ULP_COPROC_RESERVE_MEM CONFIG_ESP32_ULP_COPROC_RESERVE_MEM
 #define CONFIG_WARN_WRITE_STRINGS CONFIG_COMPILER_WARN_WRITE_STRINGS
-#define CONFIG_ARDUINO_IDF_COMMIT "3e370c4296"
+#define CONFIG_ARDUINO_IDF_COMMIT "b86fe0c66c"
 #define CONFIG_ARDUINO_IDF_BRANCH "master"
diff --git a/tools/sdk/esp32/include/driver/include/driver/rmt.h b/tools/sdk/esp32/include/driver/include/driver/rmt.h
index a7e2aad5f24..bc07954a080 100644
--- a/tools/sdk/esp32/include/driver/include/driver/rmt.h
+++ b/tools/sdk/esp32/include/driver/include/driver/rmt.h
@@ -856,16 +856,35 @@ esp_err_t rmt_remove_channel_from_group(rmt_channel_t channel);
 
 #if SOC_RMT_SUPPORT_TX_LOOP_COUNT
 /**
- * @brief Set loop count for RMT TX channel
+ * @brief Set loop count threshold value for RMT TX channel
+ *
+ * When tx loop count reaches this value, an ISR callback will notify user
  *
  * @param channel RMT channel
- * @param count loop count
+ * @param count loop count, 1 ~ 1023
  * @return
  *      - ESP_ERR_INVALID_ARG Parameter error
  *      - ESP_OK Success
  */
 esp_err_t rmt_set_tx_loop_count(rmt_channel_t channel, uint32_t count);
-#endif
+
+/**
+ * @brief Enable or disable the feature that when loop count reaches the threshold, RMT will stop transmitting.
+ *
+ * - When the loop auto-stop feature is enabled will halt RMT transmission after the loop count reaches a certain threshold
+ * - When disabled, the RMT transmission continue indefinitely until halted by the users
+ *
+ * @note The auto-stop feature is implemented in hardware on particular targets (i.e. those with SOC_RMT_SUPPORT_TX_LOOP_AUTOSTOP defined).
+ *       Otherwise, the auto-stop feature is implemented in software via the interrupt.
+ *
+ * @param channel RMT channel
+ * @param en enable bit
+ * @return
+ *      - ESP_ERR_INVALID_ARG Parameter error
+ *      - ESP_OK Success
+ */
+esp_err_t rmt_enable_tx_loop_autostop(rmt_channel_t channel, bool en);
+#endif // SOC_RMT_SUPPORT_TX_LOOP_COUNT
 
 /**
 * @brief Reset RMT TX/RX memory index.
diff --git a/tools/sdk/esp32/include/esp-face/include/dl_define.hpp b/tools/sdk/esp32/include/esp-face/include/dl_define.hpp
index 16eb0881562..7809768a5bc 100644
--- a/tools/sdk/esp32/include/esp-face/include/dl_define.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/dl_define.hpp
@@ -10,7 +10,7 @@
 #define DL_LOG_LAYER_LATENCY 0 /*<! - 1: print the latency of each parts of layer */
                                /*<! - 0: mute */
 
-#if CONFIG_SPIRAM_SUPPORT || CONFIG_ESP32_SPIRAM_SUPPORT || CONFIG_ESP32S3_SPIRAM_SUPPORT
+#if CONFIG_SPIRAM_SUPPORT || CONFIG_ESP32_SPIRAM_SUPPORT || CONFIG_ESP32S2_SPIRAM_SUPPORT || CONFIG_ESP32S3_SPIRAM_SUPPORT
 #define DL_SPIRAM_SUPPORT 1
 #else
 #define DL_SPIRAM_SUPPORT 0
@@ -83,8 +83,17 @@ namespace dl
 
     typedef enum
     {
-        PADDING_VALID,     /*<! no padding >*/
-        PADDING_SAME,      /*<! SAME in TensorFlow style >*/
-        PADDING_SAME_MXNET /*<! SAME in MXNET style >*/
+        PADDING_NOT_SET,
+        PADDING_VALID,      /*<! no padding >*/
+        PADDING_SAME_BEGIN,  /*<! SAME in MXNET style >*/
+        PADDING_SAME_END,   /*<! SAME in TensorFlow style >*/
     } padding_type_t;
-} // namespace dl
\ No newline at end of file
+    
+    typedef enum
+    {
+        CONSTANT,
+        EDGE, 
+        REFLECT,
+        SYMMETRIC,
+    } padding_mode_t;
+} // namespace dl
diff --git a/tools/sdk/esp32/include/esp-face/include/image/dl_image.hpp b/tools/sdk/esp32/include/esp-face/include/image/dl_image.hpp
index c7fecdc49cc..4a974df063a 100644
--- a/tools/sdk/esp32/include/esp-face/include/image/dl_image.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/image/dl_image.hpp
@@ -370,11 +370,70 @@ namespace dl
          */
         uint32_t get_moving_point_number(uint8_t *f1, uint8_t *f2, const uint32_t height, const uint32_t width, const uint32_t stride, const uint32_t threshold = 5);
 
-
+        /**
+         * @brief Apply an affine transformation to an image.
+         * 
+         * @tparam T 
+         * @param input     the input image.
+         * @param output    the output image.
+         * @param M_inv     the inverse transformation matrix.
+         */
         template <typename T>
         void warp_affine(dl::Tensor<T> *input, dl::Tensor<T> *output, dl::math::Matrix<float> *M_inv);
+
+        /**
+         * @brief Apply an affine transformation to an image.
+         * 
+         * @tparam T 
+         * @param input    the pointer of the input image.
+         * @param shape    the shape of the input image.
+         * @param output   the output image.
+         * @param M_inv    the inverse transformation matrix.
+         */
         template <typename T>
         void warp_affine(uint16_t *input, std::vector<int> shape, dl::Tensor<T> *output, dl::math::Matrix<float> *M_inv);
 
+        /**
+         * @brief Get the otsu thresh object.
+         * 
+         * @param image  the gray image.
+         * @return uint8_t the otsu thresh.
+         */
+        uint8_t get_otsu_thresh(Tensor<uint8_t> &image);
+
+        /**
+         * @brief Convert RGB image to gray image
+         * 
+         * @param image  input image
+         * @param bgr    true: the image is in BGR format
+         *               false: the image is in RGB format
+         * @return Tensor<uint8_t>* output image in gray format
+         */
+        Tensor<uint8_t> *rgb2gray(Tensor<uint8_t> &image, bool bgr = false);
+
+        /**
+         * @brief Convert RGB image to LAB image
+         * 
+         * @param image  input image
+         * @param bgr    true: the image is in BGR format
+         *               false: the image is in RGB format 
+         * @param fast   true: use the fast alogrithm， but the accuracy will be reduced
+         *               false: do not use the fast alogrithm
+         * @return Tensor<uint8_t>* output image in LAB foramt
+         */
+        Tensor<uint8_t> *rgb2lab(Tensor<uint8_t> &image, bool bgr = false, bool fast = true);
+
+        /**
+         * @brief Convert RGB image to HSV image
+         * 
+         * @param image   input image
+         * @param bgr     true: the image is in BGR format
+         *                false: the image is in RGB format 
+         * @param fast    true: use the fast alogrithm， but the accuracy will be reduced
+         *                false: do not use the fast alogrithm
+         * @return Tensor<uint8_t>* output image in HSV format
+         */
+        Tensor<uint8_t> *rgb2hsv(Tensor<uint8_t> &image, bool bgr = false, bool fast = true);
+
     } // namespace image
 } // namespace dl
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_add2d.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_add2d.hpp
index f0c5964b3d1..c43282b42de 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_add2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_add2d.hpp
@@ -25,7 +25,8 @@ namespace dl
             const int output_exponent;               /*<! exponent of output >*/
             Tensor<feature_t> *output;               /*<! output ptr of add2d >*/
             bool inplace;                            /*<! true: the output will store to input0
-                                                          false: the output will store to a seperate memeory >*/
+                                                          false: the output will store to a separate memory >*/
+            std::vector<int> output_shape;           /*<! output shape of add2d >*/
 
         public:
             /**
@@ -35,19 +36,21 @@ namespace dl
              * @param activation      activation of add2d, if you don't specify anything, no activation is applied
              * @param name            name of add2d
              * @param inplace         true: the output will store to input0
-             *                        false: the output will store to a seperate memeory
+             *                        false: the output will store to a separate memory
              */
-            Add2D(const int output_exponent, const Activation<feature_t> *activation = NULL, const char *name = NULL, bool inplace = false) : Layer(name), activation(activation), output_exponent(output_exponent), output(NULL)
-            {
-                this->inplace = inplace;
-            }
+            Add2D(const int output_exponent, const Activation<feature_t> *activation = NULL, const char *name = "Add2D", bool inplace = false) : Layer(name),
+                                                                                                                                                 activation(activation),
+                                                                                                                                                 output_exponent(output_exponent),
+                                                                                                                                                 output(NULL),
+                                                                                                                                                 inplace(inplace),
+                                                                                                                                                 output_shape({}) {}
 
             /**
              * @brief Destroy the Add2D object
              */
             ~Add2D()
             {
-                if((!this->inplace) && (this->output != NULL))
+                if ((!this->inplace) && (this->output != NULL))
                 {
                     delete this->output;
                 }
@@ -59,10 +62,12 @@ namespace dl
              * 
              * @param input0 as one input
              * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1)
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
             {
                 assert(input0.is_same_shape(input1));
+                this->output_shape = input0.shape;
 
                 if (!this->inplace)
                 {
@@ -78,6 +83,11 @@ namespace dl
                 {
                     this->output = &input0;
                 }
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -105,7 +115,11 @@ namespace dl
                 if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output->apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
                     this->output->set_exponent(this->output_exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -116,6 +130,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::add2d(*this->output, input0, input1, this->activation, assign_core, this->output_exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "add2d");
                 }
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_avg_pool2d.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_avg_pool2d.hpp
index f78b262beb9..8a9aaa8dfbe 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_avg_pool2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_avg_pool2d.hpp
@@ -24,23 +24,26 @@ namespace dl
             std::vector<int> filter_shape;     /*<! filter shape in [filter_height, filter_width] >*/
             const int stride_y;                /*<! stride in height >*/
             const int stride_x;                /*<! stride in width >*/
-            const padding_type_t padding_type; /*<! one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET >*/
+            const padding_type_t padding_type; /*<! one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN >*/
             std::vector<int> padding;          /*<! padding size needed in [top, bottom, left, right] of this operation >*/
-            Tensor<feature_t> *output;          /*<! output ptr of AvgPool2D >*/
+            Tensor<feature_t> *output;         /*<! output ptr of AvgPool2D >*/
+            std::vector<int> output_shape;     /*<! output shape of AvgPool2D >*/
 
         public:
-
             /**
              * @brief Construct a new AvgPool2D object.
              * 
              * @param output_exponent exponent of output
              * @param filter_shape    filter shape in [filter_height, filter_width]
-             * @param padding_type    one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+             * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN or PADDING_NOT_SET,
              *                        - PADDING_VALID means no padding
-             *                        PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+             *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
              *                        such that output has the same height/width dimension as the input,
-             *                        - PADDING_SAME results padding in TensorFlow style
-             *                        - PADDING_SAME_MXNET results padding in MXNET style
+             *                        - PADDING_SAME_END results padding in TensorFlow style
+             *                        - PADDING_SAME_BEGIN results padding in MXNET style
+             *                        - PADDING_NOT_SET means padding with the specific "padding" value below. 
+             * @param padding         if padding_type is PADDING_NOT_SET, this value will be used as padding size. 
+             *                        the shape must be 4, the value of each position is: [padding top, padding bottom, padding left, padding right]
              * @param stride_y        stride in height
              * @param stride_x        stride in width
              * @param name            name of layer
@@ -48,16 +51,23 @@ namespace dl
             AvgPool2D(const int output_exponent,
                       const std::vector<int> filter_shape,
                       const padding_type_t padding_type = PADDING_VALID,
+                      std::vector<int> padding = {},
                       const int stride_y = 1,
                       const int stride_x = 1,
-                      const char *name = NULL) : Layer(name),
-                                                 output_exponent(output_exponent),
-                                                 filter_shape(filter_shape),
-                                                 stride_y(stride_y),
-                                                 stride_x(stride_x),
-                                                 padding_type(padding_type)
+                      const char *name = "AvgPool2D") : Layer(name),
+                                                        output_exponent(output_exponent),
+                                                        filter_shape(filter_shape),
+                                                        padding_type(padding_type),
+                                                        padding(padding),
+                                                        stride_y(stride_y),
+                                                        stride_x(stride_x),
+                                                        output_shape({})
             {
                 this->output = new Tensor<feature_t>;
+                if (this->padding_type == PADDING_NOT_SET)
+                {
+                    assert(this->padding.size() == 4);
+                }
             }
 
             /**
@@ -66,7 +76,7 @@ namespace dl
             */
             ~AvgPool2D()
             {
-                if(this->output != NULL)
+                if (this->output != NULL)
                 {
                     delete this->output;
                 }
@@ -76,20 +86,31 @@ namespace dl
              * @brief Update output shape and padding.
              * 
              * @param input as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
                 assert(input.shape[0] > 0);
                 assert(input.shape[1] > 0);
-                std::vector<int> output_shape = nn::get_output_shape(input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
-                this->output->set_shape(output_shape);
+                assert(input.shape.size() == 3);
+
+                this->output_shape = nn::get_output_shape(input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type, false, this->padding);
+                this->output->set_shape(this->output_shape);
                 this->output->set_exponent(this->output_exponent);
 
-                this->padding = nn::get_pad_size(output_shape, input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
-                input.set_padding_size(this->padding);
+                if (this->padding_type != PADDING_NOT_SET)
+                {
+                    this->padding = nn::get_pad_size(this->output_shape, input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
+                }
+
                 this->output->free_element();
-            }
 
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
 
             /**
              * @brief Get the output
@@ -108,7 +129,6 @@ namespace dl
              * @param autoload_enable one of true or false, 
              *                        - true: load input and output from PSRAM to CACHE automatically
              *                        - false: do not
-             * @param assign_core     not effective yet
              * @return AvgPool2D result
              */
             Tensor<feature_t> &call(Tensor<feature_t> &input, uint8_t autoload_enable = 0)
@@ -116,7 +136,11 @@ namespace dl
                 DL_LOG_LAYER_LATENCY_INIT();
 
                 DL_LOG_LAYER_LATENCY_START();
-                this->output->apply_element();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
                 this->output->set_exponent(this->output_exponent);
                 DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_base.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_base.hpp
index 5c7d28d52b1..b265b454538 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_base.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_base.hpp
@@ -1,6 +1,7 @@
 #pragma once
 #include "dl_tool.hpp"
 #include "dl_tool_cache.hpp"
+#include <iostream>
 
 namespace dl
 {
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_concat.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_concat.hpp
new file mode 100644
index 00000000000..35ebe652e53
--- /dev/null
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_concat.hpp
@@ -0,0 +1,139 @@
+#pragma once
+
+#include <assert.h>
+#include <vector>
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+#include "dl_nn_concat.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Concat(input1, input2, input3, ...).
+         * 
+         * @tparam feature_t support all kinds of integer and float data type
+         */
+        template <typename feature_t>
+        class Concat : Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            int axis;                      /*<! The axis along which the Tensor will be concatenated. >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Concat >*/
+            std::vector<int> output_shape; /*<! output shape of Concat >*/
+        public:
+            /**
+             * @brief Construct a new Concat object.
+             * 
+             * @param name name of layer
+             * @param axis The axis along which the Tensor will be concatenated.
+             */
+            Concat(int axis, const char *name = "Concat") : Layer(name), axis(axis), output_shape({})
+            {
+                this->output = new Tensor<feature_t>;
+            }
+
+            /**
+             * @brief Destroy the Concat object
+             */
+            ~Concat()
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Collect inputs' channel and memory offset, called in Model.build().
+             * 
+             * @param args pointers of concatenated Tensor
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(std::vector<Tensor<feature_t> *> args, bool print_shape = false)
+            {
+                assert(args.size() > 1);
+                int shape_size = args[0]->shape.size();
+
+                if (this->axis < 0)
+                {
+                    this->axis = shape_size + this->axis;
+                }
+                assert((this->axis < shape_size) && (this->axis > -1));
+
+                int output_shape_axis = args[0]->shape[this->axis];
+
+                for (int i = 1; i < args.size(); i++)
+                {
+                    assert(shape_size == args[i]->shape.size());
+                    assert(args[i]->exponent == args[i - 1]->exponent);
+                    output_shape_axis += args[i]->shape[this->axis];
+
+                    for (int j = 0; j < shape_size; j++)
+                    {
+                        if (j != this->axis)
+                        {
+                            assert(args[i]->shape[j] == args[i - 1]->shape[j]);
+                        }
+                    }
+                }
+
+                this->output_exponent = args[0]->exponent;
+                this->output_shape = args[0]->shape;
+                this->output_shape[this->axis] = output_shape_axis;
+
+                this->output->set_shape(this->output_shape);
+                this->output->set_exponent(this->output_exponent);
+                this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Call Concat operation
+             * 
+             * @param inputs                the pointers of inputs 
+             * @param free_inputs           true: free the inputs after call
+             *                              false: do not free inputs
+             * @return Tensor<feature_t>&   concat result
+             */
+            Tensor<feature_t> &call(std::vector<Tensor<feature_t> *> inputs, bool free_inputs = false)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
+                this->output->set_exponent(this->output_exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                DL_LOG_LAYER_LATENCY_START();
+                nn::concat(*this->output, inputs, this->axis, free_inputs);
+                DL_LOG_LAYER_LATENCY_END(this->name, "concat");
+                return *this->output;
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>&  Concat result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_conv2d.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_conv2d.hpp
index a7c2229db09..038dd6cd79f 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_conv2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_conv2d.hpp
@@ -13,8 +13,11 @@ namespace dl
          * @tparam feature_t supports int16_t and int8_t,
          *         - int16_t: stands for operation in int16_t quantize
          *         - int8_t: stands for operation in int8_t quantize
+         * @tparam bias_t supports int16_t and int8_t, must specify when using int8 per-channel quantization
+         *         - int16_t: for int16 quantization and int8 per-channel quantization
+         *         - int8_t: for int8 per-tensor quantization
          */
-        template <typename feature_t>
+        template <typename feature_t, typename bias_t = feature_t>
         class Conv2D : public Layer
         {
         private:
@@ -22,14 +25,14 @@ namespace dl
             const Filter<feature_t> *filter;         /*<! filter of Conv2D >*/
             const int stride_y;                      /*<! stride in height >*/
             const int stride_x;                      /*<! stride in width >*/
-            const padding_type_t padding_type;       /*<! one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET >*/
-            const Bias<feature_t> *bias;             /*<! bias of Conv2D, if you don't specify anything, no bias is added >*/
+            const padding_type_t padding_type;       /*<! one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN >*/
+            const Bias<bias_t> *bias;                /*<! bias of Conv2D, if you don't specify anything, no bias is added >*/
             const Activation<feature_t> *activation; /*<! activation of Conv2D, if you don't specify anything, no activation is applied >*/
             std::vector<int> padding;                /*<! padding size needed in [top, bottom, left, right] of this operation >*/
-            Tensor<feature_t> *output;              /*<! output ptr of Conv2D >*/
+            Tensor<feature_t> *output;               /*<! output ptr of Conv2D >*/
+            std::vector<int> output_shape;           /*<! output shape of Conv2D >*/
 
         public:
-
             /**
              * @brief Construct a new Conv2D object.
              * 
@@ -37,33 +40,43 @@ namespace dl
              * @param filter          filter of Conv2D
              * @param bias            bias of Conv2D, if you don't specify anything, no bias is added
              * @param activation      activation of Conv2D, if you don't specify anything, no activation is applied
-             * @param padding_type    one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+             * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN or PADDING_NOT_SET,
              *                        - PADDING_VALID means no padding
-             *                        PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+             *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
              *                        such that output has the same height/width dimension as the input,
-             *                        - PADDING_SAME results padding in TensorFlow style
-             *                        - PADDING_SAME_MXNET results padding in MXNET style
+             *                        - PADDING_SAME_END results padding in TensorFlow style
+             *                        - PADDING_SAME_BEGIN results padding in MXNET style
+             *                        - PADDING_NOT_SET means padding with the specific "padding" value below. 
+             * @param padding         if padding_type is PADDING_NOT_SET, this value will be used as padding size. 
+             *                        the shape must be 4, the value of each position is: [padding top, padding bottom, padding left, padding right]
              * @param stride_y        stride in height
              * @param stride_x        stride in width
              * @param name            name of layer
              */
             Conv2D(const int output_exponent,
                    const Filter<feature_t> *filter,
-                   const Bias<feature_t> *bias = NULL,
+                   const Bias<bias_t> *bias = NULL,
                    const Activation<feature_t> *activation = NULL,
                    const padding_type_t padding_type = PADDING_VALID,
+                   std::vector<int> padding = {},
                    const int stride_y = 1,
                    const int stride_x = 1,
-                   const char *name = NULL) : Layer(name),
-                                              output_exponent(output_exponent),
-                                              filter(filter),
-                                              stride_y(stride_y),
-                                              stride_x(stride_x),
-                                              padding_type(padding_type),
-                                              bias(bias),
-                                              activation(activation)
+                   const char *name = "Conv2D") : Layer(name),
+                                                  output_exponent(output_exponent),
+                                                  filter(filter),
+                                                  stride_y(stride_y),
+                                                  stride_x(stride_x),
+                                                  padding_type(padding_type),
+                                                  bias(bias),
+                                                  activation(activation),
+                                                  padding(padding),
+                                                  output_shape({})
             {
                 this->output = new Tensor<feature_t>;
+                if (this->padding_type == PADDING_NOT_SET)
+                {
+                    assert(this->padding.size() == 4);
+                }
             }
 
             /**
@@ -82,19 +95,30 @@ namespace dl
              * @brief Update output padding and input padding.
              * 
              * @param input as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
                 assert(input.shape[0] > 0);
                 assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
+                assert(this->filter->shape.size() == 4);
+                assert(input.shape[2] == this->filter->shape[2]);
 
-                std::vector<int> output_shape = nn::get_output_shape(input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type, true);
-                this->output->set_shape(output_shape);
+                this->output_shape = nn::get_output_shape(input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type, true, this->padding);
+                this->output->set_shape(this->output_shape);
                 this->output->set_exponent(this->output_exponent);
                 this->output->free_element();
+                if (this->padding_type != PADDING_NOT_SET)
+                {
+                    this->padding = nn::get_pad_size(this->output_shape, input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type);
+                }
 
-                this->padding = nn::get_pad_size(output_shape, input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type);
-                input.set_padding_size(this->padding);
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -122,7 +146,11 @@ namespace dl
                 DL_LOG_LAYER_LATENCY_INIT();
 
                 DL_LOG_LAYER_LATENCY_START();
-                this->output->apply_element();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
                 this->output->set_exponent(this->output_exponent);
                 DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -153,5 +181,6 @@ namespace dl
                 dl::tool::cache::preload_func((uint32_t)(this->filter->element), size);
             }
         };
+
     } // namespace layer
 } // namespace dl
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_depthwise_conv2d.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_depthwise_conv2d.hpp
index 37475353209..30b2c2a6c77 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_depthwise_conv2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_depthwise_conv2d.hpp
@@ -13,8 +13,11 @@ namespace dl
          * @tparam feature_t supports int16_t and int8_t,
          *         - int16_t: stands for operation in int16_t quantize
          *         - int8_t: stands for operation in int8_t quantize
+         * @tparam bias_t supports int16_t and int8_t, must specify when using int8 per-channel quantization
+         *         - int16_t: for int16 quantization and int8 per-channel quantization
+         *         - int8_t: for int8 per-tensor quantization
          */
-        template <typename feature_t>
+        template <typename feature_t, typename bias_t = feature_t>
         class DepthwiseConv2D : public Layer
         {
         private:
@@ -22,14 +25,14 @@ namespace dl
             const Filter<feature_t> *filter;         /*<! filter of DepthwiseConv2D >*/
             const int stride_y;                      /*<! stride in height >*/
             const int stride_x;                      /*<! stride in width >*/
-            const padding_type_t padding_type;       /*<! one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET >*/
-            const Bias<feature_t> *bias;             /*<! bias of DepthwiseConv2D, if you don't specify anything, no bias is added >*/
+            const padding_type_t padding_type;       /*<! one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN >*/
+            const Bias<bias_t> *bias;                /*<! bias of DepthwiseConv2D, if you don't specify anything, no bias is added >*/
             const Activation<feature_t> *activation; /*<! activation of DepthwiseConv2D, if you don't specify anything, no activation is applied >*/
             std::vector<int> padding;                /*<! padding size needed in [top, bottom, left, right] of this operation >*/
             Tensor<feature_t> *output;               /*<! output ptr of DepthwiseConv2D >*/
+            std::vector<int> output_shape;           /*<! output shape of DepthwiseConv2D >*/
 
         public:
-
             /**
              * @brief Construct a new DepthwiseConv2D object.
              * 
@@ -37,40 +40,50 @@ namespace dl
              * @param filter          filter of DepthwiseConv2D
              * @param bias            bias of DepthwiseConv2D, if you don't specify anything, no bias is added
              * @param activation      activation of DepthwiseConv2D, if you don't specify anything, no activation is applied
-             * @param padding_type    one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+             * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN or PADDING_NOT_SET,
              *                        - PADDING_VALID means no padding
-             *                        PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
-             *                        such that output has the same height/width dimension as the input
-             *                        - PADDING_SAME results padding in TensorFlow style
-             *                        - PADDING_SAME_MXNET results padding in MXNET style
+             *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
+             *                        such that output has the same height/width dimension as the input,
+             *                        - PADDING_SAME_END results padding in TensorFlow style
+             *                        - PADDING_SAME_BEGIN results padding in MXNET style
+             *                        - PADDING_NOT_SET means padding with the specific "padding" value below. 
+             * @param padding         if padding_type is PADDING_NOT_SET, this value will be used as padding size. 
+             *                        the shape must be 4, the value of each position is: [padding top, padding bottom, padding left, padding right]
              * @param stride_y        - stride in height
              * @param stride_x        - stride in width
              * @param name            name of layer
              */
             DepthwiseConv2D(const int output_exponent,
                             const Filter<feature_t> *filter,
-                            const Bias<feature_t> *bias = NULL,
+                            const Bias<bias_t> *bias = NULL,
                             const Activation<feature_t> *activation = NULL,
                             const padding_type_t padding_type = PADDING_VALID,
+                            std::vector<int> padding = {},
                             const int stride_y = 1,
                             const int stride_x = 1,
-                            const char *name = NULL) : Layer(name),
-                                                       output_exponent(output_exponent),
-                                                       filter(filter),
-                                                       stride_y(stride_y),
-                                                       stride_x(stride_x),
-                                                       padding_type(padding_type),
-                                                       bias(bias),
-                                                       activation(activation)
+                            const char *name = "DepthwiseConv2D") : Layer(name),
+                                                                    output_exponent(output_exponent),
+                                                                    filter(filter),
+                                                                    stride_y(stride_y),
+                                                                    stride_x(stride_x),
+                                                                    padding_type(padding_type),
+                                                                    bias(bias),
+                                                                    activation(activation),
+                                                                    padding(padding),
+                                                                    output_shape({})
             {
                 this->output = new Tensor<feature_t>;
+                if (this->padding_type == PADDING_NOT_SET)
+                {
+                    assert(this->padding.size() == 4);
+                }
             }
 
             /**
              * @brief Destroy the DepthwiseConv2D object.
              * 
              */
-            ~DepthwiseConv2D() 
+            ~DepthwiseConv2D()
             {
                 if (this->output != NULL)
                 {
@@ -82,19 +95,31 @@ namespace dl
              * @brief Update output shape and padding.
              * 
              * @param input as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
                 assert(input.shape[0] > 0);
                 assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
+                assert(this->filter->shape.size() == 4);
+                assert(input.shape[2] == this->filter->shape[2]);
 
-                std::vector<int> output_shape = nn::get_output_shape(input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type);
-                this->output->set_shape(output_shape);
+                this->output_shape = nn::get_output_shape(input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type, false, this->padding);
+                this->output->set_shape(this->output_shape);
                 this->output->set_exponent(this->output_exponent);
 
-                this->padding = nn::get_pad_size(output_shape, input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type);
-                input.set_padding_size(this->padding);
+                if (this->padding_type != PADDING_NOT_SET)
+                {
+                    this->padding = nn::get_pad_size(this->output_shape, input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type);
+                }
                 this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -122,7 +147,12 @@ namespace dl
                 DL_LOG_LAYER_LATENCY_INIT();
 
                 DL_LOG_LAYER_LATENCY_START();
-                this->output->apply_element();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+
+                this->output->malloc_element();
                 this->output->set_exponent(this->output_exponent);
                 DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_expand_dims.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_expand_dims.hpp
new file mode 100644
index 00000000000..a59bed183fb
--- /dev/null
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_expand_dims.hpp
@@ -0,0 +1,128 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief 
+         * 
+         * @tparam feature_t 
+         */
+        template <typename feature_t>
+        class ExpandDims : public Layer
+        {
+        private:
+            std::vector<int> output_shape; /*<! output shape of ExpandDims >*/
+            std::vector<int> axis;         /*<! position where the new axis is placed. >*/
+            Tensor<feature_t> *output;     /*<! output ptr of ExpandDims >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+
+        public:
+            int output_exponent;
+
+            /**
+             * @brief Construct a new ExpandDims object
+             * 
+             * @param axis position where the new axis is placed.
+             * @param name name of layer
+             * @param inplace true: the output will store to input
+             *                false: the output will store to a separate memory
+             */
+            ExpandDims(std::vector<int> axis, const char *name = "ExpandDims", bool inplace = false) : Layer(name),
+                                                                                                       axis(axis), inplace(inplace), output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the ExpandDims object
+             * 
+             */
+            ~ExpandDims()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief  Update output shape.
+            * 
+            * @param input as an input.
+            * @param print_shape  whether to print the output shape.
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->expand_dims(this->axis);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                    this->output->expand_dims(this->axis);
+                }
+                this->output_shape = this->output->shape;
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& ExpandDims result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief call ExpandDims opeartion
+             * 
+             * @param input 
+             * @return Tensor<feature_t>& ExpandDims result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->copy_element(input, true);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "ExpandDims");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_shape(this->output_shape);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "ExpandDims");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_flatten.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_flatten.hpp
new file mode 100644
index 00000000000..70ae483a07f
--- /dev/null
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_flatten.hpp
@@ -0,0 +1,120 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief 
+         * 
+         * @tparam feature_t 
+         */
+        template <typename feature_t>
+        class Flatten : public Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Flatten >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of Flatten >*/
+
+        public:
+            /**
+             * @brief Construct a new Flatten object
+             * 
+             * @param name name of layer
+             * @param inplace true: the output will store to input0
+             *                false: the output will store to a separate memory
+             */
+            Flatten(const char *name = "Flatten", bool inplace = false) : Layer(name), inplace(inplace), output_shape({})
+            {}
+
+            /**
+             * @brief Destroy the Flatten object
+             * 
+             */
+            ~Flatten()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief  Update output shape.
+            * 
+            * @param input as an input
+            * @param print_shape  whether to print the output shape. 
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                this->output_shape = {input.get_size()};
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                }
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Flatten result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Flatten operation.
+             * 
+             * @param input as an input
+             * @return Tensor<feature_t>& Flatten result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->flatten();
+                    this->output->copy_element(input, true);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "flatten");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->flatten();
+                    DL_LOG_LAYER_LATENCY_END(this->name, "flatten");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_fullyconnected.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_fullyconnected.hpp
new file mode 100644
index 00000000000..afa7e5befba
--- /dev/null
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_fullyconnected.hpp
@@ -0,0 +1,167 @@
+#pragma once
+
+#include "dl_nn_fully_connected.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Activation(FullyConnected(input, filter) + bias).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @tparam bias_t supports int16_t and int8_t, must specify when using int8 per-channel quantization
+         *         - int16_t: for int16 quantization and int8 per-channel quantization
+         *         - int8_t: for int8 per-tensor quantization
+         */
+        template <typename feature_t, typename bias_t = feature_t>
+        class FullyConnected : public Layer
+        {
+        private:
+            const int output_exponent;               /*<! exponent of output >*/
+            const bool flatten;                      /*<! true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+                                                          false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim] >*/
+            const Filter<feature_t> *filter;         /*<! filter of FullyConnected >*/
+            const Bias<bias_t> *bias;                /*<! bias of FullyConnected, if you don't specify anything, no bias is added >*/
+            const Activation<feature_t> *activation; /*<! activation of FullyConnected, if you don't specify anything, no activation is applied >*/
+            Tensor<feature_t> *output;               /*<! output ptr of FullyConnected >*/
+            std::vector<int> output_shape;           /*<! output shape of FullyConnected >*/
+
+        public:
+            /**
+             * @brief Construct a new FullyConnected object.
+             * 
+             * @param output_exponent exponent of output
+             * @param filter          filter of FullyConnected
+             * @param bias            bias of FullyConnected, if you don't specify anything, no bias is added
+             * @param activation      activation of FullyConnected, if you don't specify anything, no activation is applied
+             * @param flatten         true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+                                      false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+             * @param name            name of layer
+             */
+            FullyConnected(const int output_exponent,
+                           const Filter<feature_t> *filter,
+                           const Bias<bias_t> *bias = NULL,
+                           const Activation<feature_t> *activation = NULL,
+                           const bool flatten = true,
+                           const char *name = "FullyConnected") : Layer(name),
+                                                                  output_exponent(output_exponent),
+                                                                  flatten(flatten),
+                                                                  filter(filter),
+                                                                  bias(bias),
+                                                                  activation(activation),
+                                                                  output_shape({})
+            {
+                this->output = new Tensor<feature_t>;
+            }
+
+            /**
+             * @brief Destroy the FullyConnected object.
+             * 
+             */
+            ~FullyConnected()
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output padding and input padding.
+             * 
+             * @param input as an input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                assert(this->filter->shape.size() == 4);
+                assert(this->filter->shape[0] == 1);
+                assert(this->filter->shape[1] == 1);
+                if (this->flatten)
+                {
+                    assert(input.get_size() == this->filter->shape[2]);
+                    this->output_shape = {this->filter->shape[3]};
+                }
+                else
+                {
+                    assert(input.shape.back() == this->filter->shape[2]);
+                    this->output_shape = input.shape;
+                    this->output_shape[this->output_shape.size() - 1] = this->filter->shape[3];
+                }
+                this->output->set_shape(this->output_shape);
+                this->output->set_exponent(this->output_exponent);
+                this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& FullyConnected result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call FullyConnected operation
+             * 
+             * @param input           as an input.
+             * @param autoload_enable one of true or false, 
+             *                        - true: load input and output from PSRAM to CACHE automatically
+             *                        - false: do not
+             * @param assign_core     not effective yet
+             * @return FullyConnected result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input, bool autoload_enable = false, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
+                this->output->set_exponent(this->output_exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                if (autoload_enable)
+                {
+                    dl::tool::cache::autoload_func((uint32_t)(this->output->element), this->output->get_size() * sizeof(feature_t),
+                                                   (uint32_t)(input.element), input.get_size() * sizeof(feature_t));
+                }
+
+                DL_LOG_LAYER_LATENCY_START();
+                nn::fully_connected(*this->output, input, *(this->filter), this->bias, this->activation, this->flatten, assign_core);
+                DL_LOG_LAYER_LATENCY_END(this->name, "fully_connected");
+                return *this->output;
+            }
+
+            /**
+             * @brief Preload the filter to Cache.
+             * NOTE: Call this layer's preload() before previous layer's call() such that filter could be loaded while previous layer is doing calculation.
+             */
+            void preload()
+            {
+                size_t size = sizeof(feature_t);
+                int shape_size = this->filter->shape.size();
+                for (int i = 0; i < shape_size; ++i)
+                {
+                    size *= filter->shape[i];
+                }
+                dl::tool::cache::preload_func((uint32_t)(this->filter->element), size);
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_global_avg_pool2d.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_global_avg_pool2d.hpp
index 044c5f61847..93f2d30ac89 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_global_avg_pool2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_global_avg_pool2d.hpp
@@ -20,8 +20,9 @@ namespace dl
         class GlobalAveragePool2D : public Layer
         {
         private:
-            const int output_exponent; /*<! exponent of output >*/
-            Tensor<feature_t> *output; /*<! output ptr of GlobalAveragePool2D >*/
+            const int output_exponent;     /*<! exponent of output >*/
+            std::vector<int> output_shape; /*<! output shape of GlobalAveragePool2D >*/
+            Tensor<feature_t> *output;     /*<! output ptr of GlobalAveragePool2D >*/
         public:
             /**
              * @brief Construct a new GlobalAveragePool2D object.
@@ -29,8 +30,9 @@ namespace dl
              * @param output_exponent exponent of output
              * @param name            name of layer
              */
-            GlobalAveragePool2D(const int output_exponent, const char *name = NULL) : Layer(name),
-                                                                                      output_exponent(output_exponent)
+            GlobalAveragePool2D(const int output_exponent, const char *name = "GlobalAveragePool2D") : Layer(name),
+                                                                                                       output_exponent(output_exponent),
+                                                                                                       output_shape({})
 
             {
                 this->output = new Tensor<feature_t>;
@@ -52,17 +54,26 @@ namespace dl
              * @brief Update output shape.
              * 
              * @param input as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
                 assert(input.shape[0] > 0);
                 assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
 
                 std::vector<int> output_shape(input.shape.size(), 1);
                 output_shape[2] = input.shape[2];
-                this->output->set_shape(output_shape);
+                this->output_shape = output_shape;
+                this->output->set_shape(this->output_shape);
                 this->output->set_exponent(this->output_exponent);
                 this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -90,7 +101,11 @@ namespace dl
                 DL_LOG_LAYER_LATENCY_INIT();
 
                 DL_LOG_LAYER_LATENCY_START();
-                this->output->apply_element();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
                 this->output->set_exponent(this->output_exponent);
                 DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_global_max_pool2d.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_global_max_pool2d.hpp
index aa146823b78..f9b7f73ff8c 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_global_max_pool2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_global_max_pool2d.hpp
@@ -20,15 +20,15 @@ namespace dl
         class GlobalMaxPool2D : public Layer
         {
         private:
-            Tensor<feature_t> *output;  /*<! output ptr of GlobalMaxPool2D >*/
+            Tensor<feature_t> *output;     /*<! output ptr of GlobalMaxPool2D >*/
+            std::vector<int> output_shape; /*<! output shape of GlobalMaxPool2D >*/
         public:
-
             /**
              * @brief Construct a new GlobalMaxPool2D object.
              * 
              * @param name         name of layer
              */
-            GlobalMaxPool2D(const char *name = NULL) : Layer(name)
+            GlobalMaxPool2D(const char *name = "GlobalMaxPool2D") : Layer(name), output_shape({})
             {
                 this->output = new Tensor<feature_t>;
             }
@@ -49,17 +49,26 @@ namespace dl
              * @brief Update output shape and exponent.
              * 
              * @param input as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
                 assert(input.shape[0] > 0);
                 assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
                 this->output->set_exponent(input.exponent);
 
                 std::vector<int> output_shape(input.shape.size(), 1);
                 output_shape[2] = input.shape[2];
-                this->output->set_shape(output_shape);
+                this->output_shape = output_shape;
+                this->output->set_shape(this->output_shape);
                 this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -87,7 +96,11 @@ namespace dl
                 DL_LOG_LAYER_LATENCY_INIT();
 
                 DL_LOG_LAYER_LATENCY_START();
-                this->output->apply_element();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
                 this->output->set_exponent(input.exponent);
                 DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_leakyrelu.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_leakyrelu.hpp
index f18d9b1c522..a972e135006 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_leakyrelu.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_leakyrelu.hpp
@@ -2,7 +2,7 @@
 
 #include "dl_constant.hpp"
 #include "dl_variable.hpp"
-#include "dl_nn_LeakyReLU.hpp"
+#include "dl_nn_leakyrelu.hpp"
 #include "dl_layer_base.hpp"
 
 namespace dl
@@ -20,13 +20,13 @@ namespace dl
         class LeakyReLU : public Layer
         {
         private:
-            feature_t activation_alpha; /*<! quantized alpha >*/
-            int activation_exponent;    /*<! exponent of quantized alpha >*/
-            Tensor<feature_t> *output;  /*<! output ptr of leakyrelu>*/
-            bool inplace;               /*<! true: the output will store to input0
-                                             false: the output will store to a seperate memeory >*/
+            feature_t activation_alpha;    /*<! quantized alpha >*/
+            int activation_exponent;       /*<! exponent of quantized alpha >*/
+            Tensor<feature_t> *output;     /*<! output ptr of leakyrelu>*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of leakyrelu >*/
         public:
-
             /**
              * @brief Construct a new LeakyReLU object
              * 
@@ -34,9 +34,9 @@ namespace dl
              * @param activation_exponent  exponent of quantized alpha
              * @param name                 name of leakyrelu
              * @param inplace              true: the output will store to input0
-             *                             false: the output will store to a seperate memeory
+             *                             false: the output will store to a separate memory
              */
-            LeakyReLU(const int activation_alpha, const int activation_exponent, const char *name = NULL, bool inplace = false) : Layer(name), output(NULL)
+            LeakyReLU(const int activation_alpha, const int activation_exponent, const char *name = "LeakyReLU", bool inplace = false) : Layer(name), output(NULL), output_shape({})
             {
                 this->activation_alpha = activation_alpha;
                 this->activation_exponent = activation_exponent;
@@ -47,7 +47,7 @@ namespace dl
              * @brief Destroy the LeakyReLU object
              * 
              */
-            ~LeakyReLU() 
+            ~LeakyReLU()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -59,24 +59,32 @@ namespace dl
              * @brief Update output shape and exponent
              * 
              * @param input       as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
-                if(!this->inplace)
+                this->output_shape = input.shape;
+                if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
-                    }  
-                    this->output->set_shape(input.shape);
+                    }
+                    this->output->set_shape(this->output_shape);
                     this->output->set_exponent(input.exponent);
                     this->output->free_element();
                 }
                 else
                 {
                     this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
                 }
-                
             }
 
             /**
@@ -100,10 +108,14 @@ namespace dl
             {
                 DL_LOG_LAYER_LATENCY_INIT();
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output->apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
                     this->output->set_exponent(input.exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -114,6 +126,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::leakyrelu<true>(*this->output, input, this->activation_alpha, this->activation_exponent, assign_core);
                     DL_LOG_LAYER_LATENCY_END(this->name, "leakyrelu");
                 }
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_max2d.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_max2d.hpp
index 8a775a2e0f0..93bc5899443 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_max2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_max2d.hpp
@@ -22,28 +22,28 @@ namespace dl
         class Max2D : public Layer
         {
         private:
-            Tensor<feature_t> *output;  /*<! output ptr of max2d >*/
-            bool inplace;               /*<! true: the output will store to input0
-                                             false: the output will store to a seperate memeory >*/
+            Tensor<feature_t> *output;     /*<! output ptr of max2d >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of max2d >*/
         public:
-            
             /**
              * @brief Construct a new Max2D object.
              * 
              * @param name            name of max2d
              * @param inplace         true: the output will store to input0
-             *                        false: the output will store to a seperate memeory
+             *                        false: the output will store to a separate memory
              */
-            Max2D(const char *name = NULL, bool inplace = false) : Layer(name), output(NULL)
+            Max2D(const char *name = "Max2D", bool inplace = false) : Layer(name),
+                                                                      output(NULL), inplace(inplace), output_shape({})
             {
-                this->inplace = inplace;
             }
 
             /**
              * @brief Destroy the Max2D object
              * 
              */
-            ~Max2D() 
+            ~Max2D()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -58,24 +58,34 @@ namespace dl
              * 
              * @param input0 as one input
              * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1)
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
             {
                 assert(input0.is_same_shape(input1));
                 assert(input0.exponent == input1.exponent);
+                this->output_shape = input0.shape;
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
                     }
                     this->output->set_exponent(this->output_exponent);
-                    this->output->set_shape(input0.shape);
+                    this->output->set_shape(this->output_shape);
                     this->output->free_element();
                 }
                 else
+                {
                     this->output = &input0;
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -100,10 +110,14 @@ namespace dl
             {
                 DL_LOG_LAYER_LATENCY_INIT();
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output->apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
                     this->output->set_exponent(input0.exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -114,6 +128,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::max2d<true>(*this->output, input0, input1, assign_core);
                     DL_LOG_LAYER_LATENCY_END(this->name, "max2d");
                 }
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_max_pool2d.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_max_pool2d.hpp
index f836a983b34..629aa87f515 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_max_pool2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_max_pool2d.hpp
@@ -23,44 +23,54 @@ namespace dl
             std::vector<int> filter_shape;     /*<! filter shape in [filter_height, filter_width] >*/
             const int stride_y;                /*<! stride in height >*/
             const int stride_x;                /*<! stride in width >*/
-            const padding_type_t padding_type; /*<! one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET >*/
+            const padding_type_t padding_type; /*<! one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN >*/
             std::vector<int> padding;          /*<! padding size needed in [top, bottom, left, right] of this operation >*/
             Tensor<feature_t> *output;         /*<! output ptr of MaxPool2D >*/
+            std::vector<int> output_shape;     /*<! output shape of MaxPool2D >*/
 
         public:
-
             /**
              * @brief Construct a new MaxPool2D object.
              * 
              * @param filter_shape filter shape in [filter_height, filter_width]
-             * @param padding_type one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+             * @param padding_type one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN or PADDING_NOT_SET,
              *                     - PADDING_VALID means no padding
-             *                     PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+             *                     PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
              *                     such that output has the same height/width dimension as the input,
-             *                     - PADDING_SAME results padding in TensorFlow style
-             *                     - PADDING_SAME_MXNET results padding in MXNET style
+             *                     - PADDING_SAME_END results padding in TensorFlow style
+             *                     - PADDING_SAME_BEGIN results padding in MXNET style
+             *                     - PADDING_NOT_SET means padding with the specific "padding" value below. 
+             * @param padding      if padding_type is PADDING_NOT_SET, this value will be used as padding size. 
+             *                     the shape must be 4, the value of each position is: [padding top, padding bottom, padding left, padding right]
              * @param stride_y     stride in height
              * @param stride_x     stride in width
              * @param name         name of layer
              */
             MaxPool2D(const std::vector<int> filter_shape,
                       const padding_type_t padding_type = PADDING_VALID,
+                      std::vector<int> padding = {},
                       const int stride_y = 1,
                       const int stride_x = 1,
-                      const char *name = NULL) : Layer(name),
-                                                 filter_shape(filter_shape),
-                                                 stride_y(stride_y),
-                                                 stride_x(stride_x),
-                                                 padding_type(padding_type)
+                      const char *name = "MaxPool2D") : Layer(name),
+                                                        filter_shape(filter_shape),
+                                                        padding_type(padding_type),
+                                                        padding(padding),
+                                                        stride_y(stride_y),
+                                                        stride_x(stride_x),
+                                                        output_shape({})
             {
                 this->output = new Tensor<feature_t>;
+                if (this->padding_type == PADDING_NOT_SET)
+                {
+                    assert(this->padding.size() == 4);
+                }
             }
 
             /**
              * @brief Destroy the MaxPool2D object.
              * 
              */
-            ~MaxPool2D() 
+            ~MaxPool2D()
             {
                 if (this->output != NULL)
                 {
@@ -72,18 +82,29 @@ namespace dl
              * @brief Update output shape and padding.
              * 
              * @param input as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
                 assert(input.shape[0] > 0);
                 assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
+
                 this->output->set_exponent(input.exponent);
-                std::vector<int> output_shape = nn::get_output_shape(input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
-                this->output->set_shape(output_shape);
+                this->output_shape = nn::get_output_shape(input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type, false, this->padding);
+                this->output->set_shape(this->output_shape);
 
-                this->padding = nn::get_pad_size(output_shape, input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
-                input.set_padding_size(this->padding);
+                if (this->padding_type != PADDING_NOT_SET)
+                {
+                    this->padding = nn::get_pad_size(this->output_shape, input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
+                }
                 this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -111,7 +132,11 @@ namespace dl
                 DL_LOG_LAYER_LATENCY_INIT();
 
                 DL_LOG_LAYER_LATENCY_START();
-                this->output->apply_element();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
                 this->output->set_exponent(input.exponent);
                 DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_min2d.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_min2d.hpp
index 71d39c9b285..e38fbf3d0d2 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_min2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_min2d.hpp
@@ -22,28 +22,28 @@ namespace dl
         class Min2D : public Layer
         {
         private:
-            Tensor<feature_t> *output;  /*<! output of ptr min2d>*/ 
-            bool inplace;               /*<! true: the output will store to input0
-                                             false: the output will store to a seperate memeory >*/ 
-        public:    
-
+            Tensor<feature_t> *output;     /*<! output of ptr min2d>*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of min2d >*/
+        public:
             /**
              * @brief Construct a new Min2D object
              * 
              * @param name            name of min2d
              * @param inplace         true: the output will store to input0
-             *                        false: the output will store to a seperate memeory
+             *                        false: the output will store to a separate memory
              */
-            Min2D(const char *name = NULL, bool inplace = false) : Layer(name), output(NULL)
-            {
-                this->inplace = inplace;
-            }
+            Min2D(const char *name = "Min2D", bool inplace = false) : Layer(name),
+                                                                      output(NULL),
+                                                                      inplace(inplace),
+                                                                      output_shape({}) {}
 
             /**
              * @brief Destroy the Min2D object
              * 
              */
-            ~Min2D() 
+            ~Min2D()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -58,25 +58,34 @@ namespace dl
              * 
              * @param input0 as one input
              * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1)
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
             {
                 assert(input0.is_same_shape(input1));
                 assert(input0.exponent == input1.exponent);
+                this->output_shape = input0.shape;
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
                     }
-                    this->output->set_shape(input0.shape);
+                    this->output->set_shape(this->output_shape);
                     this->output->set_exponent(input0.exponent);
                     this->output->free_element();
                 }
                 else
+                {
                     this->output = &input0;
-                
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -101,10 +110,14 @@ namespace dl
             {
                 DL_LOG_LAYER_LATENCY_INIT();
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output->apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
                     this->output->set_exponent(input0.exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -115,6 +128,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::min2d<true>(*this->output, input0, input1, assign_core);
                     DL_LOG_LAYER_LATENCY_END(this->name, "min2d");
                 }
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_mul2d.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_mul2d.hpp
index 0edfc9a93f0..21bcca7a81e 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_mul2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_mul2d.hpp
@@ -21,14 +21,13 @@ namespace dl
         class Mul2D : public Layer
         {
         private:
-            const int output_exponent;              /*<! exponent of output >*/
+            const int output_exponent;               /*<! exponent of output >*/
             const Activation<feature_t> *activation; /*<! activation of Mul2D, if you don't specify anything, no activation is applied >*/
-            Tensor<feature_t> *output;  /*<! output ptr of Mul2D >*/
-            bool inplace;               /*<! true: the output will store to input0
-                                             false: the output will store to a seperate memeory >*/ 
+            Tensor<feature_t> *output;               /*<! output ptr of Mul2D >*/
+            bool inplace;                            /*<! true: the output will store to input0
+                                                            false: the output will store to a separate memory >*/
+            std::vector<int> output_shape;           /*<! output shape of Mul2D >*/
         public:
-            const int output_exponent; /*<! exponent of output >*/
-
             /**
              * @brief Construct a new Mul2D object.
              * 
@@ -36,18 +35,24 @@ namespace dl
              * @param activation      activation of Mul2D, if you don't specify anything, no activation is applied
              * @param name            name of layer
              * @param inplace         true: the output will store to input0
-             *                        false: the output will store to a seperate memeory
+             *                        false: the output will store to a separate memory
              */
-            Mul2D(const int output_exponent, const Activation<feature_t> *activation = NULL, const char *name = NULL, bool inplace = false) : Layer(name), 
-                                                                                            output_exponent(output_exponent),activation(activation), output(NULL)
+            Mul2D(const int output_exponent,
+                  const Activation<feature_t> *activation = NULL,
+                  const char *name = "Mul2D",
+                  bool inplace = false) : Layer(name),
+                                          output_exponent(output_exponent),
+                                          activation(activation),
+                                          output(NULL),
+                                          inplace(inplace),
+                                          output_shape({})
             {
-                this->inplace = inplace;
             }
 
             /**
              * @brief Destroy the Multiply2D object.
              */
-            ~Mul2D() 
+            ~Mul2D()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -61,24 +66,34 @@ namespace dl
              * 
              * @param input0 as one input
              * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1)
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
             {
                 assert(input0.is_same_shape(input1));
+                this->output_shape = input0.shape;
 
                 if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
                     }
                     this->output->set_exponent(this->output_exponent);
-                    this->output->set_shape(input0.shape);
+                    this->output->set_shape(this->output_shape);
                     this->output->free_element();
                 }
-                    
+
                 else
+                {
                     this->output = &input0;
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -106,7 +121,11 @@ namespace dl
                 if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output->apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
                     this->output->set_exponent(this->output_exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -117,6 +136,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::mul2d<true>(*this->output, input0, input1, this->activation, assign_core);
                     DL_LOG_LAYER_LATENCY_END(this->name, "mul2d");
                 }
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_prelu.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_prelu.hpp
index 4781a669130..96168a783b1 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_prelu.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_prelu.hpp
@@ -24,9 +24,9 @@ namespace dl
             int activation_exponent;       /*<! exponent of quantized alpha elements >*/
             Tensor<feature_t> *output;     /*<! output ptr of prelu >*/
             bool inplace;                  /*<! true: the output will store to input0
-                                                false: the output will store to a seperate memeory >*/ 
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of prelu >*/
         public:
-
             /**
              * @brief Construct a new PReLU object
              * 
@@ -34,20 +34,25 @@ namespace dl
              * @param activation_exponent  exponent of quantized alpha elements
              * @param name                 name of prelu
              * @param inplace              true: the output will store to input0
-             *                             false: the output will store to a seperate memeory
+             *                             false: the output will store to a separate memory
              */
-            PReLU(const feature_t *activation_element, const int activation_exponent = 0, const char *name = NULL, bool inplace = false) : Layer(name), output(NULL)
+            PReLU(const feature_t *activation_element,
+                  const int activation_exponent = 0,
+                  const char *name = NULL,
+                  bool inplace = "PReLU") : Layer(name),
+                                            activation_element(activation_element),
+                                            activation_exponent(activation_exponent),
+                                            output(NULL),
+                                            inplace(inplace),
+                                            output_shape({})
             {
-                this->activation_element = activation_element;
-                this->activation_exponent = activation_exponent;
-                this->inplace = inplace;
             }
 
             /**
              * @brief Destroy the PReLU object
              * 
              */
-            ~PReLU() 
+            ~PReLU()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -59,23 +64,31 @@ namespace dl
              * @brief Update output shape and exponent
              * 
              * @param input       as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
-                if(!this->inplace)
+                this->output_shape = input.shape;
+                if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
                     }
                     this->output->set_exponent(input.exponent);
-                    this->output->set_shape(input.shape);
+                    this->output->set_shape(this->output_shape);
                     this->output->free_element();
                 }
                 else
                 {
                     this->output = &input;
                 }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -99,11 +112,15 @@ namespace dl
             {
                 DL_LOG_LAYER_LATENCY_INIT();
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     this->output->set_exponent(input.exponent);
-                    this->output->apply_element();
+                    this->output->malloc_element();
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
                     DL_LOG_LAYER_LATENCY_START();
@@ -113,6 +130,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::prelu(*this->output, input, this->activation_element, this->activation_exponent, assign_core);
                     DL_LOG_LAYER_LATENCY_END(this->name, "leakyrelu");
                 }
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_relu.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_relu.hpp
index e70663b798c..7dd29d4a178 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_relu.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_relu.hpp
@@ -21,29 +21,28 @@ namespace dl
         class ReLU : public Layer
         {
         private:
-            Tensor<feature_t> *output;  /*<! output ptr of relu >*/
-            bool inplace;               /*<! true: the output will store to input0
-                                             false: the output will store to a seperate memeory >*/ 
+            Tensor<feature_t> *output;     /*<! output ptr of relu >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of relu >*/
         public:
-
-
             /**
              * @brief Construct a new ReLU object
              * 
              * @param name            name of relu
              * @param inplace         true: the output will store to input0
-             *                        false: the output will store to a seperate memeory
+             *                        false: the output will store to a separate memory
              */
-            ReLU(const char *name = NULL, bool inplace = false) : Layer(name), output(NULL)
+            ReLU(const char *name = "ReLU", bool inplace = false) : Layer(name),
+                                                                    output(NULL), inplace(inplace), output_shape({})
             {
-                this->inplace = inplace;
             }
 
             /**
              * @brief Destroy the ReLU object
              * 
              */
-            ~ReLU() 
+            ~ReLU()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -55,23 +54,31 @@ namespace dl
              * @brief Update output shape and exponent
              * 
              * @param input       as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
-                if(!this->inplace)
+                this->output_shape = input.shape;
+                if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
                     }
                     this->output->set_exponent(input.exponent);
-                    this->output->set_shape(input.shape);
+                    this->output->set_shape(this->output_shape);
                     this->output->free_element();
                 }
                 else
                 {
                     this->output = &input;
                 }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -95,10 +102,14 @@ namespace dl
             {
                 DL_LOG_LAYER_LATENCY_INIT();
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output->apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
                     this->output->set_exponent(input.exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -109,6 +120,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::relu(*this->output, input, assign_core);
                     DL_LOG_LAYER_LATENCY_END(this->name, "relu");
                 }
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_reshape.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_reshape.hpp
new file mode 100644
index 00000000000..3f2ed72b6e0
--- /dev/null
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_reshape.hpp
@@ -0,0 +1,124 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief  Reshape(input)
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class Reshape : public Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Reshape >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of Reshape >*/
+        public:
+            /**
+             * @brief Construct a new Reshape object
+             * 
+             * @param shape      the target shape
+             * @param name       name of Reshape layer
+             * @param inplace    true: the output will store to input0
+             *                   false: the output will store to a separate memory
+             */
+            Reshape(std::vector<int> shape, const char *name = "Reshape", bool inplace = false) : Layer(name),
+                                                                                                  output_shape(shape), inplace(inplace)
+            {
+            }
+
+            /**
+             * @brief Destroy the Reshape object
+             * 
+             */
+            ~Reshape()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief             Update output shape and exponent
+            * 
+            * @param input        as an input
+            * @param print_shape  whether to print the output shape.
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Reshape result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Reshape operation.
+             * 
+             * @param input  as an input
+             * @return Tensor<feature_t>& Reshape result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->reshape(this->output_shape);
+                    this->output->copy_element(input, true);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "reshape");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->reshape(this->output_shape);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "reshape");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_squeeze.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_squeeze.hpp
new file mode 100644
index 00000000000..cee92f22764
--- /dev/null
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_squeeze.hpp
@@ -0,0 +1,127 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief 
+         * 
+         * @tparam feature_t 
+         */
+        template <typename feature_t>
+        class Squeeze : public Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Squeeze >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                 false: the output will store to a separate memory >*/
+            int axis;                      /*<! the dim to to be remove. make sure the length of the dim is equal to 1.
+                                            if axis == INT32_MAX, all the dims with length==1 will be removed. >*/
+            std::vector<int> output_shape; /*<! output shape of AvgPool2D >*/
+        public:
+            /**
+             * @brief Construct a new Squeeze object
+             * 
+             * @param axis the dim to to be remove. make sure the length of the dim is equal to 1.
+             *             if axis == INT32_MAX, all the dims with length==1 will be removed.
+             * @param name      name of Squeeze layer
+             * @param inplace   true: the output will store to input0
+             *                  false: the output will store to a separate memory
+             */
+            Squeeze(int axis = INT32_MAX, const char *name = "Squeeze", bool inplace = false) : Layer(name), axis(axis), inplace(inplace), output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the Squeeze object
+             * 
+             */
+            ~Squeeze()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief Update output shape and exponent
+            * 
+            * @param input as an input
+            * @param print_shape  whether to print the output shape.
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(input.shape);
+                    this->output->squeeze(this->axis);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(input.shape);
+                    this->output->squeeze(this->axis);
+                }
+                this->output_shape = this->output->shape;
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Squeeze result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief  Call Squeeze operation.
+             * 
+             * @param input as an input
+             * @return Tensor<feature_t>& Squeeze result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->copy_element(input, true);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "Squeeze");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_shape(this->output_shape);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "Squeeze");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_sub2d.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_sub2d.hpp
index e3453c08b97..da03b4aad85 100644
--- a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_sub2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_sub2d.hpp
@@ -21,13 +21,13 @@ namespace dl
         class Sub2D : public Layer
         {
         private:
-            const int output_exponent; /*<! exponent of output >*/
-            const Activation<feature_t> *activation; /*<! activation of Mul2D, if you don't specify anything, no activation is applied >*/
-            Tensor<feature_t> *output;  /*<! output ptr of Sub2D >*/
-            bool inplace;               /*<! true: the output will store to input0
-                                             false: the output will store to a seperate memeory >*/ 
+            const int output_exponent;               /*<! exponent of output >*/
+            const Activation<feature_t> *activation; /*<! activation of Sub2D, if you don't specify anything, no activation is applied >*/
+            Tensor<feature_t> *output;               /*<! output ptr of Sub2D >*/
+            bool inplace;                            /*<! true: the output will store to input0
+                                                            false: the output will store to a separate memory >*/
+            std::vector<int> output_shape;           /*<! output shape of Sub2D >*/
         public:
-
             /**
              * @brief Construct a new Sub2D object.
              * 
@@ -35,18 +35,17 @@ namespace dl
              * @param activation      activation of Mul2D, if you don't specify anything, no activation is applied
              * @param name            name of layer
              * @param inplace         true: the output will store to input0
-             *                        false: the output will store to a seperate memeory
+             *                        false: the output will store to a separate memory
              */
-            Sub2D(const int output_exponent, const Activation<feature_t> *activation = NULL, const char *name = NULL, bool inplace = false) : Layer(name), 
-                                                        output_exponent(output_exponent), activation(activation), output(NULL)
+            Sub2D(const int output_exponent, const Activation<feature_t> *activation = NULL, const char *name = "Sub2D", bool inplace = false) : Layer(name),
+                                                                                                                                                 output_exponent(output_exponent), activation(activation), output(NULL), inplace(inplace), output_shape({})
             {
-                this->inplace = inplace;
             }
 
             /**
              * @brief Destroy the Sub2D object.
              */
-            ~Sub2D() 
+            ~Sub2D()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -60,22 +59,32 @@ namespace dl
              * 
              * @param input0 as one input
              * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1)
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
             {
                 assert(input0.is_same_shape(input1));
+                this->output_shape = input0.shape;
                 if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
                     }
                     this->output->set_exponent(this->output_exponent);
-                    this->output->set_shape(input0.shape);
+                    this->output->set_shape(this->output_shape);
                     this->output->free_element();
-                }    
+                }
                 else
+                {
                     this->output = &input0;
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -103,7 +112,11 @@ namespace dl
                 if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output.apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output.malloc_element();
                     this->output->set_exponent(input0.exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -114,6 +127,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::sub2d<true>(this->output, input0, input1, this->activation, assign_core, this->output_exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "sub2d");
                 }
diff --git a/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_transpose.hpp b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_transpose.hpp
new file mode 100644
index 00000000000..d89ba8daed5
--- /dev/null
+++ b/tools/sdk/esp32/include/esp-face/include/layer/dl_layer_transpose.hpp
@@ -0,0 +1,126 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief 
+         * 
+         * @tparam feature_t 
+         */
+        template <typename feature_t>
+        class Transpose : public Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Transpose >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> perm;         /*<! the new arangement of the dims. if perm == {}, the dims arangement will be reversed. >*/
+            std::vector<int> output_shape; /*<! output shape of Transpose >*/
+        public:
+            /**
+             * @brief Construct a new Transpose object
+             * 
+             * @param perm          the new arangement of the dims. if perm == {}, the dims arangement will be reversed.
+             * @param name          name of Transpose layer
+             * @param inplace       true: the output will store to input
+             *                      false: the output will store to a separate memory
+             */
+            Transpose(std::vector<int> perm = {}, const char *name = "Transpose", bool inplace = false) : Layer(name), perm(perm), inplace(inplace), output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the Transpose object
+             * 
+             */
+            ~Transpose()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief Update output shape and exponent
+            * 
+            * @param input as an input
+            * @param print_shape  whether to print the output shape.
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                this->output_shape = input.shape;
+                for (int i = 0; i < this->perm.size(); i++)
+                {
+                    this->output_shape[i] = input.shape[this->perm[i]];
+                }
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Transpose result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Transpose operation.
+             * 
+             * @param input as an input.
+             * @return Tensor<feature_t>& Transpose result.
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->transpose(input, this->perm);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "transpose");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->transpose(this->perm);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "transpose");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32/include/esp-face/include/model_zoo/color_detector.hpp b/tools/sdk/esp32/include/esp-face/include/model_zoo/color_detector.hpp
new file mode 100644
index 00000000000..063ab20b34a
--- /dev/null
+++ b/tools/sdk/esp32/include/esp-face/include/model_zoo/color_detector.hpp
@@ -0,0 +1,68 @@
+#pragma once
+
+#include "dl_image.hpp"
+
+typedef struct
+{
+    int area;                /*!< Area of connected domains >*/
+    std::vector<int> center; /*<! centroid of connected domains [x, y] >*/
+    std::vector<int> box;    /*<! [left_up_x, left_up_y, right_down_x, right_down_y] >*/
+} components_stats_t;
+
+class ColorDetector
+{
+private:
+    std::vector<std::vector<components_stats_t>> results; /*!< detection results >*/
+
+public:
+    std::vector<std::vector<uint8_t>> color_thresh; /*!< threshold of colors, The threshold of each color is composed of 6 numbers >*/
+    std::vector<int> area_thresh;                   /*!< the area threshold of each color, 
+                                                                the area that is smaller than the threshold is filtered >*/
+    bool bgr;                                       /*!< true: the input image is in BGR format
+                                                                false: the input image is in RGB format >*/
+
+    /**
+     * @brief get the color threshold of rectangular region in the image
+     * 
+     * @param image the input image
+     * @param box   the coordinates of the rectanglar region : [left_up_x, left_up_y, right_down_x, right_down_y]
+     * @return std::vector<uint8_t> the threshold.
+     */
+    std::vector<uint8_t> cal_color_thresh(dl::Tensor<uint8_t> &image, std::vector<int> box);
+
+    /**
+     * @brief detect the colors based on the color thresholds
+     * 
+     * @param image the input image.
+     * @return std::vector<std::vector<components_stats_t>>&  detection result.
+     */
+    std::vector<std::vector<components_stats_t>> &detect(dl::Tensor<uint8_t> &image);
+
+    /**
+     * @brief Construct a new Color Detector object
+     * 
+     * @param color_thresh  threshold of colors, The threshold of each color is composed of 6 numbers
+     * @param area_thresh   the area threshold of each color,the area that is smaller than the threshold is filtered
+     * @param bgr           true: the input image is in BGR format
+     *                      false: the input image is in RGB format
+     */
+    ColorDetector(std::vector<std::vector<uint8_t>> color_thresh, std::vector<int> area_thresh, bool bgr = false) : color_thresh(color_thresh), area_thresh(area_thresh), bgr(bgr)
+    {
+    }
+
+    /**
+     * @brief Destroy the Color Detector object
+     * 
+     */
+    ~ColorDetector() {}
+
+    /**
+     * @brief Get the results object
+     * 
+     * @return std::vector<std::vector<components_stats_t>>& the detection result.
+     */
+    std::vector<std::vector<components_stats_t>> &get_results()
+    {
+        return this->results;
+    }
+};
\ No newline at end of file
diff --git a/tools/sdk/esp32/include/esp-face/include/model_zoo/face_recognition_tool.hpp b/tools/sdk/esp32/include/esp-face/include/model_zoo/face_recognition_tool.hpp
index 12fe8a842a2..2226d32daf9 100644
--- a/tools/sdk/esp32/include/esp-face/include/model_zoo/face_recognition_tool.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/model_zoo/face_recognition_tool.hpp
@@ -92,7 +92,7 @@ namespace face_recognition_tool
      * @return dl::Tensor<T>* 
      */
     template <typename T>
-    dl::Tensor<T> *transform_mfn_input(dl::Tensor<uint8_t> &image, bool free_input = false, bool do_padding = true);
+    dl::Tensor<T> *transform_mfn_input(dl::Tensor<uint8_t> &image, bool free_input = false);
 
     /**
      * @brief  transform the image to the input of a mfn model 
@@ -106,7 +106,7 @@ namespace face_recognition_tool
      *                             false: do not pad the result
      */
     template <typename T>
-    void transform_mfn_input(dl::Tensor<uint8_t> &image, dl::Tensor<T> &output, bool free_input = false, bool do_padding = true);
+    void transform_mfn_input(dl::Tensor<uint8_t> &image, dl::Tensor<T> &output, bool free_input = false);
 
     /**
      * @brief transform the mfn output embedding to a floating embedding
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn.hpp
index 6dba8016f26..6c737c1d341 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn.hpp
@@ -14,13 +14,13 @@ namespace dl
          * @param filter_shape filter shape with dilation
          * @param stride_y     stride in height
          * @param stride_x     stride in width
-         * @param pad_type     one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET
+         * @param pad_type     one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN
          * @param is_conv2d    one of true or false,
          *                     - true: serve for Conv2D
          *                     - false: serve for other operations
          * @return std::vector<int> 
          */
-        std::vector<int> get_output_shape(const std::vector<int> &input_shape, const std::vector<int> &filter_shape, const int stride_y, const int stride_x, const padding_type_t pad_type, const bool is_conv2d = false);
+        std::vector<int> get_output_shape(const std::vector<int> &input_shape, const std::vector<int> &filter_shape, const int stride_y, const int stride_x, const padding_type_t pad_type, const bool is_conv2d = false, std::vector<int> padding = {});
 
         /**
          * @brief Get the pad size object
@@ -30,7 +30,7 @@ namespace dl
          * @param filter_shape filter shape with dilation
          * @param stride_y     stride in height
          * @param stride_x     stride in width
-         * @param padding_type one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET
+         * @param padding_type one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN
          * @return padding size
          */
         std::vector<int> get_pad_size(const std::vector<int> &output_shape, const std::vector<int> &input_shape, const std::vector<int> &filter_shape, const int stride_y, const int stride_x, const padding_type_t padding_type);
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_add2d.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_add2d.hpp
index d296be5350b..4d4daaaef05 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_add2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_add2d.hpp
@@ -58,20 +58,20 @@ namespace dl
          */
         template <bool inplace = false, typename feature_t>
         auto add2d(const int output_exponent,
-                    Tensor<feature_t> &input0,
-                    Tensor<feature_t> &input1,
-                    const Activation<feature_t> *activation,
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+                   Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const Activation<feature_t> *activation,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             assert(input0.is_same_shape(input1));
 
             DL_LOG_NN_LATENCY_INIT();
 
             Tensor<feature_t> output;
-            if constexpr(!inplace)
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(output_exponent).set_shape(input0.shape).apply_element();
+                output.set_exponent(output_exponent).set_shape(input0.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_avg_pool2d.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_avg_pool2d.hpp
index 5b298d98c44..6e7db6ed0e0 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_avg_pool2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_avg_pool2d.hpp
@@ -58,12 +58,12 @@ namespace dl
          * @param filter_shape    filter_shape in [filter_height, filter_width]
          * @param stride_y        stride in height
          * @param stride_x        stride in width
-         * @param padding_type    one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+         * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN,
          *                        - PADDING_VALID: no padding
-         *                        PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+         *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
          *                        such that output has the same height/width dimension as the input,
-         *                        - PADDING_SAME results padding in TensorFlow style
-         *                        - PADDING_SAME_MXNET results padding in MXNET style
+         *                        - PADDING_SAME_END results padding in TensorFlow style
+         *                        - PADDING_SAME_BEGIN results padding in MXNET style
          * @param assign_core     not effective yet
          * @return avg_pool2d result
          */
@@ -81,19 +81,19 @@ namespace dl
             DL_LOG_NN_LATENCY_START();
             std::vector<int> output_shape = get_output_shape(input.shape, filter_shape, stride_y, stride_x, padding_type);
             Tensor<feature_t> output;
-            output.set_exponent(output_exponent).set_shape(output_shape).apply_element();
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
             DL_LOG_NN_LATENCY_END("apply");
+            std::vector<int> padding(4, 0);
 
             DL_LOG_NN_LATENCY_START();
-            if (padding_type == PADDING_SAME || padding_type == PADDING_SAME_MXNET)
+            if (padding_type == PADDING_SAME_END || padding_type == PADDING_SAME_BEGIN)
             {
-                std::vector<int> padding = get_pad_size(output_shape, input.shape, filter_shape, stride_y, stride_x, padding_type);
-                input.set_padding_size(padding);
+                padding = get_pad_size(output_shape, input.shape, filter_shape, stride_y, stride_x, padding_type);
             }
             DL_LOG_NN_LATENCY_END("padding");
 
             DL_LOG_NN_LATENCY_START();
-            avg_pool2d(output, input, input.padding, filter_shape, stride_y, stride_x, assign_core);
+            avg_pool2d(output, input, padding, filter_shape, stride_y, stride_x, assign_core);
             DL_LOG_NN_LATENCY_END("avg_pool2d");
 
             return output;
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_concat.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_concat.hpp
new file mode 100644
index 00000000000..73ed1aae905
--- /dev/null
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_concat.hpp
@@ -0,0 +1,63 @@
+#pragma once
+
+#include <vector>
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        template <typename feature_t>
+        void concat(Tensor<feature_t> &output, std::vector<Tensor<feature_t> *> &inputs, int axis, bool free_inputs = false);
+
+        template <typename feature_t>
+        Tensor<feature_t> concat(std::vector<Tensor<feature_t> *> &inputs, int axis, bool free_inputs = false)
+        {
+            DL_LOG_NN_LATENCY_INIT();
+
+            DL_LOG_NN_LATENCY_START();
+            assert(inputs.size() > 1);
+            int shape_size = inputs[0]->shape.size();
+
+            if (axis < 0)
+            {
+                axis = shape_size + axis;
+            }
+
+            assert((axis < shape_size) && (axis > -1));
+
+            int output_shape_axis = inputs[0]->shape[axis];
+
+            for (int i = 1; i < inputs.size(); i++)
+            {
+                assert(shape_size == inputs[i]->shape.size());
+                assert(inputs[i]->exponent == inputs[i - 1]->exponent);
+                output_shape_axis += inputs[i]->shape[axis];
+
+                for (int j = 0; j < shape_size; j++)
+                {
+                    if (j != axis)
+                    {
+                        assert(inputs[i]->shape[j] == inputs[i - 1]->shape[j]);
+                    }
+                }
+            }
+            DL_LOG_NN_LATENCY_END("assert");
+
+            DL_LOG_NN_LATENCY_START();
+            Tensor<feature_t> output;
+            std::vector<int> output_shape = inputs[0]->shape;
+            output_shape[axis] = output_shape_axis;
+            output.set_shape(output_shape);
+            output.set_exponent(inputs[0]->exponent);
+            output.malloc_element();
+            DL_LOG_NN_LATENCY_END("malloc");
+
+            DL_LOG_NN_LATENCY_START();
+            concat(output, inputs, axis, free_inputs);
+            DL_LOG_NN_LATENCY_END("concat");
+            return output;
+        }
+    } // namespace nn
+} // namespace dl
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_conv2d.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_conv2d.hpp
index 77e2c617d04..27ba0372730 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_conv2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_conv2d.hpp
@@ -10,7 +10,6 @@ namespace dl
     {
         /**
          * @brief activation(conv2d(input, filter) + bias).
-         * NOTE: When padding_type is SAME, make sure padding is already added in input.
          * 
          * @param output      as an output
          * @param input       as an input
@@ -34,7 +33,6 @@ namespace dl
 
         /**
          * @brief activation(conv2d(input, filter) + bias).
-         * NOTE: When padding_type is SAME, make sure padding is already added in input.
          * 
          * @param output      as an output
          * @param input       as an input
@@ -56,6 +54,29 @@ namespace dl
                     const Activation<int8_t> *const activation = NULL,
                     const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
 
+        /**
+         * @brief activation(conv2d(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param padding     padding size needed in [top, bottom, left, right] of this operation
+         * @param filter      filter of conv2d
+         * @param stride_y    stride in height
+         * @param stride_x    stride in width
+         * @param bias        bias of conv2d, if you don't specify anything, no bias is added
+         * @param activation  activation of conv2d, if you don't specify anything, no activation is applied
+         * @param assign_core not effective yet
+         */
+        void conv2d(Tensor<int8_t> &output,
+                    Tensor<int8_t> &input,
+                    std::vector<int> &padding,
+                    const Filter<int8_t> &filter,
+                    const int stride_y,
+                    const int stride_x,
+                    const Bias<int16_t> *const bias = NULL,
+                    const Activation<int8_t> *const activation = NULL,
+                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
         /**
          * @brief activation(conv2d(input, filter) + bias).
          * 
@@ -67,25 +88,25 @@ namespace dl
          * @param filter          Filter of conv2d
          * @param stride_y        stride in height
          * @param stride_x        stride in width
-         * @param padding_type    one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+         * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN,
          *                        - PADDING_VALID: no padding
-         *                        PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+         *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
          *                        such that output has the same height/width dimension as the input,
-         *                        - PADDING_SAME results padding in TensorFlow style
-         *                        - PADDING_SAME_MXNET results padding in MXNET style
+         *                        - PADDING_SAME_END results padding in TensorFlow style
+         *                        - PADDING_SAME_BEGIN results padding in MXNET style
          * @param bias            bias of conv2d, if you don't specify anything, no bias is added
          * @param activation      activation of conv2d, if you don't specify anything, no activation is applied
          * @param assign_core     not effective yet
          * @return conv2d result
          */
-        template <typename feature_t>
+        template <typename feature_t, typename bias_t>
         Tensor<feature_t> conv2d(const int output_exponent,
                                  Tensor<feature_t> &input,
                                  const Filter<feature_t> &filter,
                                  const int stride_y,
                                  const int stride_x,
                                  const padding_type_t padding_type,
-                                 const Bias<feature_t> *bias,
+                                 const Bias<bias_t> *bias,
                                  const Activation<feature_t> *activation,
                                  const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
         {
@@ -94,20 +115,19 @@ namespace dl
             DL_LOG_NN_LATENCY_START();
             std::vector<int> output_shape = get_output_shape(input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type, true);
             Tensor<feature_t> output;
-            output.set_exponent(output_exponent).set_shape(output_shape).apply_element();
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
             DL_LOG_NN_LATENCY_END("apply");
 
+            std::vector<int> padding(4, 0);
             DL_LOG_NN_LATENCY_START();
-            if (padding_type == PADDING_SAME || padding_type == PADDING_SAME_MXNET)
+            if (padding_type == PADDING_SAME_END || padding_type == PADDING_SAME_BEGIN)
             {
-                std::vector<int> padding = get_pad_size(output_shape, input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
-                input.set_padding_size(padding);
-                input.set_padding_value(padding, 0);
+                padding = get_pad_size(output_shape, input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
             }
             DL_LOG_NN_LATENCY_END("padding");
 
             DL_LOG_NN_LATENCY_START();
-            conv2d(output, input, input.padding, filter, stride_y, stride_x, bias, activation, assign_core);
+            conv2d(output, input, padding, filter, stride_y, stride_x, bias, activation, assign_core);
             DL_LOG_NN_LATENCY_END("conv2d");
 
             return output;
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_depthwise_conv2d.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_depthwise_conv2d.hpp
index 6e972bfd81b..135815a65cb 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_depthwise_conv2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_depthwise_conv2d.hpp
@@ -10,7 +10,6 @@ namespace dl
     {
         /**
          * @brief activate(depthwise_conv2d(input, filter) + bias)
-         * NOTE: When padding_type is SAME, make sure padding is already added in input
          * 
          * @param output      as an output
          * @param input       as an input
@@ -34,7 +33,6 @@ namespace dl
 
         /**
          * @brief activate(depthwise_conv2d(input, filter) + bias)
-         * NOTE: When padding_type is SAME, make sure padding is already added in input
          * 
          * @param output      as an output
          * @param input       as an input
@@ -56,6 +54,29 @@ namespace dl
                               const Activation<int8_t> *activation = NULL,
                               const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
 
+        /**
+         * @brief activate(depthwise_conv2d(input, filter) + bias)
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param padding     padding size needed in [top, bottom, left, right] of this operation
+         * @param filter      Filter of depthwise_conv2d
+         * @param stride_y    stride in height
+         * @param stride_x    stride in width
+         * @param bias        bias of depthwise_conv2d, if you don't specify anything, no bias is added
+         * @param activation  activation of depthwise_conv2d, if you don't specify anything, no activation is applied
+         * @param assign_core not effective yet
+         */
+        void depthwise_conv2d(Tensor<int8_t> &output,
+                              Tensor<int8_t> &input,
+                              std::vector<int> &padding,
+                              const Filter<int8_t> &filter,
+                              const int stride_y,
+                              const int stride_x,
+                              const Bias<int16_t> *bias = NULL,
+                              const Activation<int8_t> *activation = NULL,
+                              const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
         /**
          * @brief activation(depthwise_conv2d(input, filter) + bias)
          * 
@@ -67,25 +88,25 @@ namespace dl
          * @param filter          filter of depthwise_conv2d
          * @param stride_y        stride in height
          * @param stride_x        stride in width
-         * @param pad_type        one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+         * @param pad_type        one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN,
          *                        - PADDING_VALID means no padding
-         *                        PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+         *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
          *                        such that output has the same height/width dimension as the input,
-         *                        - PADDING_SAME results padding in TensorFlow style
-         *                        - PADDING_SAME_MXNET results padding in MXNET style
+         *                        - PADDING_SAME_END results padding in TensorFlow style
+         *                        - PADDING_SAME_BEGIN results padding in MXNET style
          * @param bias            bias of depthwise_conv2d, if you don't specify anything, no bias is added
          * @param activation      activation of depthwise_conv2d, if you don't specify anything, no activation is applied
          * @param assign_core     not effective yet
          * @return depthwise_conv2d result
          */
-        template <typename feature_t>
+        template <typename feature_t, typename bias_t>
         Tensor<feature_t> depthwise_conv2d(const int output_exponent,
                                            Tensor<feature_t> &input,
                                            const Filter<feature_t> &filter,
                                            const int stride_y,
                                            const int stride_x,
                                            const padding_type_t padding_type,
-                                           const Bias<feature_t> *bias,
+                                           const Bias<bias_t> *bias,
                                            const Activation<feature_t> *activation,
                                            const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
         {
@@ -94,20 +115,20 @@ namespace dl
             DL_LOG_NN_LATENCY_START();
             std::vector<int> output_shape = get_output_shape(input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
             Tensor<feature_t> output;
-            output.set_exponent(output_exponent).set_shape(output_shape).apply_element();
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
             DL_LOG_NN_LATENCY_END("apply");
 
+            std::vector<int> padding(4, 0);
+
             DL_LOG_NN_LATENCY_START();
-            if (padding_type == PADDING_SAME || padding_type == PADDING_SAME_MXNET)
+            if (padding_type == PADDING_SAME_END || padding_type == PADDING_SAME_BEGIN)
             {
-                std::vector<int> padding = get_pad_size(output_shape, input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
-                input.set_padding_size(padding);
-                input.set_padding_value(padding, 0);
+                padding = get_pad_size(output_shape, input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
             }
             DL_LOG_NN_LATENCY_END("padding");
 
             DL_LOG_NN_LATENCY_START();
-            depthwise_conv2d(output, input, input.padding, filter, stride_y, stride_x, bias, activation, assign_core);
+            depthwise_conv2d(output, input, padding, filter, stride_y, stride_x, bias, activation, assign_core);
             DL_LOG_NN_LATENCY_END("depthwise_conv2d");
 
             return output;
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_fully_connected.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_fully_connected.hpp
new file mode 100644
index 00000000000..372c84825fd
--- /dev/null
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_fully_connected.hpp
@@ -0,0 +1,126 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief activation(FullyConnected(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param filter      filter of FullyConnected
+         * @param bias        bias of FullyConnected, if you don't specify anything, no bias is added
+         * @param activation  activation of FullyConnected, if you don't specify anything, no activation is applied
+         * @param flatten     true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+         *                    false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+         * @param assign_core not effective yet
+         */
+        void fully_connected(Tensor<int16_t> &output,
+                             Tensor<int16_t> &input,
+                             const Filter<int16_t> &filter,
+                             const Bias<int16_t> *const bias = NULL,
+                             const Activation<int16_t> *const activation = NULL,
+                             const bool flatten = true,
+                             const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(FullyConnected(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param filter      filter of FullyConnected
+         * @param bias        bias of FullyConnected, if you don't specify anything, no bias is added
+         * @param activation  activation of FullyConnected, if you don't specify anything, no activation is applied
+         * @param flatten     true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+         *                    false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+         * @param assign_core not effective yet
+         */
+        void fully_connected(Tensor<int8_t> &output,
+                             Tensor<int8_t> &input,
+                             const Filter<int8_t> &filter,
+                             const Bias<int8_t> *const bias = NULL,
+                             const Activation<int8_t> *const activation = NULL,
+                             const bool flatten = true,
+                             const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(FullyConnected(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param filter      filter of FullyConnected
+         * @param bias        bias of FullyConnected, if you don't specify anything, no bias is added
+         * @param activation  activation of FullyConnected, if you don't specify anything, no activation is applied
+         * @param flatten     true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+         *                    false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+         * @param assign_core not effective yet
+         */
+        void fully_connected(Tensor<int8_t> &output,
+                             Tensor<int8_t> &input,
+                             const Filter<int8_t> &filter,
+                             const Bias<int16_t> *const bias = NULL,
+                             const Activation<int8_t> *const activation = NULL,
+                             const bool flatten = true,
+                             const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(FullyConnected(input, filter) + bias).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param output_exponent exponent of output
+         * @param input           as an input
+         * @param filter          Filter of FullyConnected
+         * @param bias            bias of FullyConnected, if you don't specify anything, no bias is added
+         * @param activation      activation of FullyConnected, if you don't specify anything, no activation is applied
+         * @param flatten         true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+         *                        false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+         * @param assign_core     not effective yet
+         * @return FullyConnected result
+         */
+        template <typename feature_t>
+        Tensor<feature_t> fully_connected(const int output_exponent,
+                                          Tensor<feature_t> &input,
+                                          const Filter<feature_t> &filter,
+                                          const Bias<feature_t> *bias,
+                                          const Activation<feature_t> *activation,
+                                          const bool flatten,
+                                          const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+        {
+            DL_LOG_NN_LATENCY_INIT();
+
+            DL_LOG_NN_LATENCY_START();
+            assert(filter.shape.size() == 4);
+            assert(filter.shape[0] == 1);
+            assert(filter.shape[1] == 1);
+
+            std::vector<int> output_shape;
+            if (flatten)
+            {
+                assert(input.get_size() == filter.shape[2]);
+                output_shape = {filter.shape.back()};
+            }
+            else
+            {
+                assert(input.shape.back() == filter->shape[2]);
+                output_shape = input.shape;
+                output_shape[output_shape.size() - 1] = filter.shape.back();
+            }
+            Tensor<feature_t> output;
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
+            DL_LOG_NN_LATENCY_END("apply");
+
+            DL_LOG_NN_LATENCY_START();
+            fully_connected(output, input, filter, bias, activation, flatten, assign_core);
+            DL_LOG_NN_LATENCY_END("fully_connected");
+
+            return output;
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_global_avg_pool2d.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_global_avg_pool2d.hpp
index 723ca9eddc4..724d3ca0952 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_global_avg_pool2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_global_avg_pool2d.hpp
@@ -53,7 +53,7 @@ namespace dl
             std::vector<int> output_shape(input.shape.size(), 1);
             output_shape[2] = input.shape[2];
             Tensor<feature_t> output;
-            output.set_exponent(output_exponent).set_shape(output_shape).apply_element();
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
             DL_LOG_NN_LATENCY_END("apply");
 
             DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_global_max_pool2d.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_global_max_pool2d.hpp
index 945645cfd55..f6f15e91bd3 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_global_max_pool2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_global_max_pool2d.hpp
@@ -51,7 +51,7 @@ namespace dl
             std::vector<int> output_shape(input.shape.size(), 1);
             output_shape[2] = input.shape[2];
             Tensor<feature_t> output;
-            output.set_exponent(input.exponent).set_shape(output_shape).apply_element();
+            output.set_exponent(input.exponent).set_shape(output_shape).malloc_element();
             DL_LOG_NN_LATENCY_END("apply");
 
             DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_leakyrelu.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_leakyrelu.hpp
index d3738fc44ca..c41728b1ad5 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_leakyrelu.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_leakyrelu.hpp
@@ -52,17 +52,17 @@ namespace dl
          * @return leakyrelu result or no return(result store to input)
          */
         template <bool inplace = false, typename feature_t>
-        auto leakyrelu(Tensor<feature_t> &input, 
-                        const int activation_alpha, 
-                        const int activation_exponent, 
-                        const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        auto leakyrelu(Tensor<feature_t> &input,
+                       const int activation_alpha,
+                       const int activation_exponent,
+                       const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
-            if constexpr(!inplace)
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(input.exponent).set_shape(input.shape).apply_element();
+                output.set_exponent(input.exponent).set_shape(input.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_max2d.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_max2d.hpp
index 0a5e8f43221..466089bc386 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_max2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_max2d.hpp
@@ -48,20 +48,20 @@ namespace dl
          * @return max2d result or no return(result store to input0)
          */
         template <bool inplace = false, typename feature_t>
-        auto max2d(Tensor<feature_t> &input0, 
-                    Tensor<feature_t> &input1, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        auto max2d(Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             assert(input0.is_same_shape(input1));
             assert(input0.exponent == input1.exponent);
 
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
-            
-            if constexpr(!inplace)
+
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(input0.exponent).set_shape(input0.shape).apply_element();
+                output.set_exponent(input0.exponent).set_shape(input0.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_max_pool2d.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_max_pool2d.hpp
index 7f95f3d4278..50d51728ce0 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_max_pool2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_max_pool2d.hpp
@@ -57,12 +57,12 @@ namespace dl
          * @param filter_shape filter shape in [filter_height, filter_width]
          * @param stride_y     stride in height
          * @param stride_x     stride in width
-         * @param padding_type one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+         * @param padding_type one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN,
          *                     - PADDING_VALID: no padding
-         *                     PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+         *                     PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
          *                     such that output has the same height/width dimension as the input,
-         *                     - PADDING_SAME results padding in TensorFlow style
-         *                     - PADDING_SAME_MXNET results padding in MXNET style
+         *                     - PADDING_SAME_END results padding in TensorFlow style
+         *                     - PADDING_SAME_BEGIN results padding in MXNET style
          * @param assign_core  not effective yet
          * @return max_pool2d result
          */
@@ -79,20 +79,20 @@ namespace dl
             DL_LOG_NN_LATENCY_START();
             std::vector<int> output_shape = get_output_shape(input.shape, filter_shape, stride_y, stride_x, padding_type);
             Tensor<feature_t> output;
-            output.set_exponent(input.exponent).set_shape(output_shape).apply_element();
+            output.set_exponent(input.exponent).set_shape(output_shape).malloc_element();
             DL_LOG_NN_LATENCY_END("apply");
 
+            std::vector<int> padding(4, 0);
+
             DL_LOG_NN_LATENCY_START();
-            if (padding_type == PADDING_SAME || padding_type == PADDING_SAME_MXNET)
+            if (padding_type == PADDING_SAME_END || padding_type == PADDING_SAME_BEGIN)
             {
-                std::vector<int> padding = get_pad_size(output_shape, input.shape, filter_shape, stride_y, stride_x, padding_type);
-                input.set_padding_size(padding);
-                input.set_padding_value(padding, 0);
+                padding = get_pad_size(output_shape, input.shape, filter_shape, stride_y, stride_x, padding_type);
             }
             DL_LOG_NN_LATENCY_END("padding");
 
             DL_LOG_NN_LATENCY_START();
-            max_pool2d(output, input, input.padding, filter_shape, stride_y, stride_x, assign_core);
+            max_pool2d(output, input, padding, filter_shape, stride_y, stride_x, assign_core);
             DL_LOG_NN_LATENCY_END("max_pool2d");
 
             return output;
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_min2d.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_min2d.hpp
index 71cb87d50d5..8faddf3c228 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_min2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_min2d.hpp
@@ -47,20 +47,20 @@ namespace dl
          * @return min2d result or no return(result store to input0)
          */
         template <bool inplace = false, typename feature_t>
-        auto min2d(Tensor<feature_t> &input0, 
-                    Tensor<feature_t> &input1, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        auto min2d(Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             assert(input0.is_same_shape(input1));
             assert(input0.exponent == input1.exponent);
 
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
-            
-            if constexpr(!inplace)
+
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(input0.exponent).set_shape(input0.shape).apply_element();
+                output.set_exponent(input0.exponent).set_shape(input0.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_mul2d.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_mul2d.hpp
index 410528a05a3..909619a8767 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_mul2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_mul2d.hpp
@@ -18,12 +18,12 @@ namespace dl
          * @param assign_core       not effective yet
          * @param output_exponent   exponent of output, only and must specify if inplace operation happens
          */
-        void mul2d(Tensor<int16_t> &output, 
-                    Tensor<int16_t> &input0, 
-                    Tensor<int16_t> &input1, 
-                    const Activation<int16_t> *const activation = NULL, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
-                    const int output_exponent = INT_MIN);
+        void mul2d(Tensor<int16_t> &output,
+                   Tensor<int16_t> &input0,
+                   Tensor<int16_t> &input1,
+                   const Activation<int16_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
 
         /**
          * @brief activation(mul2d(input0, input1)).
@@ -35,12 +35,12 @@ namespace dl
          * @param assign_core       not effective yet
          * @param output_exponent   exponent of output, only and must specify if inplace operation happens
          */
-        void mul2d(Tensor<int8_t> &output, 
-                    Tensor<int8_t> &input0, 
-                    Tensor<int8_t> &input1, 
-                    const Activation<int8_t> *const activation = NULL, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
-                    const int output_exponent = INT_MIN);
+        void mul2d(Tensor<int8_t> &output,
+                   Tensor<int8_t> &input0,
+                   Tensor<int8_t> &input1,
+                   const Activation<int8_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
 
         /**
          * @brief activation(mul2d(input0, input1)).
@@ -57,21 +57,21 @@ namespace dl
          * @return mul2d result or no return(result store to input0)
          */
         template <bool inplace = false, typename feature_t>
-        auto mul2d(const int output_exponent, 
-                    Tensor<feature_t> &input0, 
-                    Tensor<feature_t> &input1, 
-                    const Activation<feature_t> *activation, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        auto mul2d(const int output_exponent,
+                   Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const Activation<feature_t> *activation,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             assert(input0.is_same_shape(input1));
 
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
 
-            if constexpr(!inplace)
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(output_exponent).set_shape(input0.shape).apply_element();
+                output.set_exponent(output_exponent).set_shape(input0.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_prelu.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_prelu.hpp
index fb4315d9fc3..e83e8975604 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_prelu.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_prelu.hpp
@@ -52,17 +52,17 @@ namespace dl
          * @return prelu result or no return(result store to input)
          */
         template <bool inplace = false, typename feature_t>
-        auto prelu(Tensor<feature_t> &input, 
-                    const feature_t *activation_element, 
-                    const int activation_exponent, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        auto prelu(Tensor<feature_t> &input,
+                   const feature_t *activation_element,
+                   const int activation_exponent,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
-            if constexpr(!inplace)
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(input.exponent).set_shape(input.shape).apply_element();
+                output.set_exponent(input.exponent).set_shape(input.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
@@ -76,7 +76,7 @@ namespace dl
                 DL_LOG_NN_LATENCY_START();
                 prelu(input, input, activation_element, activation_exponent, assign_core);
                 DL_LOG_NN_LATENCY_END("prelu");
-            } 
+            }
         }
     } // namespace nn
 } // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_relu.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_relu.hpp
index e4159fdf898..308492dfe4b 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_relu.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_relu.hpp
@@ -15,9 +15,9 @@ namespace dl
          * @param input       as an input
          * @param assign_core not effective yet
          */
-        void relu(Tensor<int16_t> &output, 
-                    Tensor<int16_t> &input, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+        void relu(Tensor<int16_t> &output,
+                  Tensor<int16_t> &input,
+                  const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
 
         /**
          * @brief relu(input).
@@ -26,9 +26,9 @@ namespace dl
          * @param input       as an input
          * @param assign_core not effective yet
          */
-        void relu(Tensor<int8_t> &output, 
-                    Tensor<int8_t> &input, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+        void relu(Tensor<int8_t> &output,
+                  Tensor<int8_t> &input,
+                  const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
 
         /**
          * @brief relu(input)
@@ -46,11 +46,11 @@ namespace dl
         {
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
-            
-            if constexpr(!inplace)
+
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(input.exponent).set_shape(input.shape).apply_element();
+                output.set_exponent(input.exponent).set_shape(input.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_sub2d.hpp b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_sub2d.hpp
index 385188e4ba2..5bbd4940b10 100644
--- a/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_sub2d.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/nn/dl_nn_sub2d.hpp
@@ -18,12 +18,12 @@ namespace dl
          * @param assign_core not effective yet
          * @param output_exponent   exponent of output, only and must specify if inplace operation happens 
          */
-        void sub2d(Tensor<int16_t> &output, 
-                    Tensor<int16_t> &input0, 
-                    Tensor<int16_t> &input1, 
-                    const Activation<int16_t> *const activation = NULL, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
-                    const int output_exponent = INT_MIN);
+        void sub2d(Tensor<int16_t> &output,
+                   Tensor<int16_t> &input0,
+                   Tensor<int16_t> &input1,
+                   const Activation<int16_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
 
         /**
          * @brief activation(sub2d(input0, input1)).
@@ -35,12 +35,12 @@ namespace dl
          * @param assign_core not effective yet
          * @param output_exponent   exponent of output, only and must specify if inplace operation happens
          */
-        void sub2d(Tensor<int8_t> &output, 
-                    Tensor<int8_t> &input0, 
-                    Tensor<int8_t> &input1, 
-                    const Activation<int8_t> *const activation = NULL, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
-                    const int output_exponent = INT_MIN);
+        void sub2d(Tensor<int8_t> &output,
+                   Tensor<int8_t> &input0,
+                   Tensor<int8_t> &input1,
+                   const Activation<int8_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
 
         /**
          * @brief activation(sub2d(input0, input1)).
@@ -57,20 +57,20 @@ namespace dl
          * @return sub2d result or no return(result store to input0)
          */
         template <bool inplace = false, typename feature_t>
-        auto sub2d(const int output_exponent, 
-                    Tensor<feature_t> &input0, 
-                    Tensor<feature_t> &input1, 
-                    const Activation<feature_t> *activation, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        auto sub2d(const int output_exponent,
+                   Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const Activation<feature_t> *activation,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             assert(input0.is_same_shape(input1));
 
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
-            if constexpr(!inplace)
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(output_exponent).set_shape(input0.shape).apply_element();
+                output.set_exponent(output_exponent).set_shape(input0.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32/include/esp-face/include/tool/dl_tool.hpp b/tools/sdk/esp32/include/esp-face/include/tool/dl_tool.hpp
index f8e0871a04f..e5490e073d1 100644
--- a/tools/sdk/esp32/include/esp-face/include/tool/dl_tool.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/tool/dl_tool.hpp
@@ -67,62 +67,49 @@ namespace dl
         void copy_memory(void *dst, void *src, const int n);
 
         /**
-         * @brief Apply memory without initialized. Must use free_aligned() to free the memory.
+         * @brief Apply memory without initialized. Can use free_aligned() to free the memory.
          * 
          * @param number number of elements
          * @param size   size of element
-         * @param align  number of aligned, e.g., 16 means 16-byte aligned
+         * @param align  number of byte aligned, e.g., 16 means 16-byte aligned
          * @return pointer of allocated memory. NULL for failed
          */
-        inline void *malloc_aligned(int number, int size, int align = 0)
+        inline void *malloc_aligned(int number, int size, int align = 4)
         {
-            int n = number * size;
-            n >>= 4;
-            n += 2;
-            n <<= 4;
-            int total_size = n + align + sizeof(void *) + sizeof(int);
-            void *res = malloc(total_size);
+            assert((align > 0) && (((align & (align-1)) == 0)));
+            int total_size = number * size;
+
+            void *res = heap_caps_aligned_alloc(align, total_size, MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL);
 #if DL_SPIRAM_SUPPORT
             if (NULL == res)
-                res = heap_caps_malloc(total_size, MALLOC_CAP_SPIRAM);
+                res = heap_caps_aligned_alloc(align, total_size, MALLOC_CAP_SPIRAM);
 #endif
             if (NULL == res)
             {
                 printf("Fail to malloc %d bytes from DRAM(%d bytyes) and PSRAM(%d bytes), PSRAM is %s.\n",
                        total_size,
-                       heap_caps_get_free_size(MALLOC_CAP_INTERNAL),
+                       heap_caps_get_free_size(MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL),
                        heap_caps_get_free_size(MALLOC_CAP_SPIRAM),
                        DL_SPIRAM_SUPPORT ? "on" : "off");
                 return NULL;
             }
-            void **data = (void **)res + 2; // 4-byte for pointer, 4-bytes for n
-            void **aligned;
-            if (align)
-                aligned = (void **)(((size_t)data + (align - 1)) & -align);
-            else
-                aligned = data;
-
-            aligned[-1] = res;
-            int *temp = (int *)aligned;
-            temp[-2] = n;
 
-            return (void *)aligned;
+            return (void *)res;
         }
 
         /**
-         * @brief Apply memory with zero-initialized. Must use dl_lib_free() to free the memory.
+         * @brief Apply memory with zero-initialized. Can use free_aligned() to free the memory.
          * 
          * @param number number of elements
          * @param size   size of element
-         * @param align  number of aligned, e.g., 16 means 16-byte aligned
+         * @param align  number of byte aligned, e.g., 16 means 16-byte aligned
          * @return pointer of allocated memory. NULL for failed
          */
-        inline void *calloc_aligned(int number, int size, int align = 0)
+        inline void *calloc_aligned(int number, int size, int align = 4)
         {
 
             void *aligned = malloc_aligned(number, size, align);
-            int n = *((int *)aligned - 2);
-            set_zero(aligned, n);
+            set_zero(aligned, number * size);
 
             return (void *)aligned;
         }
@@ -137,7 +124,70 @@ namespace dl
             if (NULL == address)
                 return;
 
-            free(((void **)address)[-1]);
+            heap_caps_free(address);
+        }
+
+        /**
+         * @brief Apply memory without initialized in preference order: internal aligned, internal, external aligned
+         * 
+         * @param number number of elements
+         * @param size   size of element
+         * @param align  number of byte aligned, e.g., 16 means 16-byte aligned
+         * @return pointer of allocated memory. NULL for failed
+         */
+        inline void *malloc_aligned_prefer(int number, int size, int align = 4)
+        {
+            assert((align > 0) && (((align & (align-1)) == 0)));
+            int total_size = number * size;
+            void *res = heap_caps_aligned_alloc(align, total_size, MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL);
+            if (NULL == res){
+                res = heap_caps_malloc(total_size, MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL);
+            }
+#if DL_SPIRAM_SUPPORT
+            if (NULL == res){
+                res = heap_caps_aligned_alloc(align, total_size, MALLOC_CAP_SPIRAM);
+            }
+#endif
+            if (NULL == res)
+            {
+                printf("Fail to malloc %d bytes from DRAM(%d bytyes) and PSRAM(%d bytes), PSRAM is %s.\n",
+                       total_size,
+                       heap_caps_get_free_size(MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL),
+                       heap_caps_get_free_size(MALLOC_CAP_SPIRAM),
+                       DL_SPIRAM_SUPPORT ? "on" : "off");
+                return NULL;
+            }
+
+            return res;
+        }
+
+        /**
+         * @brief Apply memory with zero-initialized in preference order: internal aligned, internal, external aligned
+         * 
+         * @param number number of elements
+         * @param size   size of element
+         * @param align  number of byte aligned, e.g., 16 means 16-byte aligned
+         * @return pointer of allocated memory. NULL for failed
+         */
+        inline void *calloc_aligned_prefer(int number, int size, int align = 4)
+        {
+            void *res = malloc_aligned_prefer(number, size, align);
+            set_zero(res, number * size);
+
+            return (void *)res;
+        }
+
+        /**
+         * @brief Free the calloc_aligned_prefer() and malloc_aligned_prefer() memory
+         * 
+         * @param address pointer of memory to free
+         */
+        inline void free_aligned_prefer(void *address)
+        {
+            if (NULL == address)
+                return;
+
+            heap_caps_free(address);
         }
 
         /**
diff --git a/tools/sdk/esp32/include/esp-face/include/typedef/dl_constant.hpp b/tools/sdk/esp32/include/esp-face/include/typedef/dl_constant.hpp
index 73e3b0832e7..07b2dd24ee1 100644
--- a/tools/sdk/esp32/include/esp-face/include/typedef/dl_constant.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/typedef/dl_constant.hpp
@@ -57,7 +57,8 @@ namespace dl
          * @param exponent exponent of element
          * @param shape    shape of Filter,
          *                 - 1D: reserved
-         *                 - 2D: [filter_height, filter_width, input_channel, output_channel]
+         *                 - 2D: for convolution is [filter_height, filter_width, input_channel, output_channel],
+         *                       for depthwise convolution is [filter_height, filter_width, input_channel, 1]
          * @param dilation dilation of Filter
          *                 - 1D: reserved
          *                 - 2D: [dilation_in_height, dilation_in_width]
@@ -97,6 +98,9 @@ namespace dl
     {
     public:
         using Constant<T>::Constant;
+        std::vector<int> channel_exponent;    /*<! exponent for per-channel >*/
+
+        Bias(const T *element, const std::vector<int> channel_exponent, const std::vector<int> shape);
     };
 
     /**
diff --git a/tools/sdk/esp32/include/esp-face/include/typedef/dl_variable.hpp b/tools/sdk/esp32/include/esp-face/include/typedef/dl_variable.hpp
index bf6e8630856..18cbe9707e9 100644
--- a/tools/sdk/esp32/include/esp-face/include/typedef/dl_variable.hpp
+++ b/tools/sdk/esp32/include/esp-face/include/typedef/dl_variable.hpp
@@ -3,6 +3,7 @@
 #include <stdio.h>
 #include <vector>
 #include <assert.h>
+#include <iostream>
 
 #include "dl_tool.hpp"
 
@@ -17,27 +18,20 @@ namespace dl
     class Tensor
     {
     private:
-        int size;       /*<! size of element including padding */
-        bool auto_free; /*<! free element when object destroy */
+        int size;                     /*<! size of element including padding */
+        bool auto_free;               /*<! free element when object destroy */
+        std::vector<int> axis_offset; /*<! element offset of each axis */
 
     public:
-        T *element;                          /*<! point to element */
-        int exponent;                        /*<! exponent of element */
-        std::vector<int> shape;              /*<! shape of Tensor */
-                                             /*<! 2D: shape is [height, width, channel] */
-                                             /*<! 1D: reserved */
-        std::vector<int> shape_with_padding; /*<! shape with padding of Tensor */
-                                             /*<! 2D: shape_with_padding is [height_with_padding, width_with_padding, channel_with_padding] */
-                                             /*<! 1D: reserved */
-        std::vector<int> padding;            /*<! padding of Tensor */
-                                             /*<!- 2D: padding format is [top, bottom, left, right] */
-                                             /*<! - 1D: reserved */
+        T *element;             /*<! point to element */
+        int exponent;           /*<! exponent of element */
+        std::vector<int> shape; /*<! shape of Tensor */
 
         /**
          * @brief Construct a new Tensor object
          * 
          */
-        Tensor() : size(-1), auto_free(true), element(NULL), exponent(0) {}
+        Tensor() : auto_free(true), element(NULL), exponent(0) { this->set_shape({0}); }
 
         /**
          * @brief Construct a new Tensor object by copying from input.
@@ -49,21 +43,20 @@ namespace dl
          */
         Tensor(Tensor<T> &input, bool deep) : size(input.size),
                                               auto_free(input.auto_free),
-                                              exponent(input.exponent),
-                                              shape(input.shape),
-                                              shape_with_padding(input.shape_with_padding),
-                                              padding(input.padding)
+                                              exponent(input.exponent)
         {
-            if (deep)
+            this->set_shape(input.shape);
+            if (deep && (input.element != NULL))
             {
-                int size_real = input.shape_with_padding.size() ? input.shape_with_padding[0] * input.shape_with_padding[1] * input.shape_with_padding[2] : 0;
-                T *new_element = (T *)tool::calloc_aligned(size_real, sizeof(T), 16);
+                int size_real = input.get_size();
+                T *new_element = (T *)tool::calloc_aligned_prefer(size_real, sizeof(T), 16);
                 tool::copy_memory(new_element, input.element, size_real * sizeof(T));
                 this->element = new_element;
             }
             else
             {
                 this->element = input.element;
+                this->auto_free = false;
             }
         }
 
@@ -77,6 +70,33 @@ namespace dl
                 this->free_element();
         }
 
+        /**
+         * @brief 
+         * 
+         * @param input an input Tensor
+         * @param deep one of true or false
+         *              - true: apply a new memory, copy value from input.element to this new memory
+         *              - false: take over input.element to this->element
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &copy_element(Tensor<T> &input, bool deep)
+        {
+            assert(this->get_size() == input.get_size());
+            assert(input.element != NULL);
+
+            this->malloc_element();
+            if (deep)
+            {
+                tool::copy_memory(this->element, input.element, this->get_size() * sizeof(T));
+            }
+            else
+            {
+                this->element = input.element;
+                this->auto_free = false;
+            }
+            return *this;
+        }
+
         /**
          * @brief Set the auto free object.
          * 
@@ -120,190 +140,144 @@ namespace dl
         }
 
         /**
-         * @brief Set the shape of Tensor. Initial this->padding = {0}. Initial this->size = -1.
+         * @brief Set the shape of Tensor.
          * 
-         * @param shape shape in 
-         *              - 2D: [height, width]
+         * @param shape the target shape 
+         *        
          * @return self
          */
-        Tensor<T> &set_shape(const std::vector<int> shape)
-        {
-            for (int i = 0; i < shape.size(); ++i)
-            {
-                assert(shape[i] > 0);
-            }
-            this->shape = shape;
-            this->shape_with_padding = shape;
-            this->size = -1;
-            this->padding = std::vector<int>(((this->shape.size() - 1) << 1), 0);
-            return *this;
-        }
+        Tensor<T> &set_shape(const std::vector<int> shape);
 
         /**
-         * @brief Set the padding size object.
+         * @brief print the shape of the Tensor
          * 
-         * @param padding padding size in
-         *                - 2D: [top, bottom, left, right]
-         * @return self
          */
-        Tensor &set_padding_size(std::vector<int> &padding)
+        void print_shape()
         {
-            assert(this->shape.size());      // call Tensor.set_shape() first
-            assert(this->shape.size() == 3); // TODO: || this->shape.size() == 2
-
-            if (this->shape.size() == 3)
+            if (this->shape.size())
             {
-                std::vector<int> new_padding = this->padding;
-                bool dont_update = true;
-
-                if (padding[0] > this->padding[0])
+                printf("shape = (");
+                for (int i = 0; i < this->shape.size() - 1; i++)
                 {
-                    new_padding[0] = padding[0];
-                    dont_update = false;
-                }
-
-                if (padding[1] > this->padding[1])
-                {
-                    new_padding[1] = padding[1];
-                    dont_update = false;
-                }
-
-                if (padding[2] > this->padding[2])
-                {
-                    new_padding[2] = padding[2];
-                    dont_update = false;
-                }
-
-                if (padding[3] > this->padding[3])
-                {
-                    new_padding[3] = padding[3];
-                    dont_update = false;
+                    printf("%d, ", this->shape[i]);
                 }
+                printf("%d)\n", this->shape.back());
+            }
+            else
+            {
+                printf("shape = ()\n");
+            }
+        }
 
-                if (dont_update)
-                {
-                    return *this;
-                }
+        /**
+         * @brief flatten the Tensor
+         * 
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &flatten();
 
-                std::vector<int> new_shape_with_padding = this->shape;
+        /**
+         * @brief Change a new shape to the Tensor without changing its data.
+         * 
+         * @param shape  the target shape
+         * @return Tensor<T>&  self
+         */
+        Tensor<T> &reshape(std::vector<int> shape);
 
-                new_shape_with_padding[0] += (new_padding[0] + new_padding[1]);
-                new_shape_with_padding[1] += (new_padding[2] + new_padding[3]);
-                int new_size = new_shape_with_padding[0] * new_shape_with_padding[1] * new_shape_with_padding[2];
+        /**
+         * @brief Remove dims with length==1 from Tensor
+         * 
+         * @param axis the dim to to be remove. make sure the length of the dim is equal to 1.
+         *              if axis == INT32_MAX, all the dims with length==1 will be removed.
+         * @return Tensor<T>& self 
+         */
+        Tensor<T> &squeeze(int axis = INT32_MAX);
 
-                if (this->element) // if this->element != NULL, do padding by copy memory
-                {
-                    T *new_element = (T *)tool::malloc_aligned(new_size, sizeof(T), 16);
-                    T *dst = new_element + ((new_padding[0] * new_shape_with_padding[1]) + new_padding[2]) * new_shape_with_padding[2];
-                    T *src = this->get_element_ptr();
-                    int offset_dst_next_y = new_shape_with_padding[1] * new_shape_with_padding[2];     // width * channel
-                    int src_copy_length = this->shape[1] * this->shape[2];                             // width * channel
-                    int offset_src_next_y = this->shape_with_padding[1] * this->shape_with_padding[2]; // width * channel
-                    for (int y = 0; y < this->shape[0]; y++)
-                    {
-                        tool::copy_memory(dst, src, src_copy_length * sizeof(T));
-                        dst += offset_dst_next_y;
-                        src += offset_src_next_y;
-                    }
+        /**
+         * @brief Insert a new dim that will appear at the axis position in the expanded Tensor shape.
+         * 
+         * @param axis the dim to be inserted
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &expand_dims(int axis);
 
-                    if (this->auto_free)
-                        tool::free_aligned(this->element);
-                    this->element = new_element;
-                    this->auto_free = true;
-                }
-                this->padding = new_padding;
-                this->shape_with_padding = new_shape_with_padding;
-                this->size = new_size;
-            }
-            else if (this->shape.size() == 2)
-            {
-                printf("Tensor.set_padding_size with this->shape.size() == 2 not implement yet.\n");
-            }
+        /**
+         * @brief Insert a new dim that will appear at the axis position in the expanded Tensor shape.
+         * 
+         * @param axis  the dim to be inserted
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &expand_dims(std::vector<int> axis);
 
-            return *this;
-        }
+        /**
+         * @brief Reverse or permute the axes of the Tensor
+         * 
+         * @param perm the new arangement of the dims. if perm == {}, the dims arangement will be reversed. 
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &transpose(std::vector<int> perm = {});
 
         /**
-         * @brief Set the padding value object.
+         * @brief Reverse or permute the axes of the input Tensor
          * 
-         * @param padding padding size in
-         *                - 2D: [top, bottom, left, right]
-         * @param value   value to set
-         * @return self
+         * @param input the input Tensor
+         * @param perm the new arangement of the dims. if perm == {}, the dims arangement will be reversed. 
+         * @return Tensor<T>& self
          */
-        Tensor<T> &set_padding_value(std::vector<int> &padding, T value);
+        Tensor<T> &transpose(Tensor<T> &input, std::vector<int> perm = {});
 
         /**
          * @brief Get the element pointer.
          * 
-         * @param padding padding size in
-         *                - 2D: [top, bottom, left, right]
-         * @return pointer to memory with padding
+         * @return pointer to memory
          */
-        T *get_element_ptr(const std::vector<int> padding = {0, 0, 0, 0})
+        T *get_element_ptr()
         {
-            assert(this->shape.size() == 3); // TODO: || this->shape.size() == 2
-
-            if (this->shape.size() == 3)
-            {
-                return this->element + ((this->padding[0] - padding[0]) * this->shape_with_padding[1] + (this->padding[2] - padding[2])) * this->shape_with_padding[2];
-            }
-            else if (this->shape.size() == 2)
-            {
-                printf("Tensor.get_element_ptr with this->shape.size() == 2 is not implemented.\n");
-            }
-
-            return NULL;
+            return this->element;
         }
 
         /**
          * @brief Get the element value.
          * 
-         * @param index        index in
-         *                     - 2D: [y, x, c]
-         * @param with_padding one of true or false,
-         *                     - true: make padding size in count
-         *                     - false: do not
-         * @return element value
+         * @param index   the index of each dim. 
+         * @return T element value
          */
-        T &get_element_value(const std::vector<int> index, const bool with_padding = false)
+        T get_element_value(const std::vector<int> index)
         {
-            assert(index.size() == this->shape.size());
-            assert(this->shape.size() == 3); // TODO: || this->shape() == 2
-
-            int i = 0;
-            if (this->shape.size() == 3)
-            {
-                int y = index[0];
-                int x = index[1];
-                int c = index[2];
-                i = with_padding ? (y * this->shape_with_padding[1] + x) * this->shape_with_padding[2] + c : ((y + this->padding[0]) * this->shape_with_padding[1] + x + this->padding[2]) * this->shape_with_padding[2] + c;
-            }
-            else if (this->shape.size() == 2)
-            {
-                printf("Tensor.get_element_value with this->shape.size() == 2 is not implemented.\n");
-            }
+            return this->element[this->get_element_index(index)];
+        }
 
-            return this->element[i];
+        /**
+         * @brief Get the element value.
+         * 
+         * @param index  the index of the element.
+         * @return T  element value
+         */
+        T get_element_value(int index)
+        {
+            return this->element[index];
         }
 
         /**
-         * @brief Get the size of element.
+         * @brief Get the size of Tensor.
          * 
-         * @return size of element including padding
+         * @return  the size of Tensor.
          */
         int get_size()
         {
-            if (this->size == -1) // didn't call Tensor.set_padding_size() before
-            {
-                this->size = 1;
-                for (std::vector<int>::iterator d = this->shape.begin(); d != this->shape.end(); d++)
-                    this->size *= *d;
-            }
-
             return this->size;
         }
 
+        /**
+         * @brief Get the axis offset
+         * 
+         * @return std::vector<int> the axis offset
+         */
+        std::vector<int> get_axis_offset()
+        {
+            return this->axis_offset;
+        }
+
         /**
          * @brief Apply memory with zero-initialized only if this->element is NULL.
          * 
@@ -319,7 +293,7 @@ namespace dl
             if (this->element != NULL)
                 return false;
 
-            this->element = (T *)dl::tool::calloc_aligned(this->get_size(), sizeof(T), 16);
+            this->element = (T *)dl::tool::calloc_aligned_prefer(this->get_size(), sizeof(T), 16);
             this->auto_free = auto_free;
 
             return true;
@@ -340,31 +314,7 @@ namespace dl
             if (this->element != NULL)
                 return false;
 
-            this->element = (T *)tool::malloc_aligned(this->get_size(), sizeof(T), 16);
-            this->auto_free = auto_free;
-
-            return true;
-        }
-
-        /**
-         * @brief If this->element != NULL no memory will be applied and no value will be set in padding.
-         * Else apply memory without initialized and set value to padding.
-         * 
-         * @param padding_value value to set in padding
-         * @param auto_free     one of true of false
-         *                      - true: free element when object destroyed
-         *                      - false: do not
-         * @return 
-         *         - true: apply memory and set padding value successfully
-         *         - false: no memory applied and no padding value set
-         */
-        bool apply_element(const T padding_value = 0, const bool auto_free = true)
-        {
-            if (this->element != NULL)
-                return false;
-
-            this->element = (T *)tool::malloc_aligned(this->get_size(), sizeof(T), 16);
-            this->set_padding_value(this->padding, padding_value);
+            this->element = (T *)tool::malloc_aligned_prefer(this->get_size(), sizeof(T), 16);
             this->auto_free = auto_free;
 
             return true;
@@ -379,258 +329,56 @@ namespace dl
         {
             if (this->auto_free && this->element)
             {
-                tool::free_aligned(this->element);
+                tool::free_aligned_prefer(this->element);
                 this->element = NULL;
             }
         }
 
         /**
-         * @brief Print the shape of Tensor in format "shape = ({top_padding} + {height} + {bottom_padding}, {left_padding} + {width} + {right_padding}, {channel}(channel_with_padding))\n".
+         * @brief print the element of the tensor
+         * 
+         * @param axis_index_range  the element range of each dims to be print. if axis_index_range == {}, all the element will be print. 
+         * @param message  to print  
          */
-        void print_shape()
-        {
-            printf("shape = (%d + %d + %d, %d + %d + %d, %d(%d))\n",
-                   this->padding[0], this->shape[0], this->padding[1],
-                   this->padding[2], this->shape[1], this->padding[3],
-                   this->shape[2], this->shape_with_padding[2]);
-        }
+        void print(std::vector<int> axis_index_range = {}, const char *message = "");
 
         /**
-         * @brief Take numpy for example, this function print Tensor[y_start:y_end, x_start:x_end, c_start:c_end].
+         * @brief  print all the element of the Tensor.
          * 
-         * inner box is effective value of Tensor, "0" around is padding.
-         * 
-         * (with padding)
-         *               00000000000000000000000000000000000000000000000000
-         *               00000000000000000000000000000000000000000000000000
-         *               00000000000000000000000000000000000000000000000000
-         *               000000(without padding)                   00000000
-         *               000000                                    00000000
-         *               000000                                    00000000
-         *               000000          effective value           00000000
-         *               000000                                    00000000
-         *               000000                                    00000000
-         *               00000000000000000000000000000000000000000000000000
-         *               00000000000000000000000000000000000000000000000000
-         *               00000000000000000000000000000000000000000000000000
-         * 
-         * @param y_start start index in height
-         * @param y_end   end index in height
-         * @param x_start start index in width
-         * @param x_end   end index in width
-         * @param c_start start index in channel
-         * @param c_end   end index in channel
-         * @param message to print
-         * @param axis    print aligned this axis, effective only if all y_end - y_start, x_end - x_start and c_end - c_start equals to 1
+         * @param message to print  
          * @param with_padding one of true or false,
-         *                     - true: count from (with padding) in upper image
-         *                     - false: count from (without padding) in upper image
+         *                     - true: the padding element will also be ed
+         *                     - false: the padding element will not be ed
          */
-        void print(int y_start, int y_end,
-                   int x_start, int x_end,
-                   int c_start, int c_end,
-                   const char *message, int axis = 0, const bool with_padding = false)
+        void print_all(const char *message = "")
         {
-            assert(y_end > y_start);
-            assert(x_end > x_start);
-            assert(c_end > c_start);
-
-            y_start = DL_MAX(y_start, 0);
-            x_start = DL_MAX(x_start, 0);
-            c_start = DL_MAX(c_start, 0);
-            if (with_padding)
-            {
-                y_end = DL_MIN(y_end, this->shape_with_padding[0]);
-                x_end = DL_MIN(x_end, this->shape_with_padding[1]);
-                c_end = DL_MIN(c_end, this->shape_with_padding[2]);
-            }
-            else
-            {
-                y_end = DL_MIN(y_end, this->shape[0]);
-                x_end = DL_MIN(x_end, this->shape[1]);
-                c_end = DL_MIN(c_end, this->shape[2]);
-            }
-
-            printf("%s[%d:%d, %d:%d, %d:%d] | ", message, y_start, y_end, x_start, x_end, c_start, c_end);
+            std::cout << "\n"
+                      << message << " | ";
             this->print_shape();
 
-            if (y_end - y_start == 1)
-            {
-                if (x_end - x_start == 1)
-                {
-                    for (int c = c_start; c < c_end; c++)
-                        printf("%7d", c);
-                    printf("\n");
-
-                    for (int c = c_start; c < c_end; c++)
-                        printf("%7d", this->get_element_value({y_start, x_start, c}, with_padding));
-                    printf("\n");
-
-                    return;
-                }
-                else
-                {
-                    if (c_end - c_start == 1)
-                    {
-                        for (int x = x_start; x < x_end; x++)
-                            printf("%7d", x);
-                        printf("\n");
-
-                        for (int x = x_start; x < x_end; x++)
-                            printf("%7d", this->get_element_value({y_start, x, c_start}, with_padding));
-                        printf("\n");
-
-                        return;
-                    }
-                }
-            }
-            else
+            for (int i = 0; i < this->get_size(); i++)
             {
-                if (x_end - x_start == 1)
-                {
-                    if (c_end - c_start == 1)
-                    {
-                        for (int y = y_start; y < y_end; y++)
-                            printf("%7d", y);
-                        printf("\n");
-
-                        for (int y = y_start; y < y_end; y++)
-                            printf("%7d", this->get_element_value({y, x_start, c_start}, with_padding));
-                        printf("\n");
-
-                        return;
-                    }
-                }
+                std::cout << this->element[i] << " ";
             }
-
-            if (y_end - y_start == 1)
-                axis = 0;
-
-            if (x_end - x_start == 1)
-                axis = 1;
-
-            if (c_end - c_start == 1)
-                axis = 2;
-
-            if (axis == 0)
-            {
-                // ______c
-                // |
-                // |
-                // x
-                //
-                for (int y = y_start; y < y_end; y++)
-                {
-                    printf("y = %d\n     ", y);
-
-                    for (int c = c_start; c < c_end; c++)
-                        printf("%7d", c);
-                    printf("\n");
-
-                    for (int x = x_start; x < x_end; x++)
-                    {
-                        printf("%5d", x);
-                        for (int c = c_start; c < c_end; c++)
-                            printf("%7d", this->get_element_value({y, x, c}, with_padding));
-                        printf("\n");
-                    }
-                    printf("\n");
-                }
-            }
-            else if (axis == 1)
-            {
-                // ______c
-                // |
-                // |
-                // y
-                //
-                for (int x = x_start; x < x_end; x++)
-                {
-                    printf("x = %d\n     ", x);
-
-                    for (int c = c_start; c < c_end; c++)
-                        printf("%7d", c);
-                    printf("\n");
-
-                    for (int y = y_start; y < y_end; y++)
-                    {
-                        printf("%5d", y);
-                        for (int c = c_start; c < c_end; c++)
-                            printf("%7d", this->get_element_value({y, x, c}, with_padding));
-                        printf("\n");
-                    }
-                    printf("\n");
-                }
-            }
-            else
-            {
-                // ______x
-                // |
-                // |
-                // y
-                //
-                for (int c = c_start; c < c_end; c++)
-                {
-                    printf("c = %d\n     ", c);
-
-                    for (int x = x_start; x < x_end; x++)
-                        printf("%7d", x);
-                    printf("\n");
-
-                    for (int y = y_start; y < y_end; y++)
-                    {
-                        printf("%5d", y);
-                        for (int x = x_start; x < x_end; x++)
-                            printf("%7d", this->get_element_value({y, x, c}, with_padding));
-                        printf("\n");
-                    }
-                    printf("\n");
-                }
-            }
-
+            std::cout << "\n";
             return;
         }
 
         /**
-         * @brief print all the element of the Tensor.
+         * @brief Get the index of each dims
          * 
-         * @param message to print
-         * @param with_padding one of true or false,
-         *                     - true: the padding element will also be printed
-         *                     - false: the padding element will not be printed
+         * @param element_index the index of the element
+         * @return std::vector<int> the index of each dims
          */
-        void print_all(const char *message, const bool with_padding = false)
-        {
-            int y_end;
-            int x_end;
-            int c_end;
-            if (with_padding)
-            {
-                y_end = this->shape_with_padding[0];
-                x_end = this->shape_with_padding[1];
-                c_end = this->shape_with_padding[2];
-            }
-            else
-            {
-                y_end = this->shape[0];
-                x_end = this->shape[1];
-                c_end = this->shape[2];
-            }
+        std::vector<int> get_axis_index(int element_index);
 
-            printf("\n%s | ", message);
-            this->print_shape();
-
-            for (int y = 0; y < y_end; y++)
-            {
-                for (int x = 0; x < x_end; x++)
-                {
-                    for (int c = 0; c < c_end; c++)
-                        printf("%d ", this->get_element_value({y, x, c}, with_padding));
-                }
-            }
-            printf("\n");
-            return;
-        }
+        /**
+         * @brief Get the index of element 
+         * 
+         * @param axis_index the index of each dims
+         * @return int the index of element 
+         */
+        int get_element_index(const std::vector<int> axis_index);
 
         /**
          * @brief Check the element value with input ground-truth.
@@ -638,35 +386,39 @@ namespace dl
          * @param gt_element ground-truth value of element
          * @param bias permissible error
          * @param info one of true or false
-         *             - true: print shape and result
+         *             - true:  shape and result
          *             - false: do not
+         * @param failed_number maximum number of wrong element that will be printed
+         * 
          * @return 
          *         - true: in permissible error
          *         - false: not 
          */
-        bool check_element(T *gt_element, int bias = 2, bool info = true)
+        bool check_element(T *gt_element, int bias = 2, bool info = true, int failed_number = 0)
         {
+            int count = 0;
             if (info)
                 this->print_shape();
-            int i = 0;
-            for (int y = 0; y < this->shape[0]; y++)
+            int size = this->get_size();
+            for (int i = 0; i < size; i++)
             {
-                for (int x = 0; x < this->shape[1]; x++)
+                if (DL_ABS(this->element[i] - gt_element[i]) > bias)
                 {
-                    for (int c = 0; c < this->shape[2]; c++)
+                    std::vector<int> index = get_axis_index(i);
+                    std::cout << "element[";
+                    for (int j = 0; j < index.size() - 1; j++)
                     {
-                        int a = this->get_element_value({y, x, c});
-                        int b = gt_element[i];
-                        int offset = DL_ABS(a - b);
-                        if (offset > bias)
-                        {
-                            printf("element[%d, %d, %d]: %d v.s. %d\n", y, x, c, a, b);
-                            return false;
-                        }
-                        i++;
+                        std::cout << index[j] << ", ";
                     }
+                    std::cout << index.back() << "]: ";
+                    std::cout << +this->element[i] << " v.s. " << +gt_element[i] << "\n";
+                    count++;
+                    if (count > failed_number)
+                        return false;
                 }
             }
+            if (count)
+                return false;
 
             if (info)
                 printf("PASS\n");
@@ -700,35 +452,44 @@ namespace dl
 
         Tensor<T> &operator=(const Tensor<T> &input)
         {
-            this->size = input.size;
             this->auto_free = input.auto_free;
             this->exponent = input.exponent;
-            this->shape = input.shape;
-            this->padding = input.padding;
-            int size_real_tmp = this->shape_with_padding.size() ? this->shape_with_padding[0] * this->shape_with_padding[1] * this->shape_with_padding[2] : 0;
-            int size_input_real = input.shape_with_padding.size() ? input.shape_with_padding[0] * input.shape_with_padding[1] * input.shape_with_padding[2] : 0;
-            this->shape_with_padding = input.shape_with_padding;
-            if (this->element)
+            int size_real_tmp = this->size;
+            int size_input_real = input.size;
+            this->set_shape(input.shape);
+            if (input.element)
             {
-                if (size_real_tmp != size_input_real)
+                if (this->element)
                 {
-                    tool::free_aligned(this->element);
-                    T *new_element = (T *)tool::calloc_aligned(size_input_real, sizeof(T), 16);
-                    tool::copy_memory(new_element, input.element, size_input_real * sizeof(T));
-                    this->element = new_element;
+                    if (size_real_tmp != size_input_real)
+                    {
+                        tool::free_aligned_prefer(this->element);
+                        T *new_element = (T *)tool::malloc_aligned_prefer(size_input_real, sizeof(T), 16);
+                        tool::copy_memory(new_element, input.element, size_input_real * sizeof(T));
+                        this->element = new_element;
+                    }
+                    else
+                    {
+                        tool::copy_memory(this->element, input.element, size_input_real * sizeof(T));
+                    }
                 }
                 else
                 {
-                    tool::copy_memory(this->element, input.element, size_input_real * sizeof(T));
+                    T *new_element = (T *)tool::malloc_aligned_prefer(size_input_real, sizeof(T), 16);
+                    tool::copy_memory(new_element, input.element, size_input_real * sizeof(T));
+                    this->element = new_element;
                 }
+                return *this;
             }
             else
             {
-                T *new_element = (T *)tool::calloc_aligned(size_input_real, sizeof(T), 16);
-                tool::copy_memory(new_element, input.element, size_input_real * sizeof(T));
-                this->element = new_element;
+                if (this->element)
+                {
+                    tool::free_aligned_prefer(this->element);
+                    this->element = NULL;
+                }
+                return *this;
             }
-            return *this;
         }
     };
 } // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32/include/esp_hw_support/include/esp_sleep.h b/tools/sdk/esp32/include/esp_hw_support/include/esp_sleep.h
index cfdfbc04186..75d2f9726c3 100644
--- a/tools/sdk/esp32/include/esp_hw_support/include/esp_sleep.h
+++ b/tools/sdk/esp32/include/esp_hw_support/include/esp_sleep.h
@@ -44,6 +44,7 @@ typedef enum {
 #if SOC_PM_SUPPORT_CPU_PD
     ESP_PD_DOMAIN_CPU,             //!< CPU core
 #endif
+    ESP_PD_DOMAIN_RTC8M,           //!< Internal 8M oscillator
     ESP_PD_DOMAIN_VDDSDIO,         //!< VDD_SDIO
     ESP_PD_DOMAIN_MAX              //!< Number of domains
 } esp_sleep_pd_domain_t;
diff --git a/tools/sdk/esp32/include/esp_lcd/include/esp_lcd_panel_io.h b/tools/sdk/esp32/include/esp_lcd/include/esp_lcd_panel_io.h
index eebcabf42b4..330f4d9a165 100644
--- a/tools/sdk/esp32/include/esp_lcd/include/esp_lcd_panel_io.h
+++ b/tools/sdk/esp32/include/esp_lcd/include/esp_lcd_panel_io.h
@@ -65,6 +65,22 @@ esp_err_t esp_lcd_panel_io_tx_color(esp_lcd_panel_io_handle_t io, int lcd_cmd, c
  */
 esp_err_t esp_lcd_panel_io_del(esp_lcd_panel_io_handle_t io);
 
+/**
+ * @brief Type of LCD panel IO event data
+ */
+typedef struct {
+} esp_lcd_panel_io_event_data_t;
+
+/**
+ * @brief Declare the prototype of the function that will be invoked when panel IO finishes transferring color data
+ *
+ * @param[in] panel_io LCD panel IO handle, which is created by factory API like `esp_lcd_new_panel_io_spi()`
+ * @param[in] edata Panel IO event data, fed by driver
+ * @param[in] user_ctx User data, passed from `esp_lcd_panel_io_xxx_config_t`
+ * @return Whether a high priority task has been waken up by this function
+ */
+typedef bool (*esp_lcd_panel_io_color_trans_done_cb_t)(esp_lcd_panel_io_handle_t panel_io, esp_lcd_panel_io_event_data_t *edata, void *user_ctx);
+
 /**
  * @brief Panel IO configuration structure, for SPI interface
  */
@@ -74,8 +90,8 @@ typedef struct {
     int spi_mode;    /*!< Traditional SPI mode (0~3) */
     unsigned int pclk_hz;    /*!< Frequency of pixel clock */
     size_t trans_queue_depth; /*!< Size of internal transaction queue */
-    bool (*on_color_trans_done)(esp_lcd_panel_io_handle_t panel_io, void *user_data, void *event_data); /*!< Callback, invoked when color data transfer has finished */
-    void *user_data;    /*!< User private data, passed directly to on_trans_frame_done's user_data */
+    esp_lcd_panel_io_color_trans_done_cb_t on_color_trans_done; /*!< Callback invoked when color data transfer has finished */
+    void *user_ctx;    /*!< User private data, passed directly to on_color_trans_done's user_ctx */
     int lcd_cmd_bits;   /*!< Bit-width of LCD command */
     int lcd_param_bits; /*!< Bit-width of LCD parameter */
     struct {
@@ -100,8 +116,8 @@ esp_err_t esp_lcd_new_panel_io_spi(esp_lcd_spi_bus_handle_t bus, const esp_lcd_p
 
 typedef struct {
     uint32_t dev_addr; /*!< I2C device address */
-    bool (*on_color_trans_done)(esp_lcd_panel_io_handle_t panel_io, void *user_data, void *event_data); /*!< Callback, invoked when color data transfer has finished */
-    void *user_data; /*!< User private data, passed directly to on_trans_frame_done's user_data */
+    esp_lcd_panel_io_color_trans_done_cb_t on_color_trans_done; /*!< Callback invoked when color data transfer has finished */
+    void *user_ctx; /*!< User private data, passed directly to on_color_trans_done's user_ctx */
     size_t control_phase_bytes; /*!< I2C LCD panel will encode control information (e.g. D/C seclection) into control phase, in several bytes */
     unsigned int dc_bit_offset; /*!< Offset of the D/C selection bit in control phase */
     int lcd_cmd_bits;           /*!< Bit-width of LCD command */
@@ -168,8 +184,8 @@ typedef struct {
     int cs_gpio_num;         /*!< GPIO used for CS line, set to -1 will declaim exclusively use of I80 bus */
     unsigned int pclk_hz;    /*!< Frequency of pixel clock */
     size_t trans_queue_depth; /*!< Transaction queue size, larger queue, higher throughput */
-    bool (*on_color_trans_done)(esp_lcd_panel_io_handle_t panel_io, void *user_data, void *event_data); /*!< Callback, invoked when color data was tranferred done */
-    void *user_data;    /*!< User private data, passed directly to on_trans_done's user_data */
+    esp_lcd_panel_io_color_trans_done_cb_t on_color_trans_done; /*!< Callback invoked when color data was tranferred done */
+    void *user_ctx;    /*!< User private data, passed directly to on_color_trans_done's user_ctx */
     int lcd_cmd_bits;   /*!< Bit-width of LCD command */
     int lcd_param_bits; /*!< Bit-width of LCD parameter */
     struct {
diff --git a/tools/sdk/esp32/include/esp_lcd/include/esp_lcd_panel_rgb.h b/tools/sdk/esp32/include/esp_lcd/include/esp_lcd_panel_rgb.h
index 2ddd2b6b9a6..1368bb787f6 100644
--- a/tools/sdk/esp32/include/esp_lcd/include/esp_lcd_panel_rgb.h
+++ b/tools/sdk/esp32/include/esp_lcd/include/esp_lcd_panel_rgb.h
@@ -18,6 +18,37 @@ extern "C" {
 #if SOC_LCD_RGB_SUPPORTED
 /**
  * @brief LCD RGB timing structure
+ *
+ *                                                 Total Width
+ *                             <--------------------------------------------------->
+ *                       Hsync width HBP             Active Width                HFP
+ *                             <---><--><--------------------------------------><--->
+ *                         ____    ____|_______________________________________|____|
+ *                             |___|   |                                       |    |
+ *                                     |                                       |    |
+ *                         __|         |                                       |    |
+ *            /|\    /|\  |            |                                       |    |
+ *             | VSYNC|   |            |                                       |    |
+ *             |Width\|/  |__          |                                       |    |
+ *             |     /|\     |         |                                       |    |
+ *             |  VBP |      |         |                                       |    |
+ *             |     \|/_____|_________|_______________________________________|    |
+ *             |     /|\     |         | / / / / / / / / / / / / / / / / / / / |    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *    Total    |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *    Heigh    |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |Active|      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |Heigh |      |         |/ / / / / / Active Display Area / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |     \|/_____|_________|_______________________________________|    |
+ *             |     /|\     |                                                      |
+ *             |  VFP |      |                                                      |
+ *            \|/    \|/_____|______________________________________________________|
+ *
  */
 typedef struct {
     unsigned int pclk_hz;           /*!< Frequency of pixel clock */
@@ -38,6 +69,22 @@ typedef struct {
     } flags;
 } esp_lcd_rgb_timing_t;
 
+/**
+ * @brief Type of RGB LCD panel event data
+ */
+typedef struct {
+} esp_lcd_rgb_panel_event_data_t;
+
+/**
+ * @brief Declare the prototype of the function that will be invoked when panel IO finishes transferring color data
+ *
+ * @param[in] panel LCD panel handle, returned from `esp_lcd_new_rgb_panel`
+ * @param[in] edata Panel event data, fed by driver
+ * @param[in] user_ctx User data, passed from `esp_lcd_rgb_panel_config_t`
+ * @return Whether a high priority task has been waken up by this function
+ */
+typedef bool (*esp_lcd_rgb_panel_frame_trans_done_cb_t)(esp_lcd_panel_handle_t panel, esp_lcd_rgb_panel_event_data_t *edata, void *user_ctx);
+
 /**
  * @brief LCD RGB panel configuration structure
  */
@@ -51,8 +98,8 @@ typedef struct {
     int pclk_gpio_num;            /*!< GPIO used for PCLK signal */
     int data_gpio_nums[SOC_LCD_RGB_DATA_WIDTH]; /*!< GPIOs used for data lines */
     int disp_gpio_num; /*!< GPIO used for display control signal, set to -1 if it's not used */
-    bool (*on_frame_trans_done)(esp_lcd_panel_handle_t panel, void *user_data); /*!< Callback, invoked when one frame buffer has transferred done */
-    void *user_data; /*!< User data which would be passed to on_frame_trans_done's user_data */
+    esp_lcd_rgb_panel_frame_trans_done_cb_t on_frame_trans_done; /*!< Callback invoked when one frame buffer has transferred done */
+    void *user_ctx; /*!< User data which would be passed to on_frame_trans_done's user_ctx */
     struct {
         unsigned int disp_active_low: 1; /*!< If this flag is enabled, a low level of display control signal can turn the screen on; vice versa */
         unsigned int relax_on_idle: 1;   /*!< If this flag is enabled, the host won't refresh the LCD if nothing changed in host's frame buffer (this is usefull for LCD with built-in GRAM) */
diff --git a/tools/sdk/esp32/include/esp_littlefs/include/esp_littlefs.h b/tools/sdk/esp32/include/esp_littlefs/include/esp_littlefs.h
index dbb1028790d..7b5abe9248f 100644
--- a/tools/sdk/esp32/include/esp_littlefs/include/esp_littlefs.h
+++ b/tools/sdk/esp32/include/esp_littlefs/include/esp_littlefs.h
@@ -14,7 +14,7 @@
 #include <sys/stat.h>
 #include <sys/time.h>
 #include <sys/poll.h>
-#include <dirent.h>
+#include <sys/dirent.h>
 #include <string.h>
 #include "sdkconfig.h"
 
diff --git a/tools/sdk/esp32/include/esp_wifi/include/esp_coexist.h b/tools/sdk/esp32/include/esp_wifi/include/esp_coexist.h
index 1ff624d9b55..14b7c9aa506 100644
--- a/tools/sdk/esp32/include/esp_wifi/include/esp_coexist.h
+++ b/tools/sdk/esp32/include/esp_wifi/include/esp_coexist.h
@@ -1,16 +1,8 @@
-// Copyright 2015-2018 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #ifndef __ESP_COEXIST_H__
 #define __ESP_COEXIST_H__
@@ -32,6 +24,13 @@ typedef enum {
     ESP_COEX_PREFER_NUM,            /*!< Prefer value numbers */
 } esp_coex_prefer_t;
 
+typedef enum {
+    EXTERN_COEX_WIRE_1 = 0,
+    EXTERN_COEX_WIRE_2,
+    EXTERN_COEX_WIRE_3,
+    EXTERN_COEX_WIRE_NUM,
+} external_coex_wire_t;
+
 /**
  * @brief coex status type
  */
@@ -41,6 +40,36 @@ typedef enum {
     ESP_COEX_ST_TYPE_BT,
 } esp_coex_status_type_t;
 
+/**
+ * @brief external coex gpio pti
+ */
+typedef struct {
+    int32_t in_pin0;
+    int32_t in_pin1;
+    int32_t out_pin0;
+} esp_external_coex_gpio_set_t;
+
+/**
+ * @brief external coex pti level
+ */
+typedef enum {
+    EXTERN_COEX_PTI_MID = 0,
+    EXTERN_COEX_PTI_HIGH,
+    EXTERN_COEX_PTI_NUM,
+} esp_coex_pti_level_t;
+
+/**
+ * @brief external coex pti
+ */
+typedef struct {
+    uint32_t in_pti1;
+    uint32_t in_pti2;
+    uint32_t in_pti3;
+    uint32_t out_pti1;
+    uint32_t out_pti2;
+    uint32_t out_pti3;
+} esp_external_coex_pti_set_t;
+
 #define ESP_COEX_BLE_ST_MESH_CONFIG        0x08
 #define ESP_COEX_BLE_ST_MESH_TRAFFIC       0x10
 #define ESP_COEX_BLE_ST_MESH_STANDBY       0x20
@@ -84,6 +113,18 @@ esp_err_t esp_coex_status_bit_set(esp_coex_status_type_t type, uint32_t status);
  */
 esp_err_t esp_coex_status_bit_clear(esp_coex_status_type_t type, uint32_t status);
 
+#if CONFIG_EXTERNAL_COEX_ENABLE
+/**
+ * @brief Setup gpio pin and corresponding pti level, start external coex.
+ * @param wire_type : to select the whole external coex gpio number.
+ * @param gpio_pin : gpio pin number to choose.
+ * @return : ESP_OK - success, other - failed
+ */
+esp_err_t esp_enable_extern_coex_gpio_pin(external_coex_wire_t wire_type,
+                    esp_external_coex_gpio_set_t gpio_pin);
+
+esp_err_t esp_disable_extern_coex_gpio_pin();
+#endif
 
 #ifdef __cplusplus
 }
diff --git a/tools/sdk/esp32/include/esp_wifi/include/esp_coexist_internal.h b/tools/sdk/esp32/include/esp_wifi/include/esp_coexist_internal.h
index 3501f0da6fb..7ba06d4c690 100644
--- a/tools/sdk/esp32/include/esp_wifi/include/esp_coexist_internal.h
+++ b/tools/sdk/esp32/include/esp_wifi/include/esp_coexist_internal.h
@@ -1,21 +1,14 @@
-// Copyright 2018-2018 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2018-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #ifndef __ESP_COEXIST_INTERNAL_H__
 #define __ESP_COEXIST_INTERNAL_H__
 
 #include <stdbool.h>
+#include "esp_coexist.h"
 #include "esp_coexist_adapter.h"
 
 #ifdef __cplusplus
@@ -210,6 +203,29 @@ int coex_schm_curr_phase_idx_get(void);
  */
 esp_err_t esp_coex_adapter_register(coex_adapter_funcs_t *funcs);
 
+#if CONFIG_EXTERNAL_COEX_ENABLE
+/**
+  * @brief     Set external coexistence pti level and enable it.
+  *
+  * @param     level1    external coex low pti
+  * @param     level2    external coex mid pti
+  * @param     level3    external coex high pti
+  *
+  * @return
+  *    - ESP_OK: succeed
+  */
+esp_err_t esp_coex_external_set(esp_coex_pti_level_t level1,
+         esp_coex_pti_level_t level2, esp_coex_pti_level_t level3);
+
+/**
+  * @brief     Disable external coexist
+  *
+  * @return
+  *    - ESP_OK: succeed
+  */
+void esp_coex_external_stop(void);
+#endif    /*External Coex*/
+
 /**
   * @brief     Check the MD5 values of the coexistence adapter header files in IDF and WiFi library
   *
diff --git a/tools/sdk/esp32/include/esp_wifi/include/esp_wifi_types.h b/tools/sdk/esp32/include/esp_wifi/include/esp_wifi_types.h
index 503e8d7bb05..cee23f3fafc 100644
--- a/tools/sdk/esp32/include/esp_wifi/include/esp_wifi_types.h
+++ b/tools/sdk/esp32/include/esp_wifi/include/esp_wifi_types.h
@@ -1,16 +1,8 @@
-// Copyright 2015-2016 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 
 #ifndef __ESP_WIFI_TYPES_H__
@@ -80,6 +72,7 @@ typedef enum {
     WIFI_REASON_ASSOC_NOT_AUTHED         = 9,
     WIFI_REASON_DISASSOC_PWRCAP_BAD      = 10,
     WIFI_REASON_DISASSOC_SUPCHAN_BAD     = 11,
+    WIFI_REASON_BSS_TRANSITION_DISASSOC  = 12,
     WIFI_REASON_IE_INVALID               = 13,
     WIFI_REASON_MIC_FAILURE              = 14,
     WIFI_REASON_4WAY_HANDSHAKE_TIMEOUT   = 15,
@@ -250,7 +243,8 @@ typedef struct {
     wifi_pmf_config_t pmf_cfg;    /**< Configuration for Protected Management Frame. Will be advertized in RSN Capabilities in RSN IE. */
     uint32_t rm_enabled:1;        /**< Whether Radio Measurements are enabled for the connection */
     uint32_t btm_enabled:1;       /**< Whether BSS Transition Management is enabled for the connection */
-    uint32_t reserved:30;         /**< Reserved for future feature set */
+    uint32_t mbo_enabled:1;       /**< Whether MBO is enabled for the connection */
+    uint32_t reserved:29;         /**< Reserved for future feature set */
 } wifi_sta_config_t;
 
 /** @brief Configuration data for ESP32 AP or STA.
diff --git a/tools/sdk/esp32/include/freertos/include/esp_additions/freertos/FreeRTOSConfig.h b/tools/sdk/esp32/include/freertos/include/esp_additions/freertos/FreeRTOSConfig.h
index a01a56e9fd4..675af141ebc 100644
--- a/tools/sdk/esp32/include/freertos/include/esp_additions/freertos/FreeRTOSConfig.h
+++ b/tools/sdk/esp32/include/freertos/include/esp_additions/freertos/FreeRTOSConfig.h
@@ -90,7 +90,6 @@
 #define portNUM_PROCESSORS                              1
 #endif
 
-#define configASSERT_2                                  0
 #define portUSING_MPU_WRAPPERS                          0
 #define configUSE_MUTEX                                 1
 
@@ -206,7 +205,6 @@
 #define configGENERATE_RUN_TIME_STATS                   1       /* Used by vTaskGetRunTimeStats() */
 #endif
 
-#define configUSE_TRACE_FACILITY_2                      0
 #define configBENCHMARK                                 0
 #define configUSE_16_BIT_TICKS                          0
 #define configIDLE_SHOULD_YIELD                         0
@@ -306,4 +304,9 @@ extern void vPortCleanUpTCB ( void *pxTCB );
 
 #define configTASK_NOTIFICATION_ARRAY_ENTRIES           1
 
+// backward compatibility for 4.4
+#define xTaskRemoveFromUnorderedEventList vTaskRemoveFromUnorderedEventList
+
+#define configNUM_CORES                                 portNUM_PROCESSORS
+
 #endif /* FREERTOS_CONFIG_H */
diff --git a/tools/sdk/esp32/include/freertos/include/freertos/event_groups.h b/tools/sdk/esp32/include/freertos/include/freertos/event_groups.h
index 84505ddaaa0..9792296e566 100644
--- a/tools/sdk/esp32/include/freertos/include/freertos/event_groups.h
+++ b/tools/sdk/esp32/include/freertos/include/freertos/event_groups.h
@@ -64,7 +64,7 @@
  * used to create a synchronisation point between multiple tasks (a
  * 'rendezvous').
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup EventGroup EventGroup
  * @endcond
  */
@@ -78,7 +78,7 @@
  * xEventGroupCreate() returns an EventGroupHandle_t variable that can then
  * be used as a parameter to other event group functions.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup EventGroupHandle_t EventGroupHandle_t
  * @endcond
  * \ingroup EventGroup
@@ -94,7 +94,7 @@ typedef struct EventGroupDef_t * EventGroupHandle_t;
  * number of bits it holds is set by configUSE_16_BIT_TICKS (16 bits if set to 1,
  * 32 bits if set to 0.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup EventBits_t EventBits_t
  * @endcond
  * \ingroup EventGroup
@@ -102,7 +102,7 @@ typedef struct EventGroupDef_t * EventGroupHandle_t;
 typedef TickType_t               EventBits_t;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  * EventGroupHandle_t xEventGroupCreate( void );
@@ -152,7 +152,7 @@ typedef TickType_t               EventBits_t;
  *      // The event group was created.
  *  }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupCreate xEventGroupCreate
  * @endcond
  * \ingroup EventGroup
@@ -162,7 +162,7 @@ typedef TickType_t               EventBits_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  * EventGroupHandle_t xEventGroupCreateStatic( EventGroupHandle_t * pxEventGroupBuffer );
@@ -217,7 +217,7 @@ typedef TickType_t               EventBits_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupWaitBits(    EventGroupHandle_t xEventGroup,
@@ -307,7 +307,7 @@ typedef TickType_t               EventBits_t;
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupWaitBits xEventGroupWaitBits
  * @endcond
  * \ingroup EventGroup
@@ -319,7 +319,7 @@ EventBits_t xEventGroupWaitBits( EventGroupHandle_t xEventGroup,
                                  TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup, const EventBits_t uxBitsToClear );
@@ -372,7 +372,7 @@ EventBits_t xEventGroupWaitBits( EventGroupHandle_t xEventGroup,
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupClearBits xEventGroupClearBits
  * @endcond
  * \ingroup EventGroup
@@ -381,7 +381,7 @@ EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup,
                                   const EventBits_t uxBitsToClear ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  BaseType_t xEventGroupClearBitsFromISR( EventGroupHandle_t xEventGroup, const EventBits_t uxBitsToSet );
@@ -432,7 +432,7 @@ EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup,
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupClearBitsFromISR xEventGroupClearBitsFromISR
  * @endcond
  * \ingroup EventGroup
@@ -446,7 +446,7 @@ EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup,
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup, const EventBits_t uxBitsToSet );
@@ -516,7 +516,7 @@ EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup,
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupSetBits xEventGroupSetBits
  * @endcond
  * \ingroup EventGroup
@@ -525,7 +525,7 @@ EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup,
                                 const EventBits_t uxBitsToSet ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  BaseType_t xEventGroupSetBitsFromISR( EventGroupHandle_t xEventGroup, const EventBits_t uxBitsToSet, BaseType_t *pxHigherPriorityTaskWoken );
@@ -595,7 +595,7 @@ EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup,
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupSetBitsFromISR xEventGroupSetBitsFromISR
  * @endcond
  * \ingroup EventGroup
@@ -610,7 +610,7 @@ EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup,
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupSync(    EventGroupHandle_t xEventGroup,
@@ -732,7 +732,7 @@ EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup,
  * }
  *
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupSync xEventGroupSync
  * @endcond
  * \ingroup EventGroup
@@ -744,7 +744,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupGetBits( EventGroupHandle_t xEventGroup );
@@ -758,7 +758,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
  *
  * @return The event group bits at the time xEventGroupGetBits() was called.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupGetBits xEventGroupGetBits
  * @endcond
  * \ingroup EventGroup
@@ -766,7 +766,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
 #define xEventGroupGetBits( xEventGroup )    xEventGroupClearBits( xEventGroup, 0 )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupGetBitsFromISR( EventGroupHandle_t xEventGroup );
@@ -779,7 +779,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
  *
  * @return The event group bits at the time xEventGroupGetBitsFromISR() was called.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupGetBitsFromISR xEventGroupGetBitsFromISR
  * @endcond
  * \ingroup EventGroup
@@ -787,7 +787,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
 EventBits_t xEventGroupGetBitsFromISR( EventGroupHandle_t xEventGroup ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  void xEventGroupDelete( EventGroupHandle_t xEventGroup );
@@ -802,7 +802,7 @@ EventBits_t xEventGroupGetBitsFromISR( EventGroupHandle_t xEventGroup ) PRIVILEG
  */
 void vEventGroupDelete( EventGroupHandle_t xEventGroup ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 
 /* For internal use only. */
 void vEventGroupSetBitsCallback( void * pvEventGroup,
diff --git a/tools/sdk/esp32/include/freertos/include/freertos/message_buffer.h b/tools/sdk/esp32/include/freertos/include/freertos/message_buffer.h
index e57c589fbac..af5c3290b7c 100644
--- a/tools/sdk/esp32/include/freertos/include/freertos/message_buffer.h
+++ b/tools/sdk/esp32/include/freertos/include/freertos/message_buffer.h
@@ -85,7 +85,7 @@ typedef void * MessageBufferHandle_t;
 /*-----------------------------------------------------------*/
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -139,7 +139,7 @@ typedef void * MessageBufferHandle_t;
  *  }
  *
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferCreate xMessageBufferCreate
  * @endcond
  * \ingroup MessageBufferManagement
@@ -148,7 +148,7 @@ typedef void * MessageBufferHandle_t;
     ( MessageBufferHandle_t ) xStreamBufferGenericCreate( xBufferSizeBytes, ( size_t ) 0, pdTRUE )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -210,7 +210,7 @@ typedef void * MessageBufferHandle_t;
  * }
  *
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferCreateStatic xMessageBufferCreateStatic
  * @endcond
  * \ingroup MessageBufferManagement
@@ -219,7 +219,7 @@ typedef void * MessageBufferHandle_t;
     ( MessageBufferHandle_t ) xStreamBufferGenericCreateStatic( xBufferSizeBytes, 0, pdTRUE, pucMessageBufferStorageArea, pxStaticMessageBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -314,7 +314,7 @@ typedef void * MessageBufferHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferSend xMessageBufferSend
  * @endcond
  * \ingroup MessageBufferManagement
@@ -323,7 +323,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferSend( ( StreamBufferHandle_t ) xMessageBuffer, pvTxData, xDataLengthBytes, xTicksToWait )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -423,7 +423,7 @@ typedef void * MessageBufferHandle_t;
  *  portYIELD_FROM_ISR( xHigherPriorityTaskWoken );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferSendFromISR xMessageBufferSendFromISR
  * @endcond
  * \ingroup MessageBufferManagement
@@ -432,7 +432,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferSendFromISR( ( StreamBufferHandle_t ) xMessageBuffer, pvTxData, xDataLengthBytes, pxHigherPriorityTaskWoken )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -516,7 +516,7 @@ typedef void * MessageBufferHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferReceive xMessageBufferReceive
  * @endcond
  * \ingroup MessageBufferManagement
@@ -526,7 +526,7 @@ typedef void * MessageBufferHandle_t;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -622,7 +622,7 @@ typedef void * MessageBufferHandle_t;
  *  portYIELD_FROM_ISR( xHigherPriorityTaskWoken );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferReceiveFromISR xMessageBufferReceiveFromISR
  * @endcond
  * \ingroup MessageBufferManagement
@@ -631,7 +631,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferReceiveFromISR( ( StreamBufferHandle_t ) xMessageBuffer, pvRxData, xBufferLengthBytes, pxHigherPriorityTaskWoken )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -654,7 +654,7 @@ typedef void * MessageBufferHandle_t;
     vStreamBufferDelete( ( StreamBufferHandle_t ) xMessageBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * BaseType_t xMessageBufferIsFull( MessageBufferHandle_t xMessageBuffer ) );
@@ -674,7 +674,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferIsFull( ( StreamBufferHandle_t ) xMessageBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * BaseType_t xMessageBufferIsEmpty( MessageBufferHandle_t xMessageBuffer ) );
@@ -693,7 +693,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferIsEmpty( ( StreamBufferHandle_t ) xMessageBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * BaseType_t xMessageBufferReset( MessageBufferHandle_t xMessageBuffer );
@@ -712,7 +712,7 @@ typedef void * MessageBufferHandle_t;
  * the message queue to wait for space to become available, or to wait for a
  * a message to be available, then pdFAIL is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferReset xMessageBufferReset
  * @endcond
  * \ingroup MessageBufferManagement
@@ -722,7 +722,7 @@ typedef void * MessageBufferHandle_t;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * size_t xMessageBufferSpaceAvailable( MessageBufferHandle_t xMessageBuffer ) );
@@ -740,7 +740,7 @@ typedef void * MessageBufferHandle_t;
  * architecture, so if xMessageBufferSpacesAvailable() returns 10, then the size
  * of the largest message that can be written to the message buffer is 6 bytes.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferSpaceAvailable xMessageBufferSpaceAvailable
  * @endcond
  * \ingroup MessageBufferManagement
@@ -751,7 +751,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferSpacesAvailable( ( StreamBufferHandle_t ) xMessageBuffer ) /* Corrects typo in original macro name. */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * size_t xMessageBufferNextLengthBytes( MessageBufferHandle_t xMessageBuffer ) );
@@ -767,7 +767,7 @@ typedef void * MessageBufferHandle_t;
  * @return The length (in bytes) of the next message in the message buffer, or 0
  * if the message buffer is empty.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferNextLengthBytes xMessageBufferNextLengthBytes
  * @endcond
  * \ingroup MessageBufferManagement
@@ -776,7 +776,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferNextMessageLengthBytes( ( StreamBufferHandle_t ) xMessageBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -811,7 +811,7 @@ typedef void * MessageBufferHandle_t;
  * @return If a task was removed from the Blocked state then pdTRUE is returned.
  * Otherwise pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferSendCompletedFromISR xMessageBufferSendCompletedFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -820,7 +820,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferSendCompletedFromISR( ( StreamBufferHandle_t ) xMessageBuffer, pxHigherPriorityTaskWoken )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -856,7 +856,7 @@ typedef void * MessageBufferHandle_t;
  * @return If a task was removed from the Blocked state then pdTRUE is returned.
  * Otherwise pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferReceiveCompletedFromISR xMessageBufferReceiveCompletedFromISR
  * @endcond
  * \ingroup StreamBufferManagement
diff --git a/tools/sdk/esp32/include/freertos/include/freertos/queue.h b/tools/sdk/esp32/include/freertos/include/freertos/queue.h
index 81cccc05df3..05ca7de4546 100644
--- a/tools/sdk/esp32/include/freertos/include/freertos/queue.h
+++ b/tools/sdk/esp32/include/freertos/include/freertos/queue.h
@@ -62,7 +62,7 @@ typedef struct QueueDefinition   * QueueSetHandle_t;
  */
 typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 
 /* For internal use only. */
 #define queueSEND_TO_BACK                     ( ( BaseType_t ) 0 )
@@ -80,7 +80,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 /** @endcond */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * QueueHandle_t xQueueCreate(
@@ -146,7 +146,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueCreate xQueueCreate
  * @endcond
  * \ingroup QueueManagement
@@ -156,7 +156,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * QueueHandle_t xQueueCreateStatic(
@@ -235,7 +235,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueCreateStatic xQueueCreateStatic
  * @endcond
  * \ingroup QueueManagement
@@ -245,7 +245,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendToToFront(
@@ -321,7 +321,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSend xQueueSend
  * @endcond
  * \ingroup QueueManagement
@@ -330,7 +330,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
     xQueueGenericSend( ( xQueue ), ( pvItemToQueue ), ( xTicksToWait ), queueSEND_TO_FRONT )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendToBack(
@@ -408,7 +408,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSend xQueueSend
  * @endcond
  * \ingroup QueueManagement
@@ -417,7 +417,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
     xQueueGenericSend( ( xQueue ), ( pvItemToQueue ), ( xTicksToWait ), queueSEND_TO_BACK )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSend(
@@ -497,7 +497,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSend xQueueSend
  * @endcond
  * \ingroup QueueManagement
@@ -506,7 +506,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
     xQueueGenericSend( ( xQueue ), ( pvItemToQueue ), ( xTicksToWait ), queueSEND_TO_BACK )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueOverwrite(
@@ -585,7 +585,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ...
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueOverwrite xQueueOverwrite
  * @endcond
  * \ingroup QueueManagement
@@ -595,7 +595,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueGenericSend(
@@ -678,7 +678,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSend xQueueSend
  * @endcond
  * \ingroup QueueManagement
@@ -689,7 +689,7 @@ BaseType_t xQueueGenericSend( QueueHandle_t xQueue,
                               const BaseType_t xCopyPosition ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueuePeek(
@@ -780,7 +780,7 @@ BaseType_t xQueueGenericSend( QueueHandle_t xQueue,
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueuePeek xQueuePeek
  * @endcond
  * \ingroup QueueManagement
@@ -790,7 +790,7 @@ BaseType_t xQueuePeek( QueueHandle_t xQueue,
                        TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueuePeekFromISR(
@@ -820,7 +820,7 @@ BaseType_t xQueuePeek( QueueHandle_t xQueue,
  * @return pdTRUE if an item was successfully received from the queue,
  * otherwise pdFALSE.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueuePeekFromISR xQueuePeekFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -829,7 +829,7 @@ BaseType_t xQueuePeekFromISR( QueueHandle_t xQueue,
                               void * const pvBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueReceive(
@@ -917,7 +917,7 @@ BaseType_t xQueuePeekFromISR( QueueHandle_t xQueue,
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueReceive xQueueReceive
  * @endcond
  * \ingroup QueueManagement
@@ -927,7 +927,7 @@ BaseType_t xQueueReceive( QueueHandle_t xQueue,
                           TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * UBaseType_t uxQueueMessagesWaiting( const QueueHandle_t xQueue );
@@ -940,7 +940,7 @@ BaseType_t xQueueReceive( QueueHandle_t xQueue,
  *
  * @return The number of messages available in the queue.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup uxQueueMessagesWaiting uxQueueMessagesWaiting
  * @endcond
  * \ingroup QueueManagement
@@ -948,7 +948,7 @@ BaseType_t xQueueReceive( QueueHandle_t xQueue,
 UBaseType_t uxQueueMessagesWaiting( const QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * UBaseType_t uxQueueSpacesAvailable( const QueueHandle_t xQueue );
@@ -963,7 +963,7 @@ UBaseType_t uxQueueMessagesWaiting( const QueueHandle_t xQueue ) PRIVILEGED_FUNC
  *
  * @return The number of spaces available in the queue.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup uxQueueMessagesWaiting uxQueueMessagesWaiting
  * @endcond
  * \ingroup QueueManagement
@@ -971,7 +971,7 @@ UBaseType_t uxQueueMessagesWaiting( const QueueHandle_t xQueue ) PRIVILEGED_FUNC
 UBaseType_t uxQueueSpacesAvailable( const QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * void vQueueDelete( QueueHandle_t xQueue );
@@ -983,7 +983,7 @@ UBaseType_t uxQueueSpacesAvailable( const QueueHandle_t xQueue ) PRIVILEGED_FUNC
  *
  * @param xQueue A handle to the queue to be deleted.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vQueueDelete vQueueDelete
  * @endcond
  * \ingroup QueueManagement
@@ -991,7 +991,7 @@ UBaseType_t uxQueueSpacesAvailable( const QueueHandle_t xQueue ) PRIVILEGED_FUNC
 void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendToFrontFromISR(
@@ -1057,7 +1057,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSendFromISR xQueueSendFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1067,7 +1067,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendToBackFromISR(
@@ -1133,7 +1133,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSendFromISR xQueueSendFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1142,7 +1142,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
     xQueueGenericSendFromISR( ( xQueue ), ( pvItemToQueue ), ( pxHigherPriorityTaskWoken ), queueSEND_TO_BACK )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueOverwriteFromISR(
@@ -1225,7 +1225,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueOverwriteFromISR xQueueOverwriteFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1234,7 +1234,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
     xQueueGenericSendFromISR( ( xQueue ), ( pvItemToQueue ), ( pxHigherPriorityTaskWoken ), queueOVERWRITE )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendFromISR(
@@ -1304,7 +1304,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSendFromISR xQueueSendFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1312,10 +1312,10 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 #define xQueueSendFromISR( xQueue, pvItemToQueue, pxHigherPriorityTaskWoken ) \
     xQueueGenericSendFromISR( ( xQueue ), ( pvItemToQueue ), ( pxHigherPriorityTaskWoken ), queueSEND_TO_BACK )
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /**@{*/
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueGenericSendFromISR(
@@ -1402,7 +1402,7 @@ BaseType_t xQueueGiveFromISR( QueueHandle_t xQueue,
 /** @endcond */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueReceiveFromISR(
@@ -1487,7 +1487,7 @@ BaseType_t xQueueGiveFromISR( QueueHandle_t xQueue,
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueReceiveFromISR xQueueReceiveFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1504,7 +1504,7 @@ BaseType_t xQueueIsQueueEmptyFromISR( const QueueHandle_t xQueue ) PRIVILEGED_FU
 BaseType_t xQueueIsQueueFullFromISR( const QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 UBaseType_t uxQueueMessagesWaitingFromISR( const QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /*
  * The functions defined above are for passing data to and from tasks.  The
  * functions below are the equivalents for passing data to and from
@@ -1778,7 +1778,7 @@ QueueSetMemberHandle_t xQueueSelectFromSet( QueueSetHandle_t xQueueSet,
  */
 QueueSetMemberHandle_t xQueueSelectFromSetFromISR( QueueSetHandle_t xQueueSet ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 
 /* Not public API functions. */
 void vQueueWaitForMessageRestricted( QueueHandle_t xQueue,
diff --git a/tools/sdk/esp32/include/freertos/include/freertos/semphr.h b/tools/sdk/esp32/include/freertos/include/freertos/semphr.h
index 7e99c0b396c..2041641b91d 100644
--- a/tools/sdk/esp32/include/freertos/include/freertos/semphr.h
+++ b/tools/sdk/esp32/include/freertos/include/freertos/semphr.h
@@ -39,7 +39,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define semSEMAPHORE_QUEUE_ITEM_LENGTH      ( ( uint8_t ) 0U )
 #define semGIVE_BLOCK_TIME                  ( ( TickType_t ) 0U )
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /**
  * semphr. h
  * @code{c}
@@ -88,7 +88,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vSemaphoreCreateBinary vSemaphoreCreateBinary
  * @endcond
  * \ingroup Semaphores
@@ -106,7 +106,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 /** @endcond */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateBinary( void );
@@ -163,7 +163,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateBinary xSemaphoreCreateBinary
  * @endcond
  * \ingroup Semaphores
@@ -173,7 +173,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateBinaryStatic( StaticSemaphore_t *pxSemaphoreBuffer );
@@ -229,7 +229,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  // Rest of task code goes here.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateBinaryStatic xSemaphoreCreateBinaryStatic
  * @endcond
  * \ingroup Semaphores
@@ -239,7 +239,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * xSemaphoreTake(
@@ -304,7 +304,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreTake xSemaphoreTake
  * @endcond
  * \ingroup Semaphores
@@ -312,7 +312,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreTake( xSemaphore, xBlockTime )    xQueueSemaphoreTake( ( xSemaphore ), ( xBlockTime ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * xSemaphoreTakeRecursive(
@@ -403,7 +403,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreTakeRecursive xSemaphoreTakeRecursive
  * @endcond
  * \ingroup Semaphores
@@ -465,7 +465,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreGive xSemaphoreGive
  * @endcond
  * \ingroup Semaphores
@@ -473,7 +473,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreGive( xSemaphore )    xQueueGenericSend( ( QueueHandle_t ) ( xSemaphore ), NULL, semGIVE_BLOCK_TIME, queueSEND_TO_BACK )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * xSemaphoreGiveRecursive( SemaphoreHandle_t xMutex );
@@ -555,7 +555,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreGiveRecursive xSemaphoreGiveRecursive
  * @endcond
  * \ingroup Semaphores
@@ -641,7 +641,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreGiveFromISR xSemaphoreGiveFromISR
  * @endcond
  * \ingroup Semaphores
@@ -649,7 +649,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreGiveFromISR( xSemaphore, pxHigherPriorityTaskWoken )    xQueueGiveFromISR( ( QueueHandle_t ) ( xSemaphore ), ( pxHigherPriorityTaskWoken ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * xSemaphoreTakeFromISR(
@@ -686,7 +686,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreTakeFromISR( xSemaphore, pxHigherPriorityTaskWoken )    xQueueReceiveFromISR( ( QueueHandle_t ) ( xSemaphore ), NULL, ( pxHigherPriorityTaskWoken ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateMutex( void );
@@ -741,7 +741,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateMutex xSemaphoreCreateMutex
  * @endcond
  * \ingroup Semaphores
@@ -751,7 +751,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateMutexStatic( StaticSemaphore_t *pxMutexBuffer );
@@ -808,7 +808,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  // so there is no need to check it.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateMutexStatic xSemaphoreCreateMutexStatic
  * @endcond
  * \ingroup Semaphores
@@ -951,7 +951,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateCounting( UBaseType_t uxMaxCount, UBaseType_t uxInitialCount );
@@ -1027,7 +1027,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateCounting xSemaphoreCreateCounting
  * @endcond
  * \ingroup Semaphores
@@ -1037,7 +1037,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateCountingStatic( UBaseType_t uxMaxCount, UBaseType_t uxInitialCount, StaticSemaphore_t *pxSemaphoreBuffer );
@@ -1118,7 +1118,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  // is no need to check its value.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateCountingStatic xSemaphoreCreateCountingStatic
  * @endcond
  * \ingroup Semaphores
@@ -1128,7 +1128,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * void vSemaphoreDelete( SemaphoreHandle_t xSemaphore );
@@ -1140,7 +1140,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *
  * @param xSemaphore A handle to the semaphore to be deleted.
  *
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * \defgroup vSemaphoreDelete vSemaphoreDelete
  * @endcond
  * \ingroup Semaphores
@@ -1148,7 +1148,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define vSemaphoreDelete( xSemaphore )                   vQueueDelete( ( QueueHandle_t ) ( xSemaphore ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr.h
  * @code{c}
  * TaskHandle_t xSemaphoreGetMutexHolder( SemaphoreHandle_t xMutex );
@@ -1167,7 +1167,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreGetMutexHolder( xSemaphore )           xQueueGetMutexHolder( ( xSemaphore ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr.h
  * @code{c}
  * TaskHandle_t xSemaphoreGetMutexHolderFromISR( SemaphoreHandle_t xMutex );
@@ -1182,7 +1182,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreGetMutexHolderFromISR( xSemaphore )    xQueueGetMutexHolderFromISR( ( xSemaphore ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr.h
  * @code{c}
  * UBaseType_t uxSemaphoreGetCount( SemaphoreHandle_t xSemaphore );
diff --git a/tools/sdk/esp32/include/freertos/include/freertos/stream_buffer.h b/tools/sdk/esp32/include/freertos/include/freertos/stream_buffer.h
index 9e58cff120c..a20dcf0375e 100644
--- a/tools/sdk/esp32/include/freertos/include/freertos/stream_buffer.h
+++ b/tools/sdk/esp32/include/freertos/include/freertos/stream_buffer.h
@@ -71,7 +71,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -134,7 +134,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferCreate xStreamBufferCreate
  * @endcond
  * \ingroup StreamBufferManagement
@@ -142,7 +142,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
 #define xStreamBufferCreate( xBufferSizeBytes, xTriggerLevelBytes )    xStreamBufferGenericCreate( xBufferSizeBytes, xTriggerLevelBytes, pdFALSE )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -220,7 +220,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
  * }
  *
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferCreateStatic xStreamBufferCreateStatic
  * @endcond
  * \ingroup StreamBufferManagement
@@ -229,7 +229,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
     xStreamBufferGenericCreateStatic( xBufferSizeBytes, xTriggerLevelBytes, pdFALSE, pucStreamBufferStorageArea, pxStaticStreamBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -319,7 +319,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSend xStreamBufferSend
  * @endcond
  * \ingroup StreamBufferManagement
@@ -330,7 +330,7 @@ size_t xStreamBufferSend( StreamBufferHandle_t xStreamBuffer,
                           TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -424,7 +424,7 @@ size_t xStreamBufferSend( StreamBufferHandle_t xStreamBuffer,
  *  taskYIELD_FROM_ISR( xHigherPriorityTaskWoken );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSendFromISR xStreamBufferSendFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -435,7 +435,7 @@ size_t xStreamBufferSendFromISR( StreamBufferHandle_t xStreamBuffer,
                                  BaseType_t * const pxHigherPriorityTaskWoken ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -517,7 +517,7 @@ size_t xStreamBufferSendFromISR( StreamBufferHandle_t xStreamBuffer,
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferReceive xStreamBufferReceive
  * @endcond
  * \ingroup StreamBufferManagement
@@ -528,7 +528,7 @@ size_t xStreamBufferReceive( StreamBufferHandle_t xStreamBuffer,
                              TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -607,7 +607,7 @@ size_t xStreamBufferReceive( StreamBufferHandle_t xStreamBuffer,
  *  taskYIELD_FROM_ISR( xHigherPriorityTaskWoken );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferReceiveFromISR xStreamBufferReceiveFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -618,7 +618,7 @@ size_t xStreamBufferReceiveFromISR( StreamBufferHandle_t xStreamBuffer,
                                     BaseType_t * const pxHigherPriorityTaskWoken ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -636,7 +636,7 @@ size_t xStreamBufferReceiveFromISR( StreamBufferHandle_t xStreamBuffer,
  *
  * @param xStreamBuffer The handle of the stream buffer to be deleted.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vStreamBufferDelete vStreamBufferDelete
  * @endcond
  * \ingroup StreamBufferManagement
@@ -644,7 +644,7 @@ size_t xStreamBufferReceiveFromISR( StreamBufferHandle_t xStreamBuffer,
 void vStreamBufferDelete( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -660,7 +660,7 @@ void vStreamBufferDelete( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTI
  * @return If the stream buffer is full then pdTRUE is returned.  Otherwise
  * pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferIsFull xStreamBufferIsFull
  * @endcond
  * \ingroup StreamBufferManagement
@@ -668,7 +668,7 @@ void vStreamBufferDelete( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTI
 BaseType_t xStreamBufferIsFull( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -684,7 +684,7 @@ BaseType_t xStreamBufferIsFull( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_
  * @return If the stream buffer is empty then pdTRUE is returned.  Otherwise
  * pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferIsEmpty xStreamBufferIsEmpty
  * @endcond
  * \ingroup StreamBufferManagement
@@ -692,7 +692,7 @@ BaseType_t xStreamBufferIsFull( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_
 BaseType_t xStreamBufferIsEmpty( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -711,7 +711,7 @@ BaseType_t xStreamBufferIsEmpty( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED
  * a task blocked waiting to send to or read from the stream buffer then the
  * stream buffer is not reset and pdFAIL is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferReset xStreamBufferReset
  * @endcond
  * \ingroup StreamBufferManagement
@@ -719,7 +719,7 @@ BaseType_t xStreamBufferIsEmpty( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED
 BaseType_t xStreamBufferReset( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -736,7 +736,7 @@ BaseType_t xStreamBufferReset( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_F
  * @return The number of bytes that can be written to the stream buffer before
  * the stream buffer would be full.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSpacesAvailable xStreamBufferSpacesAvailable
  * @endcond
  * \ingroup StreamBufferManagement
@@ -744,7 +744,7 @@ BaseType_t xStreamBufferReset( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_F
 size_t xStreamBufferSpacesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -761,7 +761,7 @@ size_t xStreamBufferSpacesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVIL
  * @return The number of bytes that can be read from the stream buffer before
  * the stream buffer would be empty.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferBytesAvailable xStreamBufferBytesAvailable
  * @endcond
  * \ingroup StreamBufferManagement
@@ -769,7 +769,7 @@ size_t xStreamBufferSpacesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVIL
 size_t xStreamBufferBytesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -802,7 +802,7 @@ size_t xStreamBufferBytesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVILE
  * then the trigger level will be updated and pdTRUE is returned.  Otherwise
  * pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSetTriggerLevel xStreamBufferSetTriggerLevel
  * @endcond
  * \ingroup StreamBufferManagement
@@ -811,7 +811,7 @@ BaseType_t xStreamBufferSetTriggerLevel( StreamBufferHandle_t xStreamBuffer,
                                          size_t xTriggerLevel ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -846,7 +846,7 @@ BaseType_t xStreamBufferSetTriggerLevel( StreamBufferHandle_t xStreamBuffer,
  * @return If a task was removed from the Blocked state then pdTRUE is returned.
  * Otherwise pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSendCompletedFromISR xStreamBufferSendCompletedFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -855,7 +855,7 @@ BaseType_t xStreamBufferSendCompletedFromISR( StreamBufferHandle_t xStreamBuffer
                                               BaseType_t * pxHigherPriorityTaskWoken ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -891,7 +891,7 @@ BaseType_t xStreamBufferSendCompletedFromISR( StreamBufferHandle_t xStreamBuffer
  * @return If a task was removed from the Blocked state then pdTRUE is returned.
  * Otherwise pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferReceiveCompletedFromISR xStreamBufferReceiveCompletedFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -899,7 +899,7 @@ BaseType_t xStreamBufferSendCompletedFromISR( StreamBufferHandle_t xStreamBuffer
 BaseType_t xStreamBufferReceiveCompletedFromISR( StreamBufferHandle_t xStreamBuffer,
                                                  BaseType_t * pxHigherPriorityTaskWoken ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /* Functions below here are not part of the public API. */
 StreamBufferHandle_t xStreamBufferGenericCreate( size_t xBufferSizeBytes,
                                                  size_t xTriggerLevelBytes,
diff --git a/tools/sdk/esp32/include/freertos/include/freertos/task.h b/tools/sdk/esp32/include/freertos/include/freertos/task.h
index 9135b76f014..125a924d061 100644
--- a/tools/sdk/esp32/include/freertos/include/freertos/task.h
+++ b/tools/sdk/esp32/include/freertos/include/freertos/task.h
@@ -76,7 +76,7 @@
  * returns (via a pointer parameter) an TaskHandle_t variable that can then
  * be used as a parameter to vTaskDelete to delete the task.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup TaskHandle_t TaskHandle_t
  * @endcond
  * \ingroup Tasks
@@ -114,7 +114,7 @@ typedef enum
     eSetValueWithoutOverwrite     /* Set the task's notification value if the previous value has been read by the task. */
 } eNotifyAction;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /**
  * Used internally only.
  */
@@ -189,11 +189,13 @@ typedef enum
 #define tskIDLE_PRIORITY    ( ( UBaseType_t ) 0U )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro for forcing a context switch.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskYIELD taskYIELD
  * @endcond
  * \ingroup SchedulerControl
@@ -201,7 +203,9 @@ typedef enum
 #define taskYIELD()                        portYIELD()
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro to mark the start of a critical code region.  Preemptive context
  * switches cannot occur when in a critical region.
@@ -209,7 +213,7 @@ typedef enum
  * @note This may alter the stack (depending on the portable implementation)
  * so must be used with care!
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskENTER_CRITICAL taskENTER_CRITICAL
  * @endcond
  * \ingroup SchedulerControl
@@ -228,7 +232,9 @@ typedef enum
 #endif //  ESP_PLATFORM
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro to mark the end of a critical code region.  Preemptive context
  * switches cannot occur when in a critical region.
@@ -236,7 +242,7 @@ typedef enum
  * @note This may alter the stack (depending on the portable implementation)
  * so must be used with care!
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskEXIT_CRITICAL taskEXIT_CRITICAL
  * @endcond
  * \ingroup SchedulerControl
@@ -255,11 +261,13 @@ typedef enum
 #define taskEXIT_CRITICAL_ISR( )        portEXIT_CRITICAL_ISR( )
 #endif // ESP_PLATFORM
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro to disable all maskable interrupts.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskDISABLE_INTERRUPTS taskDISABLE_INTERRUPTS
  * @endcond
  * \ingroup SchedulerControl
@@ -267,11 +275,13 @@ typedef enum
 #define taskDISABLE_INTERRUPTS()           portDISABLE_INTERRUPTS()
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro to enable microcontroller interrupts.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskENABLE_INTERRUPTS taskENABLE_INTERRUPTS
  * @endcond
  * \ingroup SchedulerControl
@@ -422,7 +432,7 @@ typedef enum
  *   }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCreate xTaskCreate
  * @endcond
  * \ingroup Tasks
@@ -430,14 +440,14 @@ typedef enum
 #if ( configSUPPORT_DYNAMIC_ALLOCATION == 1 )
 
     static inline IRAM_ATTR BaseType_t xTaskCreate(
-            TaskFunction_t pvTaskCode,
-            const char * const pcName,
-            const uint32_t usStackDepth,
-            void * const pvParameters,
-            UBaseType_t uxPriority,
-            TaskHandle_t * const pvCreatedTask)
+                            TaskFunction_t pvTaskCode,
+                            const char * const pcName,     /*lint !e971 Unqualified char types are allowed for strings and single characters only. */
+                            const uint32_t usStackDepth,
+                            void * const pvParameters,
+                            UBaseType_t uxPriority,
+                            TaskHandle_t * const pxCreatedTask) PRIVILEGED_FUNCTION
     {
-        return xTaskCreatePinnedToCore( pvTaskCode, pcName, usStackDepth, pvParameters, uxPriority, pvCreatedTask, tskNO_AFFINITY );
+        return xTaskCreatePinnedToCore( pvTaskCode, pcName, usStackDepth, pvParameters, uxPriority, pxCreatedTask, tskNO_AFFINITY );
     }
 
 #endif
@@ -599,20 +609,20 @@ typedef enum
 
 #if( configSUPPORT_STATIC_ALLOCATION == 1 )
     static inline IRAM_ATTR TaskHandle_t xTaskCreateStatic(
-            TaskFunction_t pvTaskCode,
-            const char * const pcName,
-            const uint32_t ulStackDepth,
-            void * const pvParameters,
-            UBaseType_t uxPriority,
-            StackType_t * const pxStackBuffer,
-            StaticTask_t * const pxTaskBuffer)
+                                    TaskFunction_t pvTaskCode,
+                                    const char * const pcName,     /*lint !e971 Unqualified char types are allowed for strings and single characters only. */
+                                    const uint32_t ulStackDepth,
+                                    void * const pvParameters,
+                                    UBaseType_t uxPriority,
+                                    StackType_t * const puxStackBuffer,
+                                    StaticTask_t * const pxTaskBuffer) PRIVILEGED_FUNCTION
     {
-        return xTaskCreateStaticPinnedToCore( pvTaskCode, pcName, ulStackDepth, pvParameters, uxPriority, pxStackBuffer, pxTaskBuffer, tskNO_AFFINITY );
+        return xTaskCreateStaticPinnedToCore( pvTaskCode, pcName, ulStackDepth, pvParameters, uxPriority, puxStackBuffer, pxTaskBuffer, tskNO_AFFINITY );
     }
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskCreateRestricted( TaskParameters_t *pxTaskDefinition, TaskHandle_t *pxCreatedTask );
@@ -683,18 +693,18 @@ typedef enum
  *  for( ;; );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCreateRestricted xTaskCreateRestricted
  * @endcond
  * \ingroup Tasks
  */
 #if ( portUSING_MPU_WRAPPERS == 1 )
     BaseType_t xTaskCreateRestricted( const TaskParameters_t * const pxTaskDefinition,
-                                      TaskHandle_t * pxCreatedTask );
+                                      TaskHandle_t * pxCreatedTask ) PRIVILEGED_FUNCTION;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskCreateRestrictedStatic( TaskParameters_t *pxTaskDefinition, TaskHandle_t *pxCreatedTask );
@@ -777,7 +787,7 @@ typedef enum
  *  for( ;; );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCreateRestrictedStatic xTaskCreateRestrictedStatic
  * @endcond
  * \ingroup Tasks
@@ -788,7 +798,7 @@ typedef enum
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskAllocateMPURegions( TaskHandle_t xTask, const MemoryRegion_t * const pxRegions );
@@ -833,7 +843,7 @@ typedef enum
  *  // defined or shared regions have been declared elsewhere).
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCreateRestricted xTaskCreateRestricted
  * @endcond
  * \ingroup Tasks
@@ -842,7 +852,7 @@ void vTaskAllocateMPURegions( TaskHandle_t xTask,
                               const MemoryRegion_t * const pxRegions ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskDelete( TaskHandle_t xTask );
@@ -881,7 +891,7 @@ void vTaskAllocateMPURegions( TaskHandle_t xTask,
  *   vTaskDelete( xHandle );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskDelete vTaskDelete
  * @endcond
  * \ingroup Tasks
@@ -893,10 +903,12 @@ void vTaskDelete( TaskHandle_t xTaskToDelete ) PRIVILEGED_FUNCTION;
  *----------------------------------------------------------*/
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskDelay( const TickType_t xTicksToDelay );
  * @endcode
+ * @endcond
  *
  * Delay a task for a given number of ticks.  The actual time that the
  * task remains blocked depends on the tick rate.  The constant
@@ -938,7 +950,7 @@ void vTaskDelete( TaskHandle_t xTaskToDelete ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskDelay vTaskDelay
  * @endcond
  * \ingroup TaskCtrl
@@ -946,10 +958,12 @@ void vTaskDelete( TaskHandle_t xTaskToDelete ) PRIVILEGED_FUNCTION;
 void vTaskDelay( const TickType_t xTicksToDelay ) PRIVILEGED_FUNCTION;
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskDelayUntil( TickType_t *pxPreviousWakeTime, const TickType_t xTimeIncrement );
  * @endcode
+ * @endcond
  *
  * INCLUDE_xTaskDelayUntil must be defined as 1 for this function to be available.
  * See the configuration section for more information.
@@ -1007,7 +1021,7 @@ void vTaskDelay( const TickType_t xTicksToDelay ) PRIVILEGED_FUNCTION;
  *     }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskDelayUntil xTaskDelayUntil
  * @endcond
  * \ingroup TaskCtrl
@@ -1026,7 +1040,7 @@ BaseType_t xTaskDelayUntil( TickType_t * const pxPreviousWakeTime,
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskAbortDelay( TaskHandle_t xTask );
@@ -1054,7 +1068,7 @@ BaseType_t xTaskDelayUntil( TickType_t * const pxPreviousWakeTime,
  * @return If the task referenced by xTask was not in the Blocked state then
  * pdFAIL is returned.  Otherwise pdPASS is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskAbortDelay xTaskAbortDelay
  * @endcond
  * \ingroup TaskCtrl
@@ -1062,7 +1076,7 @@ BaseType_t xTaskDelayUntil( TickType_t * const pxPreviousWakeTime,
 BaseType_t xTaskAbortDelay( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * UBaseType_t uxTaskPriorityGet( const TaskHandle_t xTask );
@@ -1107,7 +1121,7 @@ BaseType_t xTaskAbortDelay( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
  *   }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup uxTaskPriorityGet uxTaskPriorityGet
  * @endcond
  * \ingroup TaskCtrl
@@ -1115,7 +1129,7 @@ BaseType_t xTaskAbortDelay( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 UBaseType_t uxTaskPriorityGet( const TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * UBaseType_t uxTaskPriorityGetFromISR( const TaskHandle_t xTask );
@@ -1127,7 +1141,7 @@ UBaseType_t uxTaskPriorityGet( const TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 UBaseType_t uxTaskPriorityGetFromISR( const TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * eTaskState eTaskGetState( TaskHandle_t xTask );
@@ -1149,7 +1163,7 @@ UBaseType_t uxTaskPriorityGetFromISR( const TaskHandle_t xTask ) PRIVILEGED_FUNC
 eTaskState eTaskGetState( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskGetInfo( TaskHandle_t xTask, TaskStatus_t *pxTaskStatus, BaseType_t xGetFreeStackSpace, eTaskState eState );
@@ -1203,7 +1217,7 @@ eTaskState eTaskGetState( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
  *                eInvalid ); // Include the task state in xTaskDetails.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskGetInfo vTaskGetInfo
  * @endcond
  * \ingroup TaskCtrl
@@ -1214,7 +1228,7 @@ void vTaskGetInfo( TaskHandle_t xTask,
                    eTaskState eState ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskPrioritySet( TaskHandle_t xTask, UBaseType_t uxNewPriority );
@@ -1254,7 +1268,7 @@ void vTaskGetInfo( TaskHandle_t xTask,
  *   vTaskPrioritySet( NULL, tskIDLE_PRIORITY + 1 );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskPrioritySet vTaskPrioritySet
  * @endcond
  * \ingroup TaskCtrl
@@ -1263,7 +1277,7 @@ void vTaskPrioritySet( TaskHandle_t xTask,
                        UBaseType_t uxNewPriority ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskSuspend( TaskHandle_t xTaskToSuspend );
@@ -1312,7 +1326,7 @@ void vTaskPrioritySet( TaskHandle_t xTask,
  *   // with our handle as the parameter.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskSuspend vTaskSuspend
  * @endcond
  * \ingroup TaskCtrl
@@ -1320,7 +1334,7 @@ void vTaskPrioritySet( TaskHandle_t xTask,
 void vTaskSuspend( TaskHandle_t xTaskToSuspend ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskResume( TaskHandle_t xTaskToResume );
@@ -1367,7 +1381,7 @@ void vTaskSuspend( TaskHandle_t xTaskToSuspend ) PRIVILEGED_FUNCTION;
  *   // time in accordance with its priority within the system.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskResume vTaskResume
  * @endcond
  * \ingroup TaskCtrl
@@ -1375,7 +1389,7 @@ void vTaskSuspend( TaskHandle_t xTaskToSuspend ) PRIVILEGED_FUNCTION;
 void vTaskResume( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void xTaskResumeFromISR( TaskHandle_t xTaskToResume );
@@ -1402,7 +1416,7 @@ void vTaskResume( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
  * otherwise pdFALSE. This is used by the ISR to determine if a context switch
  * may be required following the ISR.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskResumeFromISR vTaskResumeFromISR
  * @endcond
  * \ingroup TaskCtrl
@@ -1412,9 +1426,9 @@ BaseType_t xTaskResumeFromISR( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
 /*-----------------------------------------------------------
  * SCHEDULER CONTROL
  *----------------------------------------------------------*/
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskStartScheduler( void );
@@ -1445,7 +1459,7 @@ BaseType_t xTaskResumeFromISR( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskStartScheduler vTaskStartScheduler
  * @endcond
  * \ingroup SchedulerControl
@@ -1453,7 +1467,7 @@ BaseType_t xTaskResumeFromISR( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
 void vTaskStartScheduler( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskEndScheduler( void );
@@ -1507,7 +1521,7 @@ void vTaskStartScheduler( void ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskEndScheduler vTaskEndScheduler
  * @endcond
  * \ingroup SchedulerControl
@@ -1517,7 +1531,7 @@ void vTaskEndScheduler( void ) PRIVILEGED_FUNCTION;
 /** @endcond */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskSuspendAll( void );
@@ -1566,7 +1580,7 @@ void vTaskEndScheduler( void ) PRIVILEGED_FUNCTION;
  *   }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskSuspendAll vTaskSuspendAll
  * @endcond
  * \ingroup SchedulerControl
@@ -1574,7 +1588,7 @@ void vTaskEndScheduler( void ) PRIVILEGED_FUNCTION;
 void vTaskSuspendAll( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskResumeAll( void );
@@ -1626,7 +1640,7 @@ void vTaskSuspendAll( void ) PRIVILEGED_FUNCTION;
  *   }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskResumeAll xTaskResumeAll
  * @endcond
  * \ingroup SchedulerControl
@@ -1638,7 +1652,7 @@ BaseType_t xTaskResumeAll( void ) PRIVILEGED_FUNCTION;
  *----------------------------------------------------------*/
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * TickType_t xTaskGetTickCount( void );
@@ -1647,7 +1661,7 @@ BaseType_t xTaskResumeAll( void ) PRIVILEGED_FUNCTION;
  *
  * @return The count of ticks since vTaskStartScheduler was called.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskGetTickCount xTaskGetTickCount
  * @endcond
  * \ingroup TaskUtils
@@ -1655,7 +1669,7 @@ BaseType_t xTaskResumeAll( void ) PRIVILEGED_FUNCTION;
 TickType_t xTaskGetTickCount( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * TickType_t xTaskGetTickCountFromISR( void );
@@ -1669,7 +1683,7 @@ TickType_t xTaskGetTickCount( void ) PRIVILEGED_FUNCTION;
  * microcontroller being used or interrupt nesting is either not supported or
  * not being used.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskGetTickCountFromISR xTaskGetTickCountFromISR
  * @endcond
  * \ingroup TaskUtils
@@ -1677,7 +1691,7 @@ TickType_t xTaskGetTickCount( void ) PRIVILEGED_FUNCTION;
 TickType_t xTaskGetTickCountFromISR( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * uint16_t uxTaskGetNumberOfTasks( void );
@@ -1689,7 +1703,7 @@ TickType_t xTaskGetTickCountFromISR( void ) PRIVILEGED_FUNCTION;
  * has been deleted but not yet freed by the idle task will also be
  * included in the count.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup uxTaskGetNumberOfTasks uxTaskGetNumberOfTasks
  * @endcond
  * \ingroup TaskUtils
@@ -1697,7 +1711,7 @@ TickType_t xTaskGetTickCountFromISR( void ) PRIVILEGED_FUNCTION;
 UBaseType_t uxTaskGetNumberOfTasks( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * char *pcTaskGetName( TaskHandle_t xTaskToQuery );
@@ -1708,7 +1722,7 @@ UBaseType_t uxTaskGetNumberOfTasks( void ) PRIVILEGED_FUNCTION;
  * xTaskToQuery.  A task can query its own name by either passing in its own
  * handle, or by setting xTaskToQuery to NULL.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup pcTaskGetName pcTaskGetName
  * @endcond
  * \ingroup TaskUtils
@@ -1716,7 +1730,7 @@ UBaseType_t uxTaskGetNumberOfTasks( void ) PRIVILEGED_FUNCTION;
 char * pcTaskGetName( TaskHandle_t xTaskToQuery ) PRIVILEGED_FUNCTION;     /*lint !e971 Unqualified char types are allowed for strings and single characters only. */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * TaskHandle_t xTaskGetHandle( const char *pcNameToQuery );
@@ -1730,7 +1744,7 @@ char * pcTaskGetName( TaskHandle_t xTaskToQuery ) PRIVILEGED_FUNCTION;     /*lin
  * NULL is returned if no matching name is found.  INCLUDE_xTaskGetHandle
  * must be set to 1 in FreeRTOSConfig.h for pcTaskGetHandle() to be available.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup pcTaskGetHandle pcTaskGetHandle
  * @endcond
  * \ingroup TaskUtils
@@ -1813,7 +1827,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 #ifdef configUSE_APPLICATION_TASK_TAG
     #if configUSE_APPLICATION_TASK_TAG == 1
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * void vTaskSetApplicationTaskTag( TaskHandle_t xTask, TaskHookFunction_t pxHookFunction );
@@ -1830,7 +1844,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
                                          TaskHookFunction_t pxHookFunction ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * void xTaskGetApplicationTaskTag( TaskHandle_t xTask );
@@ -1844,7 +1858,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
         TaskHookFunction_t xTaskGetApplicationTaskTag( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * void xTaskGetApplicationTaskTagFromISR( TaskHandle_t xTask );
@@ -1932,7 +1946,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 #if ( configCHECK_FOR_STACK_OVERFLOW > 0 )
 
      /**
-      * @cond
+      * @cond !DOC_EXCLUDE_HEADER_SECTION
       * task.h
       * @code{c}
       * void vApplicationStackOverflowHook( TaskHandle_t xTask char *pcTaskName);
@@ -1952,7 +1966,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 
 #if  (  configUSE_TICK_HOOK > 0 )
     /**
-     * @cond
+     * @cond !DOC_EXCLUDE_HEADER_SECTION
      *  task.h
      * @code{c}
      * void vApplicationTickHook( void );
@@ -1967,7 +1981,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 
 #if ( configSUPPORT_STATIC_ALLOCATION == 1 )
     /**
-     * @cond
+     * @cond !DOC_EXCLUDE_HEADER_SECTION
      * task.h
      * @code{c}
      * void vApplicationGetIdleTaskMemory( StaticTask_t ** ppxIdleTaskTCBBuffer, StackType_t ** ppxIdleTaskStackBuffer, uint32_t *pulIdleTaskStackSize )
@@ -1986,7 +2000,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * BaseType_t xTaskCallApplicationTaskHook( TaskHandle_t xTask, void *pvParameter );
@@ -2155,7 +2169,7 @@ UBaseType_t uxTaskGetSystemState( TaskStatus_t * const pxTaskStatusArray,
  * enough to contain the generated report.  Approximately 40 bytes per
  * task should be sufficient.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskList vTaskList
  * @endcond
  * \ingroup TaskUtils
@@ -2210,7 +2224,7 @@ void vTaskList( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lint !e971 Unq
  * contain the generated report.  Approximately 40 bytes per task should
  * be sufficient.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskGetRunTimeStats vTaskGetRunTimeStats
  * @endcond
  * \ingroup TaskUtils
@@ -2218,7 +2232,7 @@ void vTaskList( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lint !e971 Unq
 void vTaskGetRunTimeStats( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lint !e971 Unqualified char types are allowed for strings and single characters only. */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code
  * uint32_t ulTaskGetIdleRunTimeCounter( void );
@@ -2246,7 +2260,7 @@ void vTaskGetRunTimeStats( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lin
  * frequency configured using the portCONFIGURE_TIMER_FOR_RUN_TIME_STATS() and
  * portGET_RUN_TIME_COUNTER_VALUE() macros.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup ulTaskGetIdleRunTimeCounter ulTaskGetIdleRunTimeCounter
  * @endcond
  * \ingroup TaskUtils
@@ -2254,11 +2268,13 @@ void vTaskGetRunTimeStats( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lin
 uint32_t ulTaskGetIdleRunTimeCounter( void ) PRIVILEGED_FUNCTION;
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyIndexed( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify, uint32_t ulValue, eNotifyAction eAction );
  * BaseType_t xTaskNotify( TaskHandle_t xTaskToNotify, uint32_t ulValue, eNotifyAction eAction );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2359,7 +2375,9 @@ uint32_t ulTaskGetIdleRunTimeCounter( void ) PRIVILEGED_FUNCTION;
  * @return Dependent on the value of eAction.  See the description of the
  * eAction parameter.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyIndexed xTaskNotifyIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
@@ -2373,11 +2391,13 @@ BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
     xTaskGenericNotify( ( xTaskToNotify ), ( uxIndexToNotify ), ( ulValue ), ( eAction ), NULL )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyAndQueryIndexed( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify, uint32_t ulValue, eNotifyAction eAction, uint32_t *pulPreviousNotifyValue );
  * BaseType_t xTaskNotifyAndQuery( TaskHandle_t xTaskToNotify, uint32_t ulValue, eNotifyAction eAction, uint32_t *pulPreviousNotifyValue );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2393,7 +2413,9 @@ BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
  * than when the function returns) in the additional pulPreviousNotifyValue
  * parameter.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyAndQueryIndexed xTaskNotifyAndQueryIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 #define xTaskNotifyAndQuery( xTaskToNotify, ulValue, eAction, pulPreviousNotifyValue ) \
@@ -2402,11 +2424,13 @@ BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
     xTaskGenericNotify( ( xTaskToNotify ), ( uxIndexToNotify ), ( ulValue ), ( eAction ), ( pulPreviousNotifyValue ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyIndexedFromISR( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify, uint32_t ulValue, eNotifyAction eAction, BaseType_t *pxHigherPriorityTaskWoken );
  * BaseType_t xTaskNotifyFromISR( TaskHandle_t xTaskToNotify, uint32_t ulValue, eNotifyAction eAction, BaseType_t *pxHigherPriorityTaskWoken );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2511,7 +2535,9 @@ BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
  * @return Dependent on the value of eAction.  See the description of the
  * eAction parameter.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyIndexedFromISR xTaskNotifyIndexedFromISR
+ * @endcond
  * \ingroup TaskNotifications
  */
 BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
@@ -2526,11 +2552,13 @@ BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
     xTaskGenericNotifyFromISR( ( xTaskToNotify ), ( uxIndexToNotify ), ( ulValue ), ( eAction ), NULL, ( pxHigherPriorityTaskWoken ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyAndQueryIndexedFromISR( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify, uint32_t ulValue, eNotifyAction eAction, uint32_t *pulPreviousNotificationValue, BaseType_t *pxHigherPriorityTaskWoken );
  * BaseType_t xTaskNotifyAndQueryFromISR( TaskHandle_t xTaskToNotify, uint32_t ulValue, eNotifyAction eAction, uint32_t *pulPreviousNotificationValue, BaseType_t *pxHigherPriorityTaskWoken );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2546,7 +2574,9 @@ BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
  * function is called rather than at the time the function returns) in the
  * additional pulPreviousNotifyValue parameter.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyAndQueryIndexedFromISR xTaskNotifyAndQueryIndexedFromISR
+ * @endcond
  * \ingroup TaskNotifications
  */
 #define xTaskNotifyAndQueryIndexedFromISR( xTaskToNotify, uxIndexToNotify, ulValue, eAction, pulPreviousNotificationValue, pxHigherPriorityTaskWoken ) \
@@ -2555,12 +2585,14 @@ BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
     xTaskGenericNotifyFromISR( ( xTaskToNotify ), ( tskDEFAULT_INDEX_TO_NOTIFY ), ( ulValue ), ( eAction ), ( pulPreviousNotificationValue ), ( pxHigherPriorityTaskWoken ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyWaitIndexed( UBaseType_t uxIndexToWaitOn, uint32_t ulBitsToClearOnEntry, uint32_t ulBitsToClearOnExit, uint32_t *pulNotificationValue, TickType_t xTicksToWait );
  *
  * BaseType_t xTaskNotifyWait( uint32_t ulBitsToClearOnEntry, uint32_t ulBitsToClearOnExit, uint32_t *pulNotificationValue, TickType_t xTicksToWait );
  * @endcode
+ * @endcond
  *
  * Waits for a direct to task notification to be pending at a given index within
  * an array of direct to task notifications.
@@ -2655,7 +2687,9 @@ BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
  * already pending when xTaskNotifyWait was called) then pdPASS is
  * returned.  Otherwise pdFAIL is returned.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyWaitIndexed xTaskNotifyWaitIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
@@ -2669,11 +2703,13 @@ BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
     xTaskGenericNotifyWait( ( uxIndexToWaitOn ), ( ulBitsToClearOnEntry ), ( ulBitsToClearOnExit ), ( pulNotificationValue ), ( xTicksToWait ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyGiveIndexed( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify );
  * BaseType_t xTaskNotifyGive( TaskHandle_t xTaskToNotify );
  * @endcode
+ * @endcond
  *
  * Sends a direct to task notification to a particular index in the target
  * task's notification array in a manner similar to giving a counting semaphore.
@@ -2737,7 +2773,9 @@ BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
  * @return xTaskNotifyGive() is a macro that calls xTaskNotify() with the
  * eAction parameter set to eIncrement - so pdPASS is always returned.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyGiveIndexed xTaskNotifyGiveIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 #define xTaskNotifyGive( xTaskToNotify ) \
@@ -2746,11 +2784,13 @@ BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
     xTaskGenericNotify( ( xTaskToNotify ), ( uxIndexToNotify ), ( 0 ), eIncrement, NULL )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskNotifyGiveIndexedFromISR( TaskHandle_t xTaskHandle, UBaseType_t uxIndexToNotify, BaseType_t *pxHigherPriorityTaskWoken );
  * void vTaskNotifyGiveFromISR( TaskHandle_t xTaskHandle, BaseType_t *pxHigherPriorityTaskWoken );
  * @endcode
+ * @endcond
  *
  * A version of xTaskNotifyGiveIndexed() that can be called from an interrupt
  * service routine (ISR).
@@ -2821,7 +2861,9 @@ BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
  * requested from an ISR is dependent on the port - see the documentation page
  * for the port in use.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskNotifyGiveIndexedFromISR vTaskNotifyGiveIndexedFromISR
+ * @endcond
  * \ingroup TaskNotifications
  */
 void vTaskGenericNotifyGiveFromISR( TaskHandle_t xTaskToNotify,
@@ -2833,12 +2875,14 @@ void vTaskGenericNotifyGiveFromISR( TaskHandle_t xTaskToNotify,
     vTaskGenericNotifyGiveFromISR( ( xTaskToNotify ), ( uxIndexToNotify ), ( pxHigherPriorityTaskWoken ) );
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * uint32_t ulTaskNotifyTakeIndexed( UBaseType_t uxIndexToWaitOn, BaseType_t xClearCountOnExit, TickType_t xTicksToWait );
  *
  * uint32_t ulTaskNotifyTake( BaseType_t xClearCountOnExit, TickType_t xTicksToWait );
  * @endcode
+ * @endcond
  *
  * Waits for a direct to task notification on a particular index in the calling
  * task's notification array in a manner similar to taking a counting semaphore.
@@ -2927,7 +2971,9 @@ void vTaskGenericNotifyGiveFromISR( TaskHandle_t xTaskToNotify,
  * @return The task's notification count before it is either cleared to zero or
  * decremented (see the xClearCountOnExit parameter).
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup ulTaskNotifyTakeIndexed ulTaskNotifyTakeIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 uint32_t ulTaskGenericNotifyTake( UBaseType_t uxIndexToWaitOn,
@@ -2939,12 +2985,14 @@ uint32_t ulTaskGenericNotifyTake( UBaseType_t uxIndexToWaitOn,
     ulTaskGenericNotifyTake( ( uxIndexToWaitOn ), ( xClearCountOnExit ), ( xTicksToWait ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyStateClearIndexed( TaskHandle_t xTask, UBaseType_t uxIndexToCLear );
  *
  * BaseType_t xTaskNotifyStateClear( TaskHandle_t xTask );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2992,7 +3040,9 @@ uint32_t ulTaskGenericNotifyTake( UBaseType_t uxIndexToWaitOn,
  * @return pdTRUE if the task's notification state was set to
  * eNotWaitingNotification, otherwise pdFALSE.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyStateClearIndexed xTaskNotifyStateClearIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 BaseType_t xTaskGenericNotifyStateClear( TaskHandle_t xTask,
@@ -3003,12 +3053,14 @@ BaseType_t xTaskGenericNotifyStateClear( TaskHandle_t xTask,
     xTaskGenericNotifyStateClear( ( xTask ), ( uxIndexToClear ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * uint32_t ulTaskNotifyValueClearIndexed( TaskHandle_t xTask, UBaseType_t uxIndexToClear, uint32_t ulBitsToClear );
  *
  * uint32_t ulTaskNotifyValueClear( TaskHandle_t xTask, uint32_t ulBitsToClear );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -3057,7 +3109,9 @@ BaseType_t xTaskGenericNotifyStateClear( TaskHandle_t xTask,
  *
  * @return The value of the target task's notification value before the bits
  * specified by ulBitsToClear were cleared.
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup ulTaskNotifyValueClear ulTaskNotifyValueClear
+ * @endcond
  * \ingroup TaskNotifications
  */
 uint32_t ulTaskGenericNotifyValueClear( TaskHandle_t xTask,
@@ -3069,7 +3123,7 @@ uint32_t ulTaskGenericNotifyValueClear( TaskHandle_t xTask,
     ulTaskGenericNotifyValueClear( ( xTask ), ( uxIndexToClear ), ( ulBitsToClear ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * void vTaskSetTimeOutState( TimeOut_t * const pxTimeOut );
@@ -3082,14 +3136,14 @@ uint32_t ulTaskGenericNotifyValueClear( TaskHandle_t xTask,
  * is to be captured.  The captured time includes the tick count and the number
  * of times the tick count has overflowed since the system first booted.
  * \defgroup vTaskSetTimeOutState vTaskSetTimeOutState
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \ingroup TaskCtrl
  * @endcond
  */
 void vTaskSetTimeOutState( TimeOut_t * const pxTimeOut ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code
  * BaseType_t xTaskCheckForTimeOut( TimeOut_t * const pxTimeOut, TickType_t * const pxTicksToWait );
@@ -3170,7 +3224,7 @@ void vTaskSetTimeOutState( TimeOut_t * const pxTimeOut ) PRIVILEGED_FUNCTION;
  *      return uxReceived;
  *  }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCheckForTimeOut xTaskCheckForTimeOut
  * @endcond
  * \ingroup TaskCtrl
@@ -3179,7 +3233,7 @@ BaseType_t xTaskCheckForTimeOut( TimeOut_t * const pxTimeOut,
                                  TickType_t * const pxTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * BaseType_t xTaskCatchUpTicks( TickType_t xTicksToCatchUp );
@@ -3204,7 +3258,7 @@ BaseType_t xTaskCheckForTimeOut( TimeOut_t * const pxTimeOut,
  * blocked state and a context switch being performed.  Otherwise pdFALSE.
  *
  * \defgroup xTaskCatchUpTicks xTaskCatchUpTicks
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \ingroup TaskCtrl
  * @endcond
  */
@@ -3214,7 +3268,7 @@ BaseType_t xTaskCatchUpTicks( TickType_t xTicksToCatchUp ) PRIVILEGED_FUNCTION;
 /*-----------------------------------------------------------
  * SCHEDULER INTERNALS AVAILABLE FOR PORTING PURPOSES
  *----------------------------------------------------------*/
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /*
  * Return the handle of the task running on a certain CPU. Because of
  * the nature of SMP processing, there is no guarantee that this
@@ -3335,8 +3389,8 @@ void vTaskPlaceOnEventListRestricted( List_t * const pxEventList,
  * making the call, otherwise pdFALSE.
  */
 BaseType_t xTaskRemoveFromEventList( const List_t * const pxEventList ) PRIVILEGED_FUNCTION;
-BaseType_t xTaskRemoveFromUnorderedEventList( ListItem_t * pxEventListItem,
-                                              const TickType_t xItemValue ) PRIVILEGED_FUNCTION;
+void vTaskRemoveFromUnorderedEventList( ListItem_t * pxEventListItem,
+                                        const TickType_t xItemValue ) PRIVILEGED_FUNCTION;
 
 /*
  * THIS FUNCTION MUST NOT BE USED FROM APPLICATION CODE.  IT IS ONLY
@@ -3399,11 +3453,6 @@ void vTaskPriorityDisinheritAfterTimeout( TaskHandle_t const pxMutexHolder,
  */
 UBaseType_t uxTaskGetTaskNumber( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
-/*
- * Get the current core affinity of a task
- */
-BaseType_t xTaskGetAffinity( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
-
 /*
  * Set the uxTaskNumber of the task referenced by the xTask parameter to
  * uxHandle.
diff --git a/tools/sdk/esp32/include/freertos/include/freertos/timers.h b/tools/sdk/esp32/include/freertos/include/freertos/timers.h
index a8bc4f38c78..af6dcb23501 100644
--- a/tools/sdk/esp32/include/freertos/include/freertos/timers.h
+++ b/tools/sdk/esp32/include/freertos/include/freertos/timers.h
@@ -450,7 +450,7 @@ void vTimerSetTimerID( TimerHandle_t xTimer,
 BaseType_t xTimerIsTimerActive( TimerHandle_t xTimer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * TaskHandle_t xTimerGetTimerDaemonTaskHandle( void );
  * @endcond
  *
@@ -1315,7 +1315,7 @@ TickType_t xTimerGetPeriod( TimerHandle_t xTimer ) PRIVILEGED_FUNCTION;
  */
 TickType_t xTimerGetExpiryTime( TimerHandle_t xTimer ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 
 /*
  * Functions beyond this part are not part of the public API and are intended
@@ -1339,7 +1339,7 @@ BaseType_t xTimerGenericCommand( TimerHandle_t xTimer,
 #if ( configSUPPORT_STATIC_ALLOCATION == 1 )
 
     /**
-     * @cond
+     * @cond !DOC_EXCLUDE_HEADER_SECTION
      * task.h
      * @code{c}
      * void vApplicationGetTimerTaskMemory( StaticTask_t ** ppxTimerTaskTCBBuffer, StackType_t ** ppxTimerTaskStackBuffer, uint32_t *pulTimerTaskStackSize )
diff --git a/tools/sdk/esp32/include/freertos/port/xtensa/include/freertos/portmacro.h b/tools/sdk/esp32/include/freertos/port/xtensa/include/freertos/portmacro.h
index 248c86d15c7..47187379c6c 100644
--- a/tools/sdk/esp32/include/freertos/port/xtensa/include/freertos/portmacro.h
+++ b/tools/sdk/esp32/include/freertos/port/xtensa/include/freertos/portmacro.h
@@ -24,317 +24,449 @@
  *
  * 1 tab == 4 spaces!
  */
+
 #ifndef PORTMACRO_H
 #define PORTMACRO_H
 
-#ifdef __cplusplus
-extern "C" {
-#endif
-
 #ifndef __ASSEMBLER__
 
-#include <sdkconfig.h>
+#include "sdkconfig.h"
 #include <stdint.h>
 #include <stdlib.h>
 #include <stdbool.h>
 #include <stdarg.h>
-#include <xtensa/hal.h>
 #include <xtensa/config/core.h>
-#include <xtensa/config/system.h>	/* required for XSHAL_CLIB */
-#include <xtensa/xtruntime.h>
-#include "esp_private/crosscore_int.h"
-#include "esp_timer.h"              /* required for FreeRTOS run time stats */
-#include "esp_system.h"
-#include "esp_newlib.h"
+#include <xtensa/hal.h>             /* required for xthal_get_ccount. [refactor-todo] use cpu_hal instead */
+#include <xtensa/xtruntime.h>       /* required for XTOS_SET_INTLEVEL. [refactor-todo] add common intr functions to esp_hw_support */
+#include "xt_instr_macros.h"
 #include "soc/spinlock.h"
-#include <esp_heap_caps.h>
+#include "hal/cpu_hal.h"
+#include "esp_private/crosscore_int.h"
+#include "esp_attr.h"
+#include "esp_timer.h"              /* required for esp_timer_get_time. [refactor-todo] make this common between archs */
+#include "esp_newlib.h"             /* required for esp_reent_init() in tasks.c */
+#include "esp_heap_caps.h"
 #include "esp_rom_sys.h"
-#include "sdkconfig.h"
-#include "freertos/xtensa_api.h"
-#include "esp_system.h"
-#include "soc/cpu.h"
-#include <limits.h>
+#include "esp_system.h"             /* required by esp_get_...() functions in portable.h. [refactor-todo] Update portable.h */
+#include "portbenchmark.h"
 
+/* [refactor-todo] These includes are not directly used in this file. They are kept into to prevent a breaking change. Remove these. */
+#include <limits.h>
+#include <xtensa/config/system.h>
+#include <xtensa/xtensa_api.h>
+#include "soc/cpu.h"
 #ifdef CONFIG_LEGACY_INCLUDE_COMMON_HEADERS
 #include "soc/soc_memory_layout.h"
 #endif
 
-/*-----------------------------------------------------------
- * Port specific definitions.
- *
- * The settings in this file configure FreeRTOS correctly for the
- * given hardware and compiler.
- *
- * These settings should not be altered.
- *-----------------------------------------------------------
- */
+#ifdef __cplusplus
+extern "C" {
+#endif
+
 
-#include "esp_system.h"
-#include "hal/cpu_hal.h"
-#include "xt_instr_macros.h"
 
-/* Type definitions. */
-#define portCHAR		int8_t
-#define portFLOAT		float
-#define portDOUBLE		double
-#define portLONG		int32_t
-#define portSHORT		int16_t
-#define portSTACK_TYPE	uint8_t
-#define portBASE_TYPE	int
+/* --------------------------------------------------- Port Types ------------------------------------------------------
+ * - Port specific types.
+ * - The settings in this file configure FreeRTOS correctly for the given hardware and compiler.
+ * - These settings should not be altered.
+ * - The port types must come before first as they are used further down the file
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-typedef portSTACK_TYPE			StackType_t;
-typedef portBASE_TYPE			BaseType_t;
-typedef unsigned portBASE_TYPE	UBaseType_t;
+#define portCHAR                    int8_t
+#define portFLOAT                   float
+#define portDOUBLE                  double
+#define portLONG                    int32_t
+#define portSHORT                   int16_t
+#define portSTACK_TYPE              uint8_t
+#define portBASE_TYPE               int
+
+typedef portSTACK_TYPE              StackType_t;
+typedef portBASE_TYPE               BaseType_t;
+typedef unsigned portBASE_TYPE      UBaseType_t;
 
 #if( configUSE_16_BIT_TICKS == 1 )
-	typedef uint16_t TickType_t;
-	#define portMAX_DELAY ( TickType_t ) 0xffff
+typedef uint16_t TickType_t;
+#define portMAX_DELAY ( TickType_t ) 0xffff
 #else
-	typedef uint32_t TickType_t;
-	#define portMAX_DELAY ( TickType_t ) 0xffffffffUL
+typedef uint32_t TickType_t;
+#define portMAX_DELAY ( TickType_t ) 0xffffffffUL
 #endif
-/*-----------------------------------------------------------*/
 
-// portbenchmark
-#include "portbenchmark.h"
+/* Task function macros as described on the FreeRTOS.org WEB site. */
+#define portTASK_FUNCTION_PROTO( vFunction, pvParameters )  void vFunction( void *pvParameters )
+#define portTASK_FUNCTION( vFunction, pvParameters )        void vFunction( void *pvParameters )
 
-#include "sdkconfig.h"
-#include "esp_attr.h"
-#include "portmacro_priv.h"
-
-// Cleaner solution allows nested interrupts disabling and restoring via local registers or stack.
-// They can be called from interrupts too.
-// WARNING: Only applies to current CPU. See notes above.
-static inline unsigned portENTER_CRITICAL_NESTED(void) {
-	unsigned state = XTOS_SET_INTLEVEL(XCHAL_EXCM_LEVEL);
-	portbenchmarkINTERRUPT_DISABLE();
-	return state;
-}
-#define portEXIT_CRITICAL_NESTED(state)   do { portbenchmarkINTERRUPT_RESTORE(state); XTOS_RESTORE_JUST_INTLEVEL(state); } while (0)
 
-/*
-Modifications to portENTER_CRITICAL.
 
-For an introduction, see "Critical Sections & Disabling Interrupts" in docs/api-guides/freertos-smp.rst
+/* ----------------------------------------------- Port Configurations -------------------------------------------------
+ * - Configurations values supplied by each port
+ * - Required by FreeRTOS
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-The original portENTER_CRITICAL only disabled the ISRs. This is enough for single-CPU operation: by
-disabling the interrupts, there is no task switch so no other tasks can meddle in the data, and because
-interrupts are disabled, ISRs can't corrupt data structures either.
+#define portCRITICAL_NESTING_IN_TCB     0
+#define portSTACK_GROWTH                ( -1 )
+#define portTICK_PERIOD_MS              ( ( TickType_t ) 1000 / configTICK_RATE_HZ )
+#define portBYTE_ALIGNMENT              4
+#define portNOP()                       XT_NOP()
 
-For multiprocessing, things get a bit more hairy. First of all, disabling the interrupts doesn't stop
-the tasks or ISRs on the other processors meddling with our CPU. For tasks, this is solved by adding
-a spinlock to the portENTER_CRITICAL macro. A task running on the other CPU accessing the same data will
-spinlock in the portENTER_CRITICAL code until the first CPU is done.
 
-For ISRs, we now also need muxes: while portENTER_CRITICAL disabling interrupts will stop ISRs on the same
-CPU from meddling with the data, it does not stop interrupts on the other cores from interfering with the
-data. For this, we also use a spinlock in the routines called by the ISR, but these spinlocks
-do not disable the interrupts (because they already are).
 
-This all assumes that interrupts are either entirely disabled or enabled. Interrupt priority levels
-will break this scheme.
+/* ---------------------------------------------- Forward Declarations -------------------------------------------------
+ * - Forward declarations of all the port functions and macros need to implement the FreeRTOS porting interface
+ * - These must come before definition/declaration of the FreeRTOS porting interface
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-Remark: For the ESP32, portENTER_CRITICAL and portENTER_CRITICAL_ISR both alias vPortEnterCritical, meaning
-that either function can be called both from ISR as well as task context. This is not standard FreeRTOS
-behaviour; please keep this in mind if you need any compatibility with other FreeRTOS implementations.
-*/
-/* "mux" data structure (spinlock) */
-typedef spinlock_t portMUX_TYPE;
+// --------------------- Interrupts ------------------------
 
-#define portMUX_FREE_VAL		SPINLOCK_FREE
-#define portMUX_NO_TIMEOUT      SPINLOCK_WAIT_FOREVER  /* When passed for 'timeout_cycles', spin forever if necessary */
-#define portMUX_TRY_LOCK        SPINLOCK_NO_WAIT       /* Try to acquire the spinlock a single time only */
-#define portMUX_INITIALIZER_UNLOCKED  SPINLOCK_INITIALIZER
+/**
+ * @brief Checks if the current core is in an ISR context
+ *
+ * - ISR context consist of Low/Mid priority ISR, or time tick ISR
+ * - High priority ISRs aren't detected here, but they normally cannot call C code, so that should not be an issue anyway.
+ *
+ * @note [refactor-todo] Check if this should be inlined
+ * @return
+ *  - pdTRUE if in ISR
+ *  - pdFALSE otherwise
+ */
+BaseType_t xPortInIsrContext(void);
 
-#define portCRITICAL_NESTING_IN_TCB 0
+/**
+ * @brief Asserts if in ISR context
+ *
+ * - Asserts on xPortInIsrContext() internally
+ *
+ * @note [refactor-todo] Check if this API is still required
+ * @note [refactor-todo] Check if this should be inlined
+ */
+void vPortAssertIfInISR(void);
 
-static inline void __attribute__((always_inline)) vPortCPUInitializeMutex(portMUX_TYPE *mux)
-{
-    spinlock_initialize(mux);
-}
+/**
+ * @brief Check if in ISR context from High priority ISRs
+ *
+ * - Called from High priority ISR
+ * - Checks if the previous context (before high priority interrupt) was in ISR context (meaning low/med priority)
+ *
+ * @note [refactor-todo] Check if this should be inlined
+ * @return
+ *  - pdTRUE if in previous in ISR context
+ *  - pdFALSE otherwise
+ */
+BaseType_t xPortInterruptedFromISRContext(void);
 
-static inline void __attribute__((always_inline)) vPortCPUAcquireMutex(portMUX_TYPE *mux)
-{
-    spinlock_acquire(mux, portMUX_NO_TIMEOUT);
-}
+/**
+ * @brief Disable interrupts in a nested manner
+ *
+ * - Cleaner solution allows nested interrupts disabling and restoring via local registers or stack.
+ * - They can be called from interrupts too.
+ * - WARNING Only applies to current CPU.
+ * @note [refactor-todo] Define this as portSET_INTERRUPT_MASK_FROM_ISR() instead
+ * @return unsigned Previous interrupt state
+ */
+static inline unsigned __attribute__((always_inline)) portENTER_CRITICAL_NESTED(void);
+
+/* ---------------------- Spinlocks ------------------------
+ * - Modifications made to critical sections to support SMP
+ * - See "Critical Sections & Disabling Interrupts" in docs/api-guides/freertos-smp.rst for more details
+ * - Remark: For the ESP32, portENTER_CRITICAL and portENTER_CRITICAL_ISR both alias vPortEnterCritical, meaning that
+ *           either function can be called both from ISR as well as task context. This is not standard FreeRTOS
+ *           behaviorr; please keep this in mind if you need any compatibility with other FreeRTOS implementations.
+ * @note [refactor-todo] Check if these comments are still true
+ * ------------------------------------------------------ */
+
+typedef spinlock_t                      portMUX_TYPE;           /**< Spinlock type used by FreeRTOS critical sections */
+#define portMUX_INITIALIZER_UNLOCKED    SPINLOCK_INITIALIZER    /**< Spinlock initializer */
+#define portMUX_FREE_VAL                SPINLOCK_FREE           /**< Spinlock is free. [refactor-todo] check if this is still required */
+#define portMUX_NO_TIMEOUT              SPINLOCK_WAIT_FOREVER   /**< When passed for 'timeout_cycles', spin forever if necessary. [refactor-todo] check if this is still required */
+#define portMUX_TRY_LOCK                SPINLOCK_NO_WAIT        /**< Try to acquire the spinlock a single time only. [refactor-todo] check if this is still required */
 
-static inline bool __attribute__((always_inline)) vPortCPUAcquireMutexTimeout(portMUX_TYPE *mux, int timeout)
-{
-    return (spinlock_acquire(mux, timeout));
-}
+/**
+ * @brief Initialize a spinlock
+ *
+ * - Initializes a spinlock that is used by FreeRTOS SMP critical sections
+ *
+ * @param[in] mux Spinlock
+ */
+static inline void __attribute__((always_inline)) vPortCPUInitializeMutex(portMUX_TYPE *mux);
 
-static inline void __attribute__((always_inline)) vPortCPUReleaseMutex(portMUX_TYPE *mux)
-{
-    spinlock_release(mux);
-}
+/**
+ * @brief Acquire a spinlock
+ *
+ * @note [refactor-todo] check if we still need this
+ *
+ * @param[in] mux Spinlock
+ */
+static inline void __attribute__((always_inline)) vPortCPUAcquireMutex(portMUX_TYPE *mux);
 
+/**
+ * @brief Acquire a spinlock but with a specified timeout
+ *
+ * @note [refactor-todo] check if we still need this
+ * @note [refactor-todo] Check if this function should be renamed (due to bool return type)
+ *
+ * @param[in] mux Spinlock
+ * @param timeout
+ * @return true Spinlock acquired
+ * @return false Timed out
+ */
+static inline bool __attribute__((always_inline)) vPortCPUAcquireMutexTimeout(portMUX_TYPE *mux, int timeout);
+
+/**
+ * @brief Release a spinlock
+ *
+ * @note [refactor-todo] check if we still need this
+ *
+ * @param[in] mux Spinlock
+ */
+static inline void __attribute__((always_inline)) vPortCPUReleaseMutex(portMUX_TYPE *mux);
+
+/**
+ * @brief Wrapper for atomic compare-and-set instruction
+ *
+ * This subroutine will atomically compare *addr to 'compare'. If *addr == compare, *addr is set to *set. *set is
+ * updated with the previous value of *addr (either 'compare' or some other value.)
+ *
+ * @warning From the ISA docs: in some (unspecified) cases, the s32c1i instruction may return the "bitwise inverse" of
+ *          the old mem if the mem wasn't written. This doesn't seem to happen on the ESP32 (portMUX assertions would
+ *          fail).
+ *
+ * @note [refactor-todo] check if we still need this
+ * @note [refactor-todo] Check if this function should be renamed (due to void return type)
+ *
+ * @param[inout] addr Pointer to target address
+ * @param[in] compare Compare value
+ * @param[inout] set Pointer to set value
+ */
+static inline void __attribute__((always_inline)) uxPortCompareSet(volatile uint32_t *addr, uint32_t compare, uint32_t *set);
+
+/**
+ * @brief Wrapper for atomic compare-and-set instruction in external RAM
+ *
+ * Atomic compare-and-set but the target address is placed in external RAM
+ *
+ * @note [refactor-todo] check if we still need this
+ *
+ * @param[inout] addr Pointer to target address
+ * @param[in] compare Compare value
+ * @param[inout] set Pointer to set value
+ */
+static inline void uxPortCompareSetExtram(volatile uint32_t *addr, uint32_t compare, uint32_t *set);
+
+// ------------------ Critical Sections --------------------
+
+/**
+ * @brief Enter a SMP critical section
+ *
+ * - Disable interrupts
+ * - Takes spinlock
+ * - Can be nested
+ *
+ * @param[in] mux Spinlock
+ */
 void vPortEnterCritical(portMUX_TYPE *mux);
+
+/**
+ * @brief Exit a SMP critical section
+ *
+ * - Releases spinlock
+ * - Reenables interrupts
+ * - Can be nested
+ *
+ * @param[in] mux Spinlock
+ */
 void vPortExitCritical(portMUX_TYPE *mux);
 
-#define portASSERT_IF_IN_ISR()  vPortAssertIfInISR()
-void vPortAssertIfInISR(void);
+/**
+ * @brief FreeRTOS compliant version of enter critical
+ *
+ * - Ensures that critical section is only entered from task context
+ *
+ * @param[in] mux Spinlock
+ */
+static inline void __attribute__((always_inline)) vPortEnterCriticalCompliance(portMUX_TYPE *mux);
 
-/*
- * Returns true if the current core is in ISR context; low prio ISR, med prio ISR or timer tick ISR. High prio ISRs
- * aren't detected here, but they normally cannot call C code, so that should not be an issue anyway.
+/**
+ * @brief FreeRTOS compliant version of exit critical
+ *
+ * @param[in] mux Spinlock
  */
-BaseType_t xPortInIsrContext(void);
+static inline void __attribute__((always_inline)) vPortExitCriticalCompliance(portMUX_TYPE *mux);
 
-static inline void __attribute__((always_inline)) vPortEnterCriticalCompliance(portMUX_TYPE *mux)
-{
-    if(!xPortInIsrContext()) {
-        vPortEnterCritical(mux);
-    } else {
-        esp_rom_printf("%s:%d (%s)- port*_CRITICAL called from ISR context!\n",
-                       __FILE__, __LINE__, __FUNCTION__);
-        abort();
-    }
-}
+/**
+ * @brief Safe version of enter critical
+ *
+ * - This function can be used to enter a critical section from both task and ISR contexts
+ *
+ * @param[in] mux Spinlock
+ */
+static inline void __attribute__((always_inline)) vPortEnterCriticalSafe(portMUX_TYPE *mux);
 
-static inline void __attribute__((always_inline)) vPortExitCriticalCompliance(portMUX_TYPE *mux)
-{
-    if(!xPortInIsrContext()) {
-        vPortExitCritical(mux);
-    } else {
-        esp_rom_printf("%s:%d (%s)- port*_CRITICAL called from ISR context!\n",
-                       __FILE__, __LINE__, __FUNCTION__);
-        abort();
-    }
-}
+/**
+ * @brief Safe version of exit critical
+ *
+ * @param[in] mux Spinlock
+ */
+static inline void __attribute__((always_inline)) vPortExitCriticalSafe(portMUX_TYPE *mux);
 
-#ifdef CONFIG_FREERTOS_CHECK_PORT_CRITICAL_COMPLIANCE
-/* Calling port*_CRITICAL from ISR context would cause an assert failure.
- * If the parent function is called from both ISR and Non-ISR context then call port*_CRITICAL_SAFE
+// ---------------------- Yielding -------------------------
+
+/**
+ * @brief Perform a solicited context switch
+ *
+ * - Defined in portasm.S
+ *
+ * @note [refactor-todo] The rest of ESP-IDF should call taskYield() instead
  */
-#define portENTER_CRITICAL(mux)        vPortEnterCriticalCompliance(mux)
-#define portEXIT_CRITICAL(mux)         vPortExitCriticalCompliance(mux)
-#else
-#define portENTER_CRITICAL(mux)        vPortEnterCritical(mux)
-#define portEXIT_CRITICAL(mux)         vPortExitCritical(mux)
-#endif
+void vPortYield( void );
 
-#define portENTER_CRITICAL_ISR(mux)    vPortEnterCritical(mux)
-#define portEXIT_CRITICAL_ISR(mux)     vPortExitCritical(mux)
+/**
+ * @brief
+ *
+ * @note [refactor-todo] Refactor this to avoid va_args
+ * @param argc
+ * @param ... Variable arguments to allow for IDF prototype without arguments, and vanilla version WITH argument
+ */
+void vPortEvaluateYieldFromISR(int argc, ...);
 
-static inline void __attribute__((always_inline)) vPortEnterCriticalSafe(portMUX_TYPE *mux)
-{
-    if (xPortInIsrContext()) {
-        portENTER_CRITICAL_ISR(mux);
-    } else {
-        portENTER_CRITICAL(mux);
-    }
-}
+/**
+ * @brief Yields the other core
+ *
+ * - Send an interrupt to another core in order to make the task running on it yield for a higher-priority task.
+ * - Can be used to yield current core as well
+ *
+ * @note [refactor-todo] Put this into private macros as its only called from task.c and is not public API
+ * @param coreid ID of core to yield
+ */
+void vPortYieldOtherCore(BaseType_t coreid);
 
-static inline void __attribute__((always_inline)) vPortExitCriticalSafe(portMUX_TYPE *mux)
-{
-    if (xPortInIsrContext()) {
-        portEXIT_CRITICAL_ISR(mux);
-    } else {
-        portEXIT_CRITICAL(mux);
-    }
-}
+/**
+ * @brief Checks if the current core can yield
+ *
+ * - A core cannot yield if its in an ISR or in a critical section
+ *
+ * @note [refactor-todo] See if this can be separated from port macro
+ * @return true Core can yield
+ * @return false Core cannot yield
+ */
+static inline bool IRAM_ATTR xPortCanYield(void);
 
-#define portENTER_CRITICAL_SAFE(mux)  vPortEnterCriticalSafe(mux)
-#define portEXIT_CRITICAL_SAFE(mux)  vPortExitCriticalSafe(mux)
+// ------------------- Hook Functions ----------------------
 
-/*
- * Wrapper for the Xtensa compare-and-set instruction. This subroutine will atomically compare
- * *addr to 'compare'. If *addr == compare, *addr is set to *set. *set is updated with the previous
- * value of *addr (either 'compare' or some other value.)
+extern void esp_vApplicationIdleHook(void);     /* Required by tasks.c */
+extern void esp_vApplicationTickHook(void);     /* Required by tasks.c */
+
+/**
+ * @brief Hook function called on entry to tickless idle
+ *
+ * - Implemented in pm_impl.c
  *
- * Warning: From the ISA docs: in some (unspecified) cases, the s32c1i instruction may return the
- * *bitwise inverse* of the old mem if the mem wasn't written. This doesn't seem to happen on the
- * ESP32 (portMUX assertions would fail).
+ * @param xExpectedIdleTime Expected idle time
  */
-static inline void __attribute__((always_inline)) uxPortCompareSet(volatile uint32_t *addr, uint32_t compare, uint32_t *set) {
-    compare_and_set_native(addr, compare, set);
-}
+void vApplicationSleep(TickType_t xExpectedIdleTime);
 
-// Critical section management. NW-TODO: replace XTOS_SET_INTLEVEL with more efficient version, if any?
-// These cannot be nested. They should be used with a lot of care and cannot be called from interrupt level.
-//
-// Only applies to one CPU. See notes above & below for reasons not to use these.
-#define portDISABLE_INTERRUPTS()      do { XTOS_SET_INTLEVEL(XCHAL_EXCM_LEVEL); portbenchmarkINTERRUPT_DISABLE(); } while (0)
-#define portENABLE_INTERRUPTS()       do { portbenchmarkINTERRUPT_RESTORE(0); XTOS_SET_INTLEVEL(0); } while (0)
+// ----------------------- System --------------------------
 
+/**
+ * @brief Get the tick rate per second
+ *
+ * @note [refactor-todo] make this inline
+ * @return uint32_t Tick rate in Hz
+ */
+uint32_t xPortGetTickRateHz(void);
 
-// These FreeRTOS versions are similar to the nested versions above
-#define portSET_INTERRUPT_MASK_FROM_ISR()            portENTER_CRITICAL_NESTED()
-#define portCLEAR_INTERRUPT_MASK_FROM_ISR(state)     portEXIT_CRITICAL_NESTED(state)
+/**
+ * @brief Set a watchpoint to watch the last 32 bytes of the stack
+ *
+ * Callback to set a watchpoint on the end of the stack. Called every context switch to change the stack watchpoint
+ * around.
+ *
+ * @param pxStackStart Pointer to the start of the stack
+ */
+void vPortSetStackWatchpoint( void *pxStackStart );
 
-//Because the ROM routines don't necessarily handle a stack in external RAM correctly, we force
-//the stack memory to always be internal.
-#define portTcbMemoryCaps (MALLOC_CAP_INTERNAL|MALLOC_CAP_8BIT)
-#define portStackMemoryCaps (MALLOC_CAP_INTERNAL|MALLOC_CAP_8BIT)
-#define pvPortMallocTcbMem(size) heap_caps_malloc(size, portTcbMemoryCaps)
-#define pvPortMallocStackMem(size)  heap_caps_malloc(size, portStackMemoryCaps)
-
-//xTaskCreateStatic uses these functions to check incoming memory.
-#define portVALID_TCB_MEM(ptr) (esp_ptr_internal(ptr) && esp_ptr_byte_accessible(ptr))
-#ifdef CONFIG_SPIRAM_ALLOW_STACK_EXTERNAL_MEMORY
-#define portVALID_STACK_MEM(ptr) esp_ptr_byte_accessible(ptr)
-#else
-#define portVALID_STACK_MEM(ptr) (esp_ptr_internal(ptr) && esp_ptr_byte_accessible(ptr))
-#endif
+/**
+ * @brief Get the current core's ID
+ *
+ * @note [refactor-todo] IDF should call a FreeRTOS like macro instead of port function directly
+ * @return BaseType_t Core ID
+ */
+static inline BaseType_t IRAM_ATTR xPortGetCoreID(void);
 
 
-static inline void uxPortCompareSetExtram(volatile uint32_t *addr, uint32_t compare, uint32_t *set)
-{
-#ifdef CONFIG_SPIRAM
-    compare_and_set_extram(addr, compare, set);
-#endif
-}
 
+/* ------------------------------------------- FreeRTOS Porting Interface ----------------------------------------------
+ * - Contains all the mappings of the macros required by FreeRTOS
+ * - Most come after forward declare as porting macros map to declared functions
+ * - Maps to forward declared functions
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-/*-----------------------------------------------------------*/
+// ----------------------- Memory --------------------------
 
-/* Architecture specifics. */
-#define portSTACK_GROWTH			( -1 )
-#define portTICK_PERIOD_MS			( ( TickType_t ) 1000 / configTICK_RATE_HZ )
-#define portBYTE_ALIGNMENT			4
-#define portNOP()					XT_NOP()
-/*-----------------------------------------------------------*/
+/**
+ * @brief Task memory allocation macros
+ *
+ * @note Because the ROM routines don't necessarily handle a stack in external RAM correctly, we force the stack
+ * memory to always be internal.
+ * @note [refactor-todo] Update portable.h to match v10.4.3 to use new malloc prototypes
+ */
+#define portTcbMemoryCaps               (MALLOC_CAP_INTERNAL|MALLOC_CAP_8BIT)
+#define portStackMemoryCaps             (MALLOC_CAP_INTERNAL|MALLOC_CAP_8BIT)
+#define pvPortMallocTcbMem(size)        heap_caps_malloc(size, portTcbMemoryCaps)
+#define pvPortMallocStackMem(size)      heap_caps_malloc(size, portStackMemoryCaps)
 
-/* Fine resolution time */
-#define portGET_RUN_TIME_COUNTER_VALUE()  xthal_get_ccount()
-//ccount or esp_timer are initialized elsewhere
-#define portCONFIGURE_TIMER_FOR_RUN_TIME_STATS()
+// --------------------- Interrupts ------------------------
 
-#ifdef CONFIG_FREERTOS_RUN_TIME_STATS_USING_ESP_TIMER
-/* Coarse resolution time (us) */
-#define portALT_GET_RUN_TIME_COUNTER_VALUE(x)     do {x = (uint32_t)esp_timer_get_time();} while(0)
-#endif
+#define portEXIT_CRITICAL_NESTED(state)     do { portbenchmarkINTERRUPT_RESTORE(state); XTOS_RESTORE_JUST_INTLEVEL(state); } while (0)
 
-void vPortYield( void );
-void vPortEvaluateYieldFromISR(int argc, ...);
-void _frxt_setup_switch( void );
+/**
+ * - Only applies to current core
+ * - These cannot be nested. They should be used with a lot of care and cannot be called from interrupt level.
+ *
+ * @note [refactor-todo] replace XTOS_SET_INTLEVEL with more efficient version, if any?
+ */
+#define portDISABLE_INTERRUPTS()            do { XTOS_SET_INTLEVEL(XCHAL_EXCM_LEVEL); portbenchmarkINTERRUPT_DISABLE(); } while (0)
+#define portENABLE_INTERRUPTS()             do { portbenchmarkINTERRUPT_RESTORE(0); XTOS_SET_INTLEVEL(0); } while (0)
 
 /**
- * Macro to count number of arguments of a __VA_ARGS__ used to support portYIELD_FROM_ISR with,
- * or without arguments. The macro counts only 0 or 1 arguments.
+ * ISR versions to enable/disable interrupts
+ */
+#define portSET_INTERRUPT_MASK_FROM_ISR()            portENTER_CRITICAL_NESTED()
+#define portCLEAR_INTERRUPT_MASK_FROM_ISR(state)     portEXIT_CRITICAL_NESTED(state)
+
+#define portASSERT_IF_IN_ISR() vPortAssertIfInISR()
+
+// ------------------ Critical Sections --------------------
+
+/**
+ * @brief FreeRTOS critical section macros
  *
- * In the future, we want to switch to C++20. We also want to become compatible with clang.
- * Hence, we provide two versions of the following macros which are using variadic arguments.
- * The first one is using the GNU extension ##__VA_ARGS__. The second one is using the C++20 feature __VA_OPT__(,).
- * This allows users to compile their code with standard C++20 enabled instead of the GNU extension.
- * Below C++20, we haven't found any good alternative to using ##__VA_ARGS__.
+ * - Added a spinlock argument for SMP
+ * - Can be nested
+ * - Compliance versions will assert if regular critical section API is used in ISR context
+ * - Safe versions can be called from either contexts
  */
-#if defined(__cplusplus) && (__cplusplus >  201703L)
-#define portGET_ARGUMENT_COUNT(...) portGET_ARGUMENT_COUNT_INNER(0 __VA_OPT__(,) __VA_ARGS__,1,0)
+#ifdef CONFIG_FREERTOS_CHECK_PORT_CRITICAL_COMPLIANCE
+#define portENTER_CRITICAL(mux)         vPortEnterCriticalCompliance(mux)
+#define portEXIT_CRITICAL(mux)          vPortExitCriticalCompliance(mux)
 #else
-#define portGET_ARGUMENT_COUNT(...) portGET_ARGUMENT_COUNT_INNER(0, ##__VA_ARGS__,1,0)
-#endif
-#define portGET_ARGUMENT_COUNT_INNER(zero, one, count, ...) count
+#define portENTER_CRITICAL(mux)         vPortEnterCritical(mux)
+#define portEXIT_CRITICAL(mux)          vPortExitCritical(mux)
+#endif /* CONFIG_FREERTOS_CHECK_PORT_CRITICAL_COMPLIANCE */
+#define portENTER_CRITICAL_ISR(mux)     vPortEnterCritical(mux)
+#define portEXIT_CRITICAL_ISR(mux)      vPortExitCritical(mux)
+#define portENTER_CRITICAL_SAFE(mux)    vPortEnterCriticalSafe(mux)
+#define portEXIT_CRITICAL_SAFE(mux)     vPortExitCriticalSafe(mux)
 
-_Static_assert(portGET_ARGUMENT_COUNT() == 0, "portGET_ARGUMENT_COUNT() result does not match for 0 arguments");
-_Static_assert(portGET_ARGUMENT_COUNT(1) == 1, "portGET_ARGUMENT_COUNT() result does not match for 1 argument");
+// ---------------------- Yielding -------------------------
 
-#define portYIELD()	vPortYield()
+#define portYIELD() vPortYield()
 
 /**
  * @note    The macro below could be used when passing a single argument, or without any argument,
  *          it was developed to support both usages of portYIELD inside of an ISR. Any other usage form
- *          might result in undesired behaviour
+ *          might result in undesired behavior
+ *
+ * @note [refactor-todo] Refactor this to avoid va_args
  */
 #if defined(__cplusplus) && (__cplusplus >  201703L)
 #define portYIELD_FROM_ISR(...) vPortEvaluateYieldFromISR(portGET_ARGUMENT_COUNT(__VA_ARGS__) __VA_OPT__(,) __VA_ARGS__)
@@ -351,121 +483,137 @@ _Static_assert(portGET_ARGUMENT_COUNT(1) == 1, "portGET_ARGUMENT_COUNT() result
 */
 #define portYIELD_WITHIN_API() esp_crosscore_int_send_yield(xPortGetCoreID())
 
-/*-----------------------------------------------------------*/
-
-/* Task function macros as described on the FreeRTOS.org WEB site. */
-#define portTASK_FUNCTION_PROTO( vFunction, pvParameters ) void vFunction( void *pvParameters )
-#define portTASK_FUNCTION( vFunction, pvParameters ) void vFunction( void *pvParameters )
-
-// When coprocessors are defined, we to maintain a pointer to coprocessors area.
-// We currently use a hack: redefine field xMPU_SETTINGS in TCB block as a structure that can hold:
-// MPU wrappers, coprocessor area pointer, trace code structure, and more if needed.
-// The field is normally used for memory protection. FreeRTOS should create another general purpose field.
-typedef struct {
-	#if XCHAL_CP_NUM > 0
-	volatile StackType_t* coproc_area; // Pointer to coprocessor save area; MUST BE FIRST
-	#endif
-
-	#if portUSING_MPU_WRAPPERS
-	// Define here mpu_settings, which is port dependent
-	int mpu_setting; // Just a dummy example here; MPU not ported to Xtensa yet
-	#endif
-
-	#if configUSE_TRACE_FACILITY_2
-	struct {
-		// Cf. porttraceStamp()
-		int taskstamp;        /* Stamp from inside task to see where we are */
-		int taskstampcount;   /* A counter usually incremented when we restart the task's loop */
-	} porttrace;
-	#endif
-} xMPU_SETTINGS;
-
-// Main hack to use MPU_wrappers even when no MPU is defined (warning: mpu_setting should not be accessed; otherwise move this above xMPU_SETTINGS)
-#if (XCHAL_CP_NUM > 0 || configUSE_TRACE_FACILITY_2) && !portUSING_MPU_WRAPPERS   // If MPU wrappers not used, we still need to allocate coproc area
-	#undef portUSING_MPU_WRAPPERS
-	#define portUSING_MPU_WRAPPERS 1   // Enable it to allocate coproc area
-	#define MPU_WRAPPERS_H             // Override mpu_wrapper.h to disable unwanted code
-	#define PRIVILEGED_FUNCTION
-	#define PRIVILEGED_DATA
-#endif
-
-extern void esp_vApplicationIdleHook( void );
-extern void esp_vApplicationTickHook( void );
+// ------------------- Hook Functions ----------------------
 
 #ifndef CONFIG_FREERTOS_LEGACY_HOOKS
-#define vApplicationIdleHook    esp_vApplicationIdleHook
-#define vApplicationTickHook    esp_vApplicationTickHook
+#define vApplicationIdleHook esp_vApplicationIdleHook
+#define vApplicationTickHook esp_vApplicationTickHook
 #endif /* !CONFIG_FREERTOS_LEGACY_HOOKS */
 
-void vApplicationSleep( TickType_t xExpectedIdleTime );
+#define portSUPPRESS_TICKS_AND_SLEEP(idleTime) vApplicationSleep(idleTime)
 
-#define portSUPPRESS_TICKS_AND_SLEEP( idleTime ) vApplicationSleep( idleTime )
+// ------------------- Run Time Stats ----------------------
 
-void _xt_coproc_release(volatile void * coproc_sa_base);
+#define portCONFIGURE_TIMER_FOR_RUN_TIME_STATS()
 
-/* Architecture specific optimisations. */
-#if configUSE_PORT_OPTIMISED_TASK_SELECTION == 1
+/**
+ * - Fine resolution uses ccount
+ * - ALT is coarse and uses esp_timer
+ * @note [refactor-todo] Make fine and alt timers mutually exclusive
+ */
+#define portGET_RUN_TIME_COUNTER_VALUE() xthal_get_ccount()
+#ifdef CONFIG_FREERTOS_RUN_TIME_STATS_USING_ESP_TIMER
+#define portALT_GET_RUN_TIME_COUNTER_VALUE(x) do {x = (uint32_t)esp_timer_get_time();} while(0)
+#endif
 
+// -------------- Optimized Task Selection -----------------
+
+#if configUSE_PORT_OPTIMISED_TASK_SELECTION == 1
 /* Check the configuration. */
 #if( configMAX_PRIORITIES > 32 )
-    #error configUSE_PORT_OPTIMISED_TASK_SELECTION can only be set to 1 when configMAX_PRIORITIES is less than or equal to 32.  It is very rare that a system requires more than 10 to 15 different priorities as tasks that share a priority will time slice.
+#error configUSE_PORT_OPTIMISED_TASK_SELECTION can only be set to 1 when configMAX_PRIORITIES is less than or equal to 32.  It is very rare that a system requires more than 10 to 15 different priorities as tasks that share a priority will time slice.
 #endif
 
 /* Store/clear the ready priorities in a bit map. */
 #define portRECORD_READY_PRIORITY( uxPriority, uxReadyPriorities ) ( uxReadyPriorities ) |= ( 1UL << ( uxPriority ) )
 #define portRESET_READY_PRIORITY( uxPriority, uxReadyPriorities ) ( uxReadyPriorities ) &= ~( 1UL << ( uxPriority ) )
+#define portGET_HIGHEST_PRIORITY( uxTopPriority, uxReadyPriorities ) uxTopPriority = ( 31 - __builtin_clz( ( uxReadyPriorities ) ) )
+#endif /* configUSE_PORT_OPTIMISED_TASK_SELECTION */
 
-/*-----------------------------------------------------------*/
 
-#define portGET_HIGHEST_PRIORITY( uxTopPriority, uxReadyPriorities ) uxTopPriority = ( 31 - __builtin_clz( ( uxReadyPriorities ) ) )
 
-#endif /* configUSE_PORT_OPTIMISED_TASK_SELECTION */
+/* --------------------------------------------- Inline Implementations ------------------------------------------------
+ * - Implementation of inline functions of the forward declares
+ * - Should come after forward declare and FreeRTOS Porting interface, as implementation may use both.
+ * - For implementation of non-inlined functions, see port.c
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-/*
- * Send an interrupt to another core in order to make the task running
- * on it yield for a higher-priority task.
- */
+// --------------------- Interrupts ------------------------
 
-void vPortYieldOtherCore( BaseType_t coreid) ;
+static inline unsigned portENTER_CRITICAL_NESTED(void)
+{
+    unsigned state = XTOS_SET_INTLEVEL(XCHAL_EXCM_LEVEL);
+    portbenchmarkINTERRUPT_DISABLE();
+    return state;
+}
 
-/*
- Callback to set a watchpoint on the end of the stack. Called every context switch to change the stack
- watchpoint around.
- */
-void vPortSetStackWatchpoint( void* pxStackStart );
+// ---------------------- Spinlocks ------------------------
 
-/*
- * Returns true if the current core is in ISR context; low prio ISR, med prio ISR or timer tick ISR. High prio ISRs
- * aren't detected here, but they normally cannot call C code, so that should not be an issue anyway.
- */
-BaseType_t xPortInIsrContext(void);
+static inline void __attribute__((always_inline)) vPortCPUInitializeMutex(portMUX_TYPE *mux)
+{
+    spinlock_initialize(mux);
+}
 
-/*
- * This function will be called in High prio ISRs. Returns true if the current core was in ISR context
- * before calling into high prio ISR context.
- */
-BaseType_t xPortInterruptedFromISRContext(void);
+static inline void __attribute__((always_inline)) vPortCPUAcquireMutex(portMUX_TYPE *mux)
+{
+    spinlock_acquire(mux, portMUX_NO_TIMEOUT);
+}
 
-/*
- * The structures and methods of manipulating the MPU are contained within the
- * port layer.
- *
- * Fills the xMPUSettings structure with the memory region information
- * contained in xRegions.
- */
-#if( portUSING_MPU_WRAPPERS == 1 )
-    struct xMEMORY_REGION;
-    void vPortStoreTaskMPUSettings( xMPU_SETTINGS *xMPUSettings, const struct xMEMORY_REGION * const xRegions, StackType_t *pxBottomOfStack, uint32_t usStackDepth ) PRIVILEGED_FUNCTION;
-    void vPortReleaseTaskMPUSettings( xMPU_SETTINGS *xMPUSettings );
+static inline bool __attribute__((always_inline)) vPortCPUAcquireMutexTimeout(portMUX_TYPE *mux, int timeout)
+{
+    return (spinlock_acquire(mux, timeout));
+}
+
+static inline void __attribute__((always_inline)) vPortCPUReleaseMutex(portMUX_TYPE *mux)
+{
+    spinlock_release(mux);
+}
+
+static inline void __attribute__((always_inline)) uxPortCompareSet(volatile uint32_t *addr, uint32_t compare, uint32_t *set)
+{
+    compare_and_set_native(addr, compare, set);
+}
+
+static inline void uxPortCompareSetExtram(volatile uint32_t *addr, uint32_t compare, uint32_t *set)
+{
+#ifdef CONFIG_SPIRAM
+    compare_and_set_extram(addr, compare, set);
 #endif
+}
 
-/* Multi-core: get current core ID */
-static inline BaseType_t IRAM_ATTR xPortGetCoreID(void) {
-    return cpu_hal_get_core_id();
+// ------------------ Critical Sections --------------------
+
+static inline void __attribute__((always_inline)) vPortEnterCriticalCompliance(portMUX_TYPE *mux)
+{
+    if (!xPortInIsrContext()) {
+        vPortEnterCritical(mux);
+    } else {
+        esp_rom_printf("%s:%d (%s)- port*_CRITICAL called from ISR context!\n",
+                       __FILE__, __LINE__, __FUNCTION__);
+        abort();
+    }
 }
 
-/* Get tick rate per second */
-uint32_t xPortGetTickRateHz(void);
+static inline void __attribute__((always_inline)) vPortExitCriticalCompliance(portMUX_TYPE *mux)
+{
+    if (!xPortInIsrContext()) {
+        vPortExitCritical(mux);
+    } else {
+        esp_rom_printf("%s:%d (%s)- port*_CRITICAL called from ISR context!\n",
+                       __FILE__, __LINE__, __FUNCTION__);
+        abort();
+    }
+}
+
+static inline void __attribute__((always_inline)) vPortEnterCriticalSafe(portMUX_TYPE *mux)
+{
+    if (xPortInIsrContext()) {
+        portENTER_CRITICAL_ISR(mux);
+    } else {
+        portENTER_CRITICAL(mux);
+    }
+}
+
+static inline void __attribute__((always_inline)) vPortExitCriticalSafe(portMUX_TYPE *mux)
+{
+    if (xPortInIsrContext()) {
+        portEXIT_CRITICAL_ISR(mux);
+    } else {
+        portEXIT_CRITICAL(mux);
+    }
+}
+
+// ---------------------- Yielding -------------------------
 
 static inline bool IRAM_ATTR xPortCanYield(void)
 {
@@ -484,22 +632,115 @@ static inline bool IRAM_ATTR xPortCanYield(void)
     return ((ps_reg & PS_INTLEVEL_MASK) == 0);
 }
 
-// porttrace
-#if configUSE_TRACE_FACILITY_2
-#include "porttrace.h"
+// ----------------------- System --------------------------
+
+static inline BaseType_t IRAM_ATTR xPortGetCoreID(void)
+{
+    return (uint32_t) cpu_hal_get_core_id();
+}
+
+
+
+/* ------------------------------------------------------ Misc ---------------------------------------------------------
+ * - Miscellaneous porting macros
+ * - These are not port of the FreeRTOS porting interface, but are used by other FreeRTOS dependent components
+ * - [refactor-todo] Remove dependency on MPU wrappers by modifying TCB
+ * ------------------------------------------------------------------------------------------------------------------ */
+
+// -------------------- Co-Processor -----------------------
+
+// When coprocessors are defined, we maintain a pointer to coprocessors area.
+// We currently use a hack: redefine field xMPU_SETTINGS in TCB block as a structure that can hold:
+// MPU wrappers, coprocessor area pointer, trace code structure, and more if needed.
+// The field is normally used for memory protection. FreeRTOS should create another general purpose field.
+typedef struct {
+#if XCHAL_CP_NUM > 0
+    volatile StackType_t *coproc_area; // Pointer to coprocessor save area; MUST BE FIRST
+#endif
+
+#if portUSING_MPU_WRAPPERS
+    // Define here mpu_settings, which is port dependent
+    int mpu_setting; // Just a dummy example here; MPU not ported to Xtensa yet
 #endif
+} xMPU_SETTINGS;
 
-// configASSERT_2 if requested
-#if configASSERT_2
-#include <stdio.h>
-void exit(int);
-#define configASSERT( x )   if (!(x)) { porttracePrint(-1); printf("\nAssertion failed in %s:%d\n", __FILE__, __LINE__); exit(-1); }
+// Main hack to use MPU_wrappers even when no MPU is defined (warning: mpu_setting should not be accessed; otherwise move this above xMPU_SETTINGS)
+#if (XCHAL_CP_NUM > 0) && !portUSING_MPU_WRAPPERS   // If MPU wrappers not used, we still need to allocate coproc area
+#undef portUSING_MPU_WRAPPERS
+#define portUSING_MPU_WRAPPERS 1   // Enable it to allocate coproc area
+#define MPU_WRAPPERS_H             // Override mpu_wrapper.h to disable unwanted code
+#define PRIVILEGED_FUNCTION
+#define PRIVILEGED_DATA
 #endif
 
-#endif // __ASSEMBLER__
+void _xt_coproc_release(volatile void *coproc_sa_base);
+
+/*
+ * The structures and methods of manipulating the MPU are contained within the
+ * port layer.
+ *
+ * Fills the xMPUSettings structure with the memory region information
+ * contained in xRegions.
+ */
+#if( portUSING_MPU_WRAPPERS == 1 )
+struct xMEMORY_REGION;
+void vPortStoreTaskMPUSettings( xMPU_SETTINGS *xMPUSettings, const struct xMEMORY_REGION *const xRegions, StackType_t *pxBottomOfStack, uint32_t usStackDepth ) PRIVILEGED_FUNCTION;
+void vPortReleaseTaskMPUSettings( xMPU_SETTINGS *xMPUSettings );
+#endif
+
+// -------------------- VA_ARGS Yield ----------------------
+
+/**
+ * Macro to count number of arguments of a __VA_ARGS__ used to support portYIELD_FROM_ISR with,
+ * or without arguments. The macro counts only 0 or 1 arguments.
+ *
+ * In the future, we want to switch to C++20. We also want to become compatible with clang.
+ * Hence, we provide two versions of the following macros which are using variadic arguments.
+ * The first one is using the GNU extension ##__VA_ARGS__. The second one is using the C++20 feature __VA_OPT__(,).
+ * This allows users to compile their code with standard C++20 enabled instead of the GNU extension.
+ * Below C++20, we haven't found any good alternative to using ##__VA_ARGS__.
+ */
+#if defined(__cplusplus) && (__cplusplus >  201703L)
+#define portGET_ARGUMENT_COUNT(...) portGET_ARGUMENT_COUNT_INNER(0 __VA_OPT__(,) __VA_ARGS__,1,0)
+#else
+#define portGET_ARGUMENT_COUNT(...) portGET_ARGUMENT_COUNT_INNER(0, ##__VA_ARGS__,1,0)
+#endif
+#define portGET_ARGUMENT_COUNT_INNER(zero, one, count, ...) count
+
+_Static_assert(portGET_ARGUMENT_COUNT() == 0, "portGET_ARGUMENT_COUNT() result does not match for 0 arguments");
+_Static_assert(portGET_ARGUMENT_COUNT(1) == 1, "portGET_ARGUMENT_COUNT() result does not match for 1 argument");
+
+// -------------------- Heap Related -----------------------
+
+/**
+ * @brief Checks if a given piece of memory can be used to store a task's TCB
+ *
+ * - Defined in port_common.c
+ *
+ * @param ptr Pointer to memory
+ * @return true Memory can be used to store a TCB
+ * @return false Otherwise
+ */
+bool xPortCheckValidTCBMem(const void *ptr);
+
+/**
+ * @brief Checks if a given piece of memory can be used to store a task's stack
+ *
+ * - Defined in port_common.c
+ *
+ * @param ptr Pointer to memory
+ * @return true Memory can be used to store a task stack
+ * @return false Otherwise
+ */
+bool xPortcheckValidStackMem(const void *ptr);
+
+#define portVALID_TCB_MEM(ptr) xPortCheckValidTCBMem(ptr)
+#define portVALID_STACK_MEM(ptr) xPortcheckValidStackMem(ptr)
 
 #ifdef __cplusplus
 }
 #endif
 
+#endif // __ASSEMBLER__
+
 #endif /* PORTMACRO_H */
diff --git a/tools/sdk/esp32/include/freertos/port/xtensa/include/freertos/portmacro_priv.h b/tools/sdk/esp32/include/freertos/port/xtensa/include/freertos/portmacro_priv.h
deleted file mode 100644
index 843456b5aa8..00000000000
--- a/tools/sdk/esp32/include/freertos/port/xtensa/include/freertos/portmacro_priv.h
+++ /dev/null
@@ -1,78 +0,0 @@
-/*
-    FreeRTOS V8.2.0 - Copyright (C) 2015 Real Time Engineers Ltd.
-    All rights reserved
-
-    VISIT http://www.FreeRTOS.org TO ENSURE YOU ARE USING THE LATEST VERSION.
-
-    ***************************************************************************
-     *                                                                       *
-     *    FreeRTOS provides completely free yet professionally developed,    *
-     *    robust, strictly quality controlled, supported, and cross          *
-     *    platform software that has become a de facto standard.             *
-     *                                                                       *
-     *    Help yourself get started quickly and support the FreeRTOS         *
-     *    project by purchasing a FreeRTOS tutorial book, reference          *
-     *    manual, or both from: http://www.FreeRTOS.org/Documentation        *
-     *                                                                       *
-     *    Thank you!                                                         *
-     *                                                                       *
-    ***************************************************************************
-
-    This file is part of the FreeRTOS distribution.
-
-    FreeRTOS is free software; you can redistribute it and/or modify it under
-    the terms of the GNU General Public License (version 2) as published by the
-    Free Software Foundation >>!AND MODIFIED BY!<< the FreeRTOS exception.
-
-    >>!   NOTE: The modification to the GPL is included to allow you to     !<<
-    >>!   distribute a combined work that includes FreeRTOS without being   !<<
-    >>!   obliged to provide the source code for proprietary components     !<<
-    >>!   outside of the FreeRTOS kernel.                                   !<<
-
-    FreeRTOS is distributed in the hope that it will be useful, but WITHOUT ANY
-    WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
-    FOR A PARTICULAR PURPOSE.  Full license text is available from the following
-    link: http://www.freertos.org/a00114.html
-
-    1 tab == 4 spaces!
-
-    ***************************************************************************
-     *                                                                       *
-     *    Having a problem?  Start by reading the FAQ "My application does   *
-     *    not run, what could be wrong?"                                     *
-     *                                                                       *
-     *    http://www.FreeRTOS.org/FAQHelp.html                               *
-     *                                                                       *
-    ***************************************************************************
-
-    http://www.FreeRTOS.org - Documentation, books, training, latest versions,
-    license and Real Time Engineers Ltd. contact details.
-
-    http://www.FreeRTOS.org/plus - A selection of FreeRTOS ecosystem products,
-    including FreeRTOS+Trace - an indispensable productivity tool, a DOS
-    compatible FAT file system, and our tiny thread aware UDP/IP stack.
-
-    http://www.OpenRTOS.com - Real Time Engineers ltd license FreeRTOS to High
-    Integrity Systems to sell under the OpenRTOS brand.  Low cost OpenRTOS
-    licenses offer ticketed support, indemnification and middleware.
-
-    http://www.SafeRTOS.com - High Integrity Systems also provide a safety
-    engineered and independently SIL3 certified version for use in safety and
-    mission critical applications that require provable dependability.
-
-    1 tab == 4 spaces!
-*/
-
-
-/* This header holds the macros for porting which should only be used inside FreeRTOS */
-
-#pragma once
-#include "soc/soc_memory_layout.h"
-
-//xTaskCreateStatic uses these functions to check incoming memory.
-#define portVALID_TCB_MEM(ptr) (esp_ptr_internal(ptr) && esp_ptr_byte_accessible(ptr))
-#ifdef CONFIG_SPIRAM_ALLOW_STACK_EXTERNAL_MEMORY
-#define portVALID_STACK_MEM(ptr) esp_ptr_byte_accessible(ptr)
-#else
-#define portVALID_STACK_MEM(ptr) (esp_ptr_internal(ptr) && esp_ptr_byte_accessible(ptr))
-#endif
diff --git a/tools/sdk/esp32/include/hal/include/hal/i2s_hal.h b/tools/sdk/esp32/include/hal/include/hal/i2s_hal.h
index a08813db808..037970fa2b3 100644
--- a/tools/sdk/esp32/include/hal/include/hal/i2s_hal.h
+++ b/tools/sdk/esp32/include/hal/include/hal/i2s_hal.h
@@ -1,16 +1,8 @@
-// Copyright 2020 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2020-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 /*******************************************************************************
  * NOTICE
@@ -176,28 +168,28 @@ void i2s_hal_enable_slave_fd_mode(i2s_hal_context_t *hal);
  *
  * @param hal Context of the HAL layer
  */
-#define i2s_hal_start_tx(hal)               i2s_ll_tx_start((hal)->dev)
+void i2s_hal_start_tx(i2s_hal_context_t *hal);
 
 /**
  * @brief Start I2S rx
  *
  * @param hal Context of the HAL layer
  */
-#define i2s_hal_start_rx(hal)               i2s_ll_rx_start((hal)->dev)
+void i2s_hal_start_rx(i2s_hal_context_t *hal);
 
 /**
  * @brief Stop I2S tx
  *
  * @param hal Context of the HAL layer
  */
-#define i2s_hal_stop_tx(hal)                i2s_ll_tx_stop((hal)->dev)
+void i2s_hal_stop_tx(i2s_hal_context_t *hal);
 
 /**
  * @brief Stop I2S rx
  *
  * @param hal Context of the HAL layer
  */
-#define i2s_hal_stop_rx(hal)                i2s_ll_rx_stop((hal)->dev)
+void i2s_hal_stop_rx(i2s_hal_context_t *hal);
 
 /**
  * @brief Set the received data length to trigger `in_suc_eof` interrupt.
diff --git a/tools/sdk/esp32/include/hal/include/hal/lcd_hal.h b/tools/sdk/esp32/include/hal/include/hal/lcd_hal.h
index 76e28dda0e4..db255b3d1e4 100644
--- a/tools/sdk/esp32/include/hal/include/hal/lcd_hal.h
+++ b/tools/sdk/esp32/include/hal/include/hal/lcd_hal.h
@@ -1,22 +1,8 @@
-// Copyright 2021 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
-
-/*******************************************************************************
- * NOTICE
- * The HAL is not public api, don't use in application code.
- * See readme.md in soc/README.md
- ******************************************************************************/
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
diff --git a/tools/sdk/esp32/include/hal/include/hal/lcd_types.h b/tools/sdk/esp32/include/hal/include/hal/lcd_types.h
index 69dab14801d..01e6d0c2949 100644
--- a/tools/sdk/esp32/include/hal/include/hal/lcd_types.h
+++ b/tools/sdk/esp32/include/hal/include/hal/lcd_types.h
@@ -1,16 +1,8 @@
-// Copyright 2021 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
@@ -21,15 +13,12 @@ extern "C" {
 /**
  * @brief LCD clock source
  * @note User should select the clock source based on the real requirement:
- * ╔═════════════════════╦══════════════════════════╦════════════════════════════╗
- * ║ LCD clock source    ║ Features                 ║ Power Management           ║
- * ╠═════════════════════╬══════════════════════════╬════════════════════════════╣
- * ║ LCD_CLK_SRC_PLL160M ║ High resolution, fixed   ║ ESP_PM_APB_FREQ_MAX lock   ║
- * ╠═════════════════════╬══════════════════════════╬════════════════════════════╣
- * ║ LCD_CLK_SRC_APLL    ║ Configurable resolution  ║ ESP_PM_NO_LIGHT_SLEEP lock ║
- * ╠═════════════════════╬══════════════════════════╬════════════════════════════╣
- * ║ LCD_CLK_SRC_XTAL    ║ Medium resolution, fixed ║ No PM lock                 ║
- * ╚═════════════════════╩══════════════════════════╩════════════════════════════╝
+ *
+ * | LCD clock source    | Features                 | Power Management           |
+ * |---------------------|--------------------------|----------------------------|
+ * | LCD_CLK_SRC_PLL160M | High resolution, fixed   | ESP_PM_APB_FREQ_MAX lock   |
+ * | LCD_CLK_SRC_APLL    | Configurable resolution  | ESP_PM_NO_LIGHT_SLEEP lock |
+ * | LCD_CLK_SRC_XTAL    | Medium resolution, fixed | No PM lock                 |
  */
 typedef enum {
     LCD_CLK_SRC_PLL160M, /*!< Select PLL160M as the source clock */
diff --git a/tools/sdk/esp32/include/hal/include/hal/touch_sensor_types.h b/tools/sdk/esp32/include/hal/include/hal/touch_sensor_types.h
index ec027bf8705..9085f5eecd8 100644
--- a/tools/sdk/esp32/include/hal/include/hal/touch_sensor_types.h
+++ b/tools/sdk/esp32/include/hal/include/hal/touch_sensor_types.h
@@ -1,16 +1,8 @@
-// Copyright 2015-2020 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
@@ -155,12 +147,23 @@ typedef enum {
     TOUCH_PAD_INTR_MASK_INACTIVE = BIT(2),  /*!<Inactive for one of the enabled channels. */
     TOUCH_PAD_INTR_MASK_SCAN_DONE = BIT(3), /*!<Measurement done for all the enabled channels. */
     TOUCH_PAD_INTR_MASK_TIMEOUT = BIT(4),   /*!<Timeout for one of the enabled channels. */
+#if SOC_TOUCH_PROXIMITY_MEAS_DONE_SUPPORTED
+    TOUCH_PAD_INTR_MASK_PROXI_MEAS_DONE = BIT(5),   /*!<For proximity sensor, when the number of measurements reaches the set count of measurements, an interrupt will be generated. */
+    TOUCH_PAD_INTR_MASK_MAX
+#define TOUCH_PAD_INTR_MASK_ALL (TOUCH_PAD_INTR_MASK_TIMEOUT    \
+                                | TOUCH_PAD_INTR_MASK_SCAN_DONE \
+                                | TOUCH_PAD_INTR_MASK_INACTIVE  \
+                                | TOUCH_PAD_INTR_MASK_ACTIVE    \
+                                | TOUCH_PAD_INTR_MASK_DONE      \
+                                | TOUCH_PAD_INTR_MASK_PROXI_MEAS_DONE) /*!<All touch interrupt type enable. */
+#else
     TOUCH_PAD_INTR_MASK_MAX
 #define TOUCH_PAD_INTR_MASK_ALL (TOUCH_PAD_INTR_MASK_TIMEOUT    \
                                 | TOUCH_PAD_INTR_MASK_SCAN_DONE \
                                 | TOUCH_PAD_INTR_MASK_INACTIVE  \
                                 | TOUCH_PAD_INTR_MASK_ACTIVE    \
                                 | TOUCH_PAD_INTR_MASK_DONE) /*!<All touch interrupt type enable. */
+#endif
 } touch_pad_intr_mask_t;
 FLAG_ATTR(touch_pad_intr_mask_t)
 
diff --git a/tools/sdk/esp32/include/lwip/port/esp32/include/lwipopts.h b/tools/sdk/esp32/include/lwip/port/esp32/include/lwipopts.h
index 4bc2ad13960..b9cfe17aca9 100644
--- a/tools/sdk/esp32/include/lwip/port/esp32/include/lwipopts.h
+++ b/tools/sdk/esp32/include/lwip/port/esp32/include/lwipopts.h
@@ -266,6 +266,12 @@
 
 #endif
 
+/**
+ * CONFIG_LWIP_DHCP_OPTIONS_LEN: The total length of outgoing DHCP option msg. If you have many options
+ * and options value is too long, you can configure the length according to your requirements
+ */
+#define DHCP_OPTIONS_LEN                CONFIG_LWIP_DHCP_OPTIONS_LEN
+
 /*
    ------------------------------------
    ---------- AUTOIP options ----------
diff --git a/tools/sdk/esp32/include/mdns/include/mdns.h b/tools/sdk/esp32/include/mdns/include/mdns.h
index 4bb288d4174..35c8d538abc 100644
--- a/tools/sdk/esp32/include/mdns/include/mdns.h
+++ b/tools/sdk/esp32/include/mdns/include/mdns.h
@@ -187,7 +187,10 @@ esp_err_t mdns_instance_name_set(const char * instance_name);
  * @note The value length of txt items will be automatically decided by strlen
  *
  * @param  instance_name    instance name to set. If NULL,
- *                          global instance name or hostname will be used
+ *                          global instance name or hostname will be used.
+ *                          Note that MDNS_MULTIPLE_INSTANCE config option
+ *                          needs to be enabled for adding multiple instances
+ *                          with the same instance type.
  * @param  service_type     service type (_http, _ftp, etc)
  * @param  proto            service protocol (_tcp, _udp)
  * @param  port             service port
@@ -209,6 +212,9 @@ esp_err_t mdns_service_add(const char * instance_name, const char * service_type
  *
  * @param  instance_name    instance name to set. If NULL,
  *                          global instance name or hostname will be used
+ *                          Note that MDNS_MULTIPLE_INSTANCE config option
+ *                          needs to be enabled for adding multiple instances
+ *                          with the same instance type.
  * @param  service_type     service type (_http, _ftp, etc)
  * @param  proto            service protocol (_tcp, _udp)
  * @param  hostname         service hostname. If NULL, local hostname will be used.
@@ -238,6 +244,22 @@ esp_err_t mdns_service_add_for_host(const char * instance_name, const char * ser
  */
 bool mdns_service_exists(const char * service_type, const char * proto, const char * hostname);
 
+
+/**
+ * @brief  Check whether a service has been added.
+ *
+ * @param  instance         instance name
+ * @param  service_type     service type (_http, _ftp, etc)
+ * @param  proto            service protocol (_tcp, _udp)
+ * @param  hostname         service hostname. If NULL, checks for the local hostname.
+ *
+ * @return
+ *     - true   Correspondding service has been added.
+ *     - false  Service not found.
+ */
+bool mdns_service_exists_with_instance(const char *instance, const char *service_type, const char *proto,
+                                       const char *hostname);
+
 /**
  * @brief  Remove service from mDNS server
  *
diff --git a/tools/sdk/esp32/include/mqtt/esp-mqtt/include/mqtt_client.h b/tools/sdk/esp32/include/mqtt/esp-mqtt/include/mqtt_client.h
index 348d0729f1d..475581b50db 100644
--- a/tools/sdk/esp32/include/mqtt/esp-mqtt/include/mqtt_client.h
+++ b/tools/sdk/esp32/include/mqtt/esp-mqtt/include/mqtt_client.h
@@ -206,6 +206,7 @@ typedef struct {
     int network_timeout_ms;                 /*!< Abort network operation if it is not completed after this value, in milliseconds (defaults to 10s) */
     bool disable_keepalive;                 /*!< Set disable_keepalive=true to turn off keep-alive mechanism, false by default (keepalive is active by default). Note: setting the config value `keepalive` to `0` doesn't disable keepalive feature, but uses a default keepalive period */
     const char *path;                       /*!< Path in the URI*/
+    int message_retransmit_timeout;         /*!< timeout for retansmit of failded packet */
 } esp_mqtt_client_config_t;
 
 /**
@@ -375,6 +376,7 @@ esp_err_t esp_mqtt_client_destroy(esp_mqtt_client_handle_t client);
  * @param config    mqtt configuration structure
  *
  * @return ESP_ERR_NO_MEM if failed to allocate
+ *         ESP_ERR_INVALID_ARG if conflicts on transport configuration.
  *         ESP_OK on success
  */
 esp_err_t esp_mqtt_set_config(esp_mqtt_client_handle_t client, const esp_mqtt_client_config_t *config);
diff --git a/tools/sdk/esp32/include/openssl/include/openssl/ssl.h b/tools/sdk/esp32/include/openssl/include/openssl/ssl.h
index 4a3376c0dbf..9a33e0d33b2 100644
--- a/tools/sdk/esp32/include/openssl/include/openssl/ssl.h
+++ b/tools/sdk/esp32/include/openssl/include/openssl/ssl.h
@@ -1,16 +1,13 @@
-// Copyright 2015-2016 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+// Allow for this warning suppression only in IDF_CI_BUILD
+#if !defined(ESP_OPENSSL_SUPPRESS_LEGACY_WARNING) || !defined(IDF_CI_BUILD)
+#warning "OpenSSL component will be removed from ESP-IDF in v5.0, please use esp_tls instead"
+#endif
 
 #ifndef _SSL_H_
 #define _SSL_H_
diff --git a/tools/sdk/esp32/include/protocomm/include/transports/protocomm_ble.h b/tools/sdk/esp32/include/protocomm/include/transports/protocomm_ble.h
index a9e7b6a209e..f899f95b188 100644
--- a/tools/sdk/esp32/include/protocomm/include/transports/protocomm_ble.h
+++ b/tools/sdk/esp32/include/protocomm/include/transports/protocomm_ble.h
@@ -1,16 +1,8 @@
-// Copyright 2018 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2018-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
@@ -86,6 +78,10 @@ typedef struct protocomm_ble_config {
      * Pointer to the Name-UUID lookup table
      */
     protocomm_ble_name_uuid_t *nu_lookup;
+
+    /* BLE bonding */
+    unsigned ble_bonding:1;
+
 } protocomm_ble_config_t;
 
 /**
diff --git a/tools/sdk/esp32/include/soc/esp32/include/soc/rtc.h b/tools/sdk/esp32/include/soc/esp32/include/soc/rtc.h
index 7e4a73a20fd..bbbf3c7550c 100644
--- a/tools/sdk/esp32/include/soc/esp32/include/soc/rtc.h
+++ b/tools/sdk/esp32/include/soc/esp32/include/soc/rtc.h
@@ -493,6 +493,7 @@ typedef struct rtc_sleep_config_s {
     uint32_t rtc_slowmem_pd_en : 1;     //!< power down RTC slow memory
     uint32_t rtc_peri_pd_en : 1;        //!< power down RTC peripherals
     uint32_t wifi_pd_en : 1;            //!< power down WiFi
+    uint32_t int_8m_pd_en : 1;          //!< Power down Internal 8M oscillator
     uint32_t rom_mem_pd_en : 1;         //!< power down main RAM and ROM
     uint32_t deep_slp : 1;              //!< power down digital domain
     uint32_t wdt_flashboot_mod_en : 1;  //!< enable WDT flashboot mode
@@ -513,6 +514,7 @@ typedef struct rtc_sleep_config_s {
  *
  * @param RTC_SLEEP_PD_x flags combined using bitwise OR
  */
+#define is_dslp(pd_flags)   ((pd_flags) & RTC_SLEEP_PD_DIG)
 #define RTC_SLEEP_CONFIG_DEFAULT(sleep_flags) { \
     .lslp_mem_inf_fpu = 0, \
     .rtc_mem_inf_fpu = 0, \
@@ -521,13 +523,18 @@ typedef struct rtc_sleep_config_s {
     .rtc_slowmem_pd_en = ((sleep_flags) & RTC_SLEEP_PD_RTC_SLOW_MEM) ? 1 : 0, \
     .rtc_peri_pd_en = ((sleep_flags) & RTC_SLEEP_PD_RTC_PERIPH) ? 1 : 0, \
     .wifi_pd_en = 0, \
+    .int_8m_pd_en = is_dslp(sleep_flags) ? 1 : ((sleep_flags) & RTC_SLEEP_PD_INT_8M) ? 1 : 0, \
     .rom_mem_pd_en = 0, \
     .deep_slp = ((sleep_flags) & RTC_SLEEP_PD_DIG) ? 1 : 0, \
     .wdt_flashboot_mod_en = 0, \
     .dig_dbias_wak = RTC_CNTL_DBIAS_1V10, \
-    .dig_dbias_slp = RTC_CNTL_DBIAS_0V90, \
+    .dig_dbias_slp = is_dslp(sleep_flags)                   ? RTC_CNTL_DBIAS_0V90 \
+                   : !((sleep_flags) & RTC_SLEEP_PD_INT_8M) ? RTC_CNTL_DBIAS_1V10 \
+                   : RTC_CNTL_DBIAS_0V90, \
     .rtc_dbias_wak = RTC_CNTL_DBIAS_1V10, \
-    .rtc_dbias_slp = RTC_CNTL_DBIAS_0V90, \
+    .rtc_dbias_slp = is_dslp(sleep_flags)                   ? RTC_CNTL_DBIAS_0V90 \
+                   : !((sleep_flags) & RTC_SLEEP_PD_INT_8M) ? RTC_CNTL_DBIAS_1V10 \
+                   : RTC_CNTL_DBIAS_0V90, \
     .lslp_meminf_pd = 1, \
     .vddsdio_pd_en = ((sleep_flags) & RTC_SLEEP_PD_VDDSDIO) ? 1 : 0, \
     .xtal_fpu = ((sleep_flags) & RTC_SLEEP_PD_XTAL) ? 0 : 1 \
@@ -540,6 +547,7 @@ typedef struct rtc_sleep_config_s {
 #define RTC_SLEEP_PD_RTC_MEM_FOLLOW_CPU BIT(4)  //!< RTC FAST and SLOW memories are automatically powered up and down along with the CPU
 #define RTC_SLEEP_PD_VDDSDIO            BIT(5)  //!< Power down VDDSDIO regulator
 #define RTC_SLEEP_PD_XTAL               BIT(6)  //!< Power down main XTAL
+#define RTC_SLEEP_PD_INT_8M             BIT(7)  //!< Power down Internal 8M oscillator
 
 /* Various delays to be programmed into power control state machines */
 #define RTC_CNTL_XTL_BUF_WAIT_SLP_US        (500)
diff --git a/tools/sdk/esp32/include/soc/esp32/include/soc/soc_caps.h b/tools/sdk/esp32/include/soc/esp32/include/soc/soc_caps.h
index f9eef1ce18e..88b6f1c8ee2 100644
--- a/tools/sdk/esp32/include/soc/esp32/include/soc/soc_caps.h
+++ b/tools/sdk/esp32/include/soc/esp32/include/soc/soc_caps.h
@@ -1,16 +1,8 @@
-// Copyright 2020 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 /*
  * Soc capabilities file, describing the following chip attributes:
@@ -239,11 +231,11 @@
 #define SOC_TIMER_GROUP_TOTAL_TIMERS (SOC_TIMER_GROUPS * SOC_TIMER_GROUP_TIMERS_PER_GROUP)
 
 /*-------------------------- TOUCH SENSOR CAPS -------------------------------*/
+#define SOC_TOUCH_VERSION_1                 (1)     /*!<Hardware version of touch sensor */
 #define SOC_TOUCH_SENSOR_NUM                (10)
 
 #define SOC_TOUCH_PAD_MEASURE_WAIT_MAX      (0xFF)  /*!<The timer frequency is 8Mhz, the max value is 0xff */
 #define SOC_TOUCH_PAD_THRESHOLD_MAX         (0)     /*!<If set touch threshold max value, The touch sensor can't be in touched status */
-#define SOC_TOUCH_PAD_WAKE_SUPPORTED        (1)     /*!<Supports waking up from touch pad trigger */
 
 /*-------------------------- TWAI CAPS ---------------------------------------*/
 #define SOC_TWAI_BRP_MIN                        2
@@ -295,6 +287,7 @@
 
 /*-------------------------- Power Management CAPS ---------------------------*/
 #define SOC_PM_SUPPORT_EXT_WAKEUP       (1)
+#define SOC_PM_SUPPORT_TOUCH_SENSOR_WAKEUP        (1)     /*!<Supports waking up from touch pad trigger */
 
 /* ---------------------------- Compatibility ------------------------------- */
 #define SOC_CAN_SUPPORTED                   SOC_TWAI_SUPPORTED
diff --git a/tools/sdk/esp32/include/soc/include/soc/lcd_periph.h b/tools/sdk/esp32/include/soc/include/soc/lcd_periph.h
index d7d3ebc68bf..7bb7c530cf2 100644
--- a/tools/sdk/esp32/include/soc/include/soc/lcd_periph.h
+++ b/tools/sdk/esp32/include/soc/include/soc/lcd_periph.h
@@ -1,16 +1,8 @@
-// Copyright 2021 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
diff --git a/tools/sdk/esp32/include/spi_flash/include/esp_partition.h b/tools/sdk/esp32/include/spi_flash/include/esp_partition.h
index 36baea29860..550bf6b6b26 100644
--- a/tools/sdk/esp32/include/spi_flash/include/esp_partition.h
+++ b/tools/sdk/esp32/include/spi_flash/include/esp_partition.h
@@ -181,7 +181,9 @@ esp_partition_iterator_t esp_partition_next(esp_partition_iterator_t iterator);
 /**
  * @brief Release partition iterator
  *
- * @param iterator Iterator obtained using esp_partition_find. Must be non-NULL.
+ * @param iterator Iterator obtained using esp_partition_find.
+ *                 The iterator is allowed to be NULL, so it is not necessary to check its value
+ *                 before calling this function.
  *
  */
 void esp_partition_iterator_release(esp_partition_iterator_t iterator);
diff --git a/tools/sdk/esp32s2/include/spi_flash/include/spi_flash_private.h b/tools/sdk/esp32/include/spi_flash/include/esp_private/spi_flash_os.h
similarity index 88%
rename from tools/sdk/esp32s2/include/spi_flash/include/spi_flash_private.h
rename to tools/sdk/esp32/include/spi_flash/include/esp_private/spi_flash_os.h
index 9d822fb9475..e302df86ea2 100644
--- a/tools/sdk/esp32s2/include/spi_flash/include/spi_flash_private.h
+++ b/tools/sdk/esp32/include/spi_flash/include/esp_private/spi_flash_os.h
@@ -4,7 +4,9 @@
  * SPDX-License-Identifier: Apache-2.0
  */
 
-
+/**
+ * System level MSPI APIs (private)
+ */
 /**
  * Currently the MSPI timing tuning related APIs are designed to be private.
  * Because:
@@ -17,6 +19,10 @@
  */
 #pragma once
 
+#include <stdint.h>
+#include <stdbool.h>
+#include "sdkconfig.h"
+#include "esp_err.h"
 #if CONFIG_IDF_TARGET_ESP32
 #include "esp32/rom/spi_flash.h"
 #elif CONFIG_IDF_TARGET_ESP32S2
@@ -34,9 +40,9 @@ extern "C" {
 #endif
 
 /**
- * @brief Register ROM functions and init flash device registers to make use of octal flash
+ * @brief To setup Flash chip
  */
-esp_err_t esp_opiflash_init(void);
+esp_err_t spi_flash_init_chip_state(void);
 
 /**
  * @brief Make MSPI work under 20Mhz
@@ -88,6 +94,12 @@ void spi_timing_get_flash_timing_param(spi_flash_hal_timing_config_t *out_timing
  */
 bool spi_timine_config_flash_is_tuned(void);
 
+/**
+ * @brief Set Flash chip specifically required MSPI register settings here
+ */
+void spi_flash_set_vendor_required_regs(void);
+
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/tools/sdk/esp32/include/wpa_supplicant/esp_supplicant/include/esp_mbo.h b/tools/sdk/esp32/include/wpa_supplicant/esp_supplicant/include/esp_mbo.h
new file mode 100644
index 00000000000..4292213943f
--- /dev/null
+++ b/tools/sdk/esp32/include/wpa_supplicant/esp_supplicant/include/esp_mbo.h
@@ -0,0 +1,64 @@
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+#ifndef _ESP_MBO_H
+#define _ESP_MBO_H
+
+#include <stdbool.h>
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/**
+  * enum non_pref_chan_reason: Reason for non preference of channel
+  */
+enum non_pref_chan_reason {
+	NON_PREF_CHAN_REASON_UNSPECIFIED = 0,
+	NON_PREF_CHAN_REASON_RSSI = 1,
+	NON_PREF_CHAN_REASON_EXT_INTERFERENCE = 2,
+	NON_PREF_CHAN_REASON_INT_INTERFERENCE = 3,
+};
+
+/**
+  * @brief  Channel structure for non preferred channel
+  *
+  * @param  reason: enum non_pref_chan_reason
+  * @param  oper_class: operating class for the channel
+  * @param  chan: channel number
+  * @param  preference: channel preference
+  */
+struct non_pref_chan {
+	enum non_pref_chan_reason reason;
+	uint8_t oper_class;
+	uint8_t chan;
+	uint8_t preference;
+};
+
+/**
+  * @brief  Array structure for non preferred channel struct
+  *
+  * @param  non_pref_chan_num: channel count
+  * @param  chan: array of non_pref_chan type
+  */
+struct non_pref_chan_s {
+	size_t non_pref_chan_num;
+	struct non_pref_chan chan[];
+};
+
+/**
+  * @brief  Update channel preference for MBO IE
+  *
+  * @param  non_pref_chan: Non preference channel list
+  *
+  * @return
+  *    - 0: success else failure
+  */
+int esp_mbo_update_non_pref_chan(struct non_pref_chan_s *non_pref_chan);
+
+#ifdef __cplusplus
+}
+#endif
+#endif
diff --git a/tools/sdk/esp32/include/wpa_supplicant/esp_supplicant/include/esp_wnm.h b/tools/sdk/esp32/include/wpa_supplicant/esp_supplicant/include/esp_wnm.h
index a1dcfb655c5..4301385d2cf 100644
--- a/tools/sdk/esp32/include/wpa_supplicant/esp_supplicant/include/esp_wnm.h
+++ b/tools/sdk/esp32/include/wpa_supplicant/esp_supplicant/include/esp_wnm.h
@@ -1,17 +1,7 @@
-/**
- * Copyright 2020 Espressif Systems (Shanghai) PTE LTD
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *   http://www.apache.org/licenses/LICENSE-2.0
+/*
+ * SPDX-FileCopyrightText: 2020-2021 Espressif Systems (Shanghai) CO LTD
  *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
+ * SPDX-License-Identifier: Apache-2.0
  */
 
 #ifndef _ESP_WNM_H
@@ -29,11 +19,13 @@ enum btm_query_reason {
 	REASON_UNSPECIFIED = 0,
 	REASON_FRAME_LOSS = 1,
 	REASON_DELAY = 2,
-	REASON_QOS_CAPACITY = 3,
-	REASON_FIRST_ASSOC = 4,
-	REASON_LOAD_BALALNCE = 5,
-	REASON_BETTER_AP = 6,
-	REASON_CURRENT_DEAUTH = 7,
+	REASON_BANDWIDTH = 3,
+	REASON_LOAD_BALANCE = 4,
+	REASON_RSSI = 5,
+	REASON_RETRANSMISSIONS = 6,
+	REASON_INTERFERENCE = 7,
+	REASON_GRAY_ZONE = 8,
+	REASON_PREMIUM_AP = 9,
 };
 
 /**
diff --git a/tools/sdk/esp32/ld/libcat_face_detect.a b/tools/sdk/esp32/ld/libcat_face_detect.a
index 18bff9da4e2..804fea7893f 100644
Binary files a/tools/sdk/esp32/ld/libcat_face_detect.a and b/tools/sdk/esp32/ld/libcat_face_detect.a differ
diff --git a/tools/sdk/esp32/ld/libdl.a b/tools/sdk/esp32/ld/libdl.a
index 8840177a68b..6917a349872 100644
Binary files a/tools/sdk/esp32/ld/libdl.a and b/tools/sdk/esp32/ld/libdl.a differ
diff --git a/tools/sdk/esp32/ld/libhuman_face_detect.a b/tools/sdk/esp32/ld/libhuman_face_detect.a
index 8a6930968aa..8656df19de8 100644
Binary files a/tools/sdk/esp32/ld/libhuman_face_detect.a and b/tools/sdk/esp32/ld/libhuman_face_detect.a differ
diff --git a/tools/sdk/esp32/ld/sections.ld b/tools/sdk/esp32/ld/sections.ld
index 85a3a35947d..8e26933d350 100644
--- a/tools/sdk/esp32/ld/sections.ld
+++ b/tools/sdk/esp32/ld/sections.ld
@@ -323,7 +323,7 @@ SECTIONS
     *libesp_system.a:ubsan.*(.literal .literal.* .text .text.*)
     *libfreertos.a:(EXCLUDE_FILE(*libfreertos.a:port.* *libfreertos.a:port_common.*) .literal EXCLUDE_FILE(*libfreertos.a:port.* *libfreertos.a:port_common.*) .literal.* EXCLUDE_FILE(*libfreertos.a:port.* *libfreertos.a:port_common.*) .text EXCLUDE_FILE(*libfreertos.a:port.* *libfreertos.a:port_common.*) .text.*)
     *libfreertos.a:port.*(.literal.pxPortInitialiseStack .literal.unlikely.vPortEndScheduler .literal.vApplicationStackOverflowHook .literal.vPortAssertIfInISR .literal.vPortEnterCritical .literal.vPortExitCritical .literal.vPortReleaseTaskMPUSettings .literal.vPortSetStackWatchpoint .literal.vPortYieldOtherCore .literal.xPortInIsrContext .literal.xPortStartScheduler .text .text.pxPortInitialiseStack .text.unlikely.vPortEndScheduler .text.vApplicationStackOverflowHook .text.vPortAssertIfInISR .text.vPortEnterCritical .text.vPortExitCritical .text.vPortReleaseTaskMPUSettings .text.vPortSetStackWatchpoint .text.vPortStoreTaskMPUSettings .text.vPortYieldOtherCore .text.xPortGetTickRateHz .text.xPortInIsrContext .text.xPortStartScheduler)
-    *libfreertos.a:port_common.*(.literal.esp_startup_start_app_common .text .text.esp_startup_start_app_common)
+    *libfreertos.a:port_common.*(.literal.esp_startup_start_app_common .literal.xPortCheckValidTCBMem .literal.xPortcheckValidStackMem .text .text.esp_startup_start_app_common .text.xPortCheckValidTCBMem .text.xPortcheckValidStackMem)
     *libgcc.a:lib2funcs.*(.literal .literal.* .text .text.*)
     *libgcov.a:(.literal .literal.* .text .text.*)
     *libhal.a:cpu_hal.*(.literal .literal.* .text .text.*)
diff --git a/tools/sdk/esp32/lib/libapp_trace.a b/tools/sdk/esp32/lib/libapp_trace.a
index 457355a9cff..9e08a79c408 100644
Binary files a/tools/sdk/esp32/lib/libapp_trace.a and b/tools/sdk/esp32/lib/libapp_trace.a differ
diff --git a/tools/sdk/esp32/lib/libapp_update.a b/tools/sdk/esp32/lib/libapp_update.a
index 551bfb86cf9..654f1094989 100644
Binary files a/tools/sdk/esp32/lib/libapp_update.a and b/tools/sdk/esp32/lib/libapp_update.a differ
diff --git a/tools/sdk/esp32/lib/libasio.a b/tools/sdk/esp32/lib/libasio.a
index 69e072d1737..f70463d71dd 100644
Binary files a/tools/sdk/esp32/lib/libasio.a and b/tools/sdk/esp32/lib/libasio.a differ
diff --git a/tools/sdk/esp32/lib/libbootloader_support.a b/tools/sdk/esp32/lib/libbootloader_support.a
index db0df17b7a6..1247c3fe92d 100644
Binary files a/tools/sdk/esp32/lib/libbootloader_support.a and b/tools/sdk/esp32/lib/libbootloader_support.a differ
diff --git a/tools/sdk/esp32/lib/libbt.a b/tools/sdk/esp32/lib/libbt.a
index cf6f28abaf3..ae90c9430e0 100644
Binary files a/tools/sdk/esp32/lib/libbt.a and b/tools/sdk/esp32/lib/libbt.a differ
diff --git a/tools/sdk/esp32/lib/libbutton.a b/tools/sdk/esp32/lib/libbutton.a
index ee6935a6eaf..2c7a5c83979 100644
Binary files a/tools/sdk/esp32/lib/libbutton.a and b/tools/sdk/esp32/lib/libbutton.a differ
diff --git a/tools/sdk/esp32/lib/libcoap.a b/tools/sdk/esp32/lib/libcoap.a
index 475f344c21a..a49fbc832e6 100644
Binary files a/tools/sdk/esp32/lib/libcoap.a and b/tools/sdk/esp32/lib/libcoap.a differ
diff --git a/tools/sdk/esp32/lib/libcoexist.a b/tools/sdk/esp32/lib/libcoexist.a
index 6a07d9fab49..1c67ea32c12 100644
Binary files a/tools/sdk/esp32/lib/libcoexist.a and b/tools/sdk/esp32/lib/libcoexist.a differ
diff --git a/tools/sdk/esp32/lib/libconsole.a b/tools/sdk/esp32/lib/libconsole.a
index f711007592c..7d5c3fdac1b 100644
Binary files a/tools/sdk/esp32/lib/libconsole.a and b/tools/sdk/esp32/lib/libconsole.a differ
diff --git a/tools/sdk/esp32/lib/libcore.a b/tools/sdk/esp32/lib/libcore.a
index 5dfca587210..842c189cef0 100644
Binary files a/tools/sdk/esp32/lib/libcore.a and b/tools/sdk/esp32/lib/libcore.a differ
diff --git a/tools/sdk/esp32/lib/libcxx.a b/tools/sdk/esp32/lib/libcxx.a
index ef0e12ce41d..16162e67920 100644
Binary files a/tools/sdk/esp32/lib/libcxx.a and b/tools/sdk/esp32/lib/libcxx.a differ
diff --git a/tools/sdk/esp32/lib/libdriver.a b/tools/sdk/esp32/lib/libdriver.a
index f238bbb787e..57c332baf54 100644
Binary files a/tools/sdk/esp32/lib/libdriver.a and b/tools/sdk/esp32/lib/libdriver.a differ
diff --git a/tools/sdk/esp32/lib/libefuse.a b/tools/sdk/esp32/lib/libefuse.a
index 3f941fd6035..24dc6fb1504 100644
Binary files a/tools/sdk/esp32/lib/libefuse.a and b/tools/sdk/esp32/lib/libefuse.a differ
diff --git a/tools/sdk/esp32/lib/libesp-dsp.a b/tools/sdk/esp32/lib/libesp-dsp.a
index 49b548eaf7a..26e601eb8ac 100644
Binary files a/tools/sdk/esp32/lib/libesp-dsp.a and b/tools/sdk/esp32/lib/libesp-dsp.a differ
diff --git a/tools/sdk/esp32/lib/libesp-tls.a b/tools/sdk/esp32/lib/libesp-tls.a
index b68a26e8905..d9828365b69 100644
Binary files a/tools/sdk/esp32/lib/libesp-tls.a and b/tools/sdk/esp32/lib/libesp-tls.a differ
diff --git a/tools/sdk/esp32/lib/libesp32-camera.a b/tools/sdk/esp32/lib/libesp32-camera.a
index 0514d6ce494..d32a4f7cd10 100644
Binary files a/tools/sdk/esp32/lib/libesp32-camera.a and b/tools/sdk/esp32/lib/libesp32-camera.a differ
diff --git a/tools/sdk/esp32/lib/libesp_common.a b/tools/sdk/esp32/lib/libesp_common.a
index 55ad6397248..8376abc4c81 100644
Binary files a/tools/sdk/esp32/lib/libesp_common.a and b/tools/sdk/esp32/lib/libesp_common.a differ
diff --git a/tools/sdk/esp32/lib/libesp_eth.a b/tools/sdk/esp32/lib/libesp_eth.a
index b1929f1e507..2e6f5cd426e 100644
Binary files a/tools/sdk/esp32/lib/libesp_eth.a and b/tools/sdk/esp32/lib/libesp_eth.a differ
diff --git a/tools/sdk/esp32/lib/libesp_event.a b/tools/sdk/esp32/lib/libesp_event.a
index 553ea1c7aee..e77be44bce7 100644
Binary files a/tools/sdk/esp32/lib/libesp_event.a and b/tools/sdk/esp32/lib/libesp_event.a differ
diff --git a/tools/sdk/esp32/lib/libesp_gdbstub.a b/tools/sdk/esp32/lib/libesp_gdbstub.a
index 4532df2f78a..ebab7a300ff 100644
Binary files a/tools/sdk/esp32/lib/libesp_gdbstub.a and b/tools/sdk/esp32/lib/libesp_gdbstub.a differ
diff --git a/tools/sdk/esp32/lib/libesp_hid.a b/tools/sdk/esp32/lib/libesp_hid.a
index 81355c91e67..3c0e96ef5c1 100644
Binary files a/tools/sdk/esp32/lib/libesp_hid.a and b/tools/sdk/esp32/lib/libesp_hid.a differ
diff --git a/tools/sdk/esp32/lib/libesp_http_client.a b/tools/sdk/esp32/lib/libesp_http_client.a
index 9d084d94942..512bc3a9dcd 100644
Binary files a/tools/sdk/esp32/lib/libesp_http_client.a and b/tools/sdk/esp32/lib/libesp_http_client.a differ
diff --git a/tools/sdk/esp32/lib/libesp_http_server.a b/tools/sdk/esp32/lib/libesp_http_server.a
index f20074b7e94..d304e1e31b0 100644
Binary files a/tools/sdk/esp32/lib/libesp_http_server.a and b/tools/sdk/esp32/lib/libesp_http_server.a differ
diff --git a/tools/sdk/esp32/lib/libesp_https_ota.a b/tools/sdk/esp32/lib/libesp_https_ota.a
index c8f743f733a..16fd6ab15c7 100644
Binary files a/tools/sdk/esp32/lib/libesp_https_ota.a and b/tools/sdk/esp32/lib/libesp_https_ota.a differ
diff --git a/tools/sdk/esp32/lib/libesp_hw_support.a b/tools/sdk/esp32/lib/libesp_hw_support.a
index 8b318d6bfdb..36e6c9cfda8 100644
Binary files a/tools/sdk/esp32/lib/libesp_hw_support.a and b/tools/sdk/esp32/lib/libesp_hw_support.a differ
diff --git a/tools/sdk/esp32/lib/libesp_ipc.a b/tools/sdk/esp32/lib/libesp_ipc.a
index 8958f861427..3ccc4e9a5e4 100644
Binary files a/tools/sdk/esp32/lib/libesp_ipc.a and b/tools/sdk/esp32/lib/libesp_ipc.a differ
diff --git a/tools/sdk/esp32/lib/libesp_lcd.a b/tools/sdk/esp32/lib/libesp_lcd.a
index 1801d9b8755..8dda0d105bd 100644
Binary files a/tools/sdk/esp32/lib/libesp_lcd.a and b/tools/sdk/esp32/lib/libesp_lcd.a differ
diff --git a/tools/sdk/esp32/lib/libesp_littlefs.a b/tools/sdk/esp32/lib/libesp_littlefs.a
index 38290da68ae..b1cb9bf4003 100644
Binary files a/tools/sdk/esp32/lib/libesp_littlefs.a and b/tools/sdk/esp32/lib/libesp_littlefs.a differ
diff --git a/tools/sdk/esp32/lib/libesp_local_ctrl.a b/tools/sdk/esp32/lib/libesp_local_ctrl.a
index e1f10c0dffe..c082eae9d9b 100644
Binary files a/tools/sdk/esp32/lib/libesp_local_ctrl.a and b/tools/sdk/esp32/lib/libesp_local_ctrl.a differ
diff --git a/tools/sdk/esp32/lib/libesp_netif.a b/tools/sdk/esp32/lib/libesp_netif.a
index d606140499b..d9762dd28d4 100644
Binary files a/tools/sdk/esp32/lib/libesp_netif.a and b/tools/sdk/esp32/lib/libesp_netif.a differ
diff --git a/tools/sdk/esp32/lib/libesp_phy.a b/tools/sdk/esp32/lib/libesp_phy.a
index 3abd25f53b2..9681fab43cc 100644
Binary files a/tools/sdk/esp32/lib/libesp_phy.a and b/tools/sdk/esp32/lib/libesp_phy.a differ
diff --git a/tools/sdk/esp32/lib/libesp_pm.a b/tools/sdk/esp32/lib/libesp_pm.a
index 39cb0ce9527..9ef0b9ac522 100644
Binary files a/tools/sdk/esp32/lib/libesp_pm.a and b/tools/sdk/esp32/lib/libesp_pm.a differ
diff --git a/tools/sdk/esp32/lib/libesp_rainmaker.a b/tools/sdk/esp32/lib/libesp_rainmaker.a
index eb552bda46d..ab5ece21d76 100644
Binary files a/tools/sdk/esp32/lib/libesp_rainmaker.a and b/tools/sdk/esp32/lib/libesp_rainmaker.a differ
diff --git a/tools/sdk/esp32/lib/libesp_ringbuf.a b/tools/sdk/esp32/lib/libesp_ringbuf.a
index e3fe56d6590..68c6e93b0d9 100644
Binary files a/tools/sdk/esp32/lib/libesp_ringbuf.a and b/tools/sdk/esp32/lib/libesp_ringbuf.a differ
diff --git a/tools/sdk/esp32/lib/libesp_schedule.a b/tools/sdk/esp32/lib/libesp_schedule.a
index 13d28e57465..c13fe9c30a5 100644
Binary files a/tools/sdk/esp32/lib/libesp_schedule.a and b/tools/sdk/esp32/lib/libesp_schedule.a differ
diff --git a/tools/sdk/esp32/lib/libesp_serial_slave_link.a b/tools/sdk/esp32/lib/libesp_serial_slave_link.a
index 0f49c488742..9dcff8989a6 100644
Binary files a/tools/sdk/esp32/lib/libesp_serial_slave_link.a and b/tools/sdk/esp32/lib/libesp_serial_slave_link.a differ
diff --git a/tools/sdk/esp32/lib/libesp_system.a b/tools/sdk/esp32/lib/libesp_system.a
index f50673a65c4..2cc68405c88 100644
Binary files a/tools/sdk/esp32/lib/libesp_system.a and b/tools/sdk/esp32/lib/libesp_system.a differ
diff --git a/tools/sdk/esp32/lib/libesp_timer.a b/tools/sdk/esp32/lib/libesp_timer.a
index 6dd4d143f6e..04672445255 100644
Binary files a/tools/sdk/esp32/lib/libesp_timer.a and b/tools/sdk/esp32/lib/libesp_timer.a differ
diff --git a/tools/sdk/esp32/lib/libesp_websocket_client.a b/tools/sdk/esp32/lib/libesp_websocket_client.a
index 4ea65a16e18..390bdc9b029 100644
Binary files a/tools/sdk/esp32/lib/libesp_websocket_client.a and b/tools/sdk/esp32/lib/libesp_websocket_client.a differ
diff --git a/tools/sdk/esp32/lib/libesp_wifi.a b/tools/sdk/esp32/lib/libesp_wifi.a
index 6d7bcc29315..7779a0a840a 100644
Binary files a/tools/sdk/esp32/lib/libesp_wifi.a and b/tools/sdk/esp32/lib/libesp_wifi.a differ
diff --git a/tools/sdk/esp32/lib/libespcoredump.a b/tools/sdk/esp32/lib/libespcoredump.a
index 25cd8820b99..c73df3b029d 100644
Binary files a/tools/sdk/esp32/lib/libespcoredump.a and b/tools/sdk/esp32/lib/libespcoredump.a differ
diff --git a/tools/sdk/esp32/lib/libespnow.a b/tools/sdk/esp32/lib/libespnow.a
index 9b53365726b..24169f6f0c2 100644
Binary files a/tools/sdk/esp32/lib/libespnow.a and b/tools/sdk/esp32/lib/libespnow.a differ
diff --git a/tools/sdk/esp32/lib/libfatfs.a b/tools/sdk/esp32/lib/libfatfs.a
index 4ee8d7990ed..a47df9006e2 100644
Binary files a/tools/sdk/esp32/lib/libfatfs.a and b/tools/sdk/esp32/lib/libfatfs.a differ
diff --git a/tools/sdk/esp32/lib/libfreemodbus.a b/tools/sdk/esp32/lib/libfreemodbus.a
index 66297b41164..f1ff8c861f8 100644
Binary files a/tools/sdk/esp32/lib/libfreemodbus.a and b/tools/sdk/esp32/lib/libfreemodbus.a differ
diff --git a/tools/sdk/esp32/lib/libfreertos.a b/tools/sdk/esp32/lib/libfreertos.a
index 310a0260efa..4951dc26239 100644
Binary files a/tools/sdk/esp32/lib/libfreertos.a and b/tools/sdk/esp32/lib/libfreertos.a differ
diff --git a/tools/sdk/esp32/lib/libhal.a b/tools/sdk/esp32/lib/libhal.a
index 1ec5baa88e8..f27a2354ee4 100644
Binary files a/tools/sdk/esp32/lib/libhal.a and b/tools/sdk/esp32/lib/libhal.a differ
diff --git a/tools/sdk/esp32/lib/libheap.a b/tools/sdk/esp32/lib/libheap.a
index 9c1f8d53e6e..e564f6767e2 100644
Binary files a/tools/sdk/esp32/lib/libheap.a and b/tools/sdk/esp32/lib/libheap.a differ
diff --git a/tools/sdk/esp32/lib/liblog.a b/tools/sdk/esp32/lib/liblog.a
index 32d9e92b187..120920f8dff 100644
Binary files a/tools/sdk/esp32/lib/liblog.a and b/tools/sdk/esp32/lib/liblog.a differ
diff --git a/tools/sdk/esp32/lib/liblwip.a b/tools/sdk/esp32/lib/liblwip.a
index ad6dcde9e98..2c7e5df5d35 100644
Binary files a/tools/sdk/esp32/lib/liblwip.a and b/tools/sdk/esp32/lib/liblwip.a differ
diff --git a/tools/sdk/esp32/lib/libmbedcrypto.a b/tools/sdk/esp32/lib/libmbedcrypto.a
index b63bf72ce9b..c0d35ee00ee 100644
Binary files a/tools/sdk/esp32/lib/libmbedcrypto.a and b/tools/sdk/esp32/lib/libmbedcrypto.a differ
diff --git a/tools/sdk/esp32/lib/libmbedtls.a b/tools/sdk/esp32/lib/libmbedtls.a
index 116564a2f60..44b1fdd7f73 100644
Binary files a/tools/sdk/esp32/lib/libmbedtls.a and b/tools/sdk/esp32/lib/libmbedtls.a differ
diff --git a/tools/sdk/esp32/lib/libmdns.a b/tools/sdk/esp32/lib/libmdns.a
index a07fbbd45ea..ce5e94a5b1e 100644
Binary files a/tools/sdk/esp32/lib/libmdns.a and b/tools/sdk/esp32/lib/libmdns.a differ
diff --git a/tools/sdk/esp32/lib/libmesh.a b/tools/sdk/esp32/lib/libmesh.a
index 9a32bef01fd..4796b538801 100644
Binary files a/tools/sdk/esp32/lib/libmesh.a and b/tools/sdk/esp32/lib/libmesh.a differ
diff --git a/tools/sdk/esp32/lib/libmqtt.a b/tools/sdk/esp32/lib/libmqtt.a
index 127053e6c78..39156994d7c 100644
Binary files a/tools/sdk/esp32/lib/libmqtt.a and b/tools/sdk/esp32/lib/libmqtt.a differ
diff --git a/tools/sdk/esp32/lib/libnet80211.a b/tools/sdk/esp32/lib/libnet80211.a
index 75b75922b13..dce7fceb9d6 100644
Binary files a/tools/sdk/esp32/lib/libnet80211.a and b/tools/sdk/esp32/lib/libnet80211.a differ
diff --git a/tools/sdk/esp32/lib/libnewlib.a b/tools/sdk/esp32/lib/libnewlib.a
index 0f96d134212..95712a71002 100644
Binary files a/tools/sdk/esp32/lib/libnewlib.a and b/tools/sdk/esp32/lib/libnewlib.a differ
diff --git a/tools/sdk/esp32/lib/libnvs_flash.a b/tools/sdk/esp32/lib/libnvs_flash.a
index 7ef82432abc..c3d40bbb2bc 100644
Binary files a/tools/sdk/esp32/lib/libnvs_flash.a and b/tools/sdk/esp32/lib/libnvs_flash.a differ
diff --git a/tools/sdk/esp32/lib/libopenssl.a b/tools/sdk/esp32/lib/libopenssl.a
index 1a87c1f7d05..fcae8b2b012 100644
Binary files a/tools/sdk/esp32/lib/libopenssl.a and b/tools/sdk/esp32/lib/libopenssl.a differ
diff --git a/tools/sdk/esp32/lib/libpp.a b/tools/sdk/esp32/lib/libpp.a
index 3ec218aa6f1..73458bfedf6 100644
Binary files a/tools/sdk/esp32/lib/libpp.a and b/tools/sdk/esp32/lib/libpp.a differ
diff --git a/tools/sdk/esp32/lib/libprotocomm.a b/tools/sdk/esp32/lib/libprotocomm.a
index 701acf9e584..666d3586672 100644
Binary files a/tools/sdk/esp32/lib/libprotocomm.a and b/tools/sdk/esp32/lib/libprotocomm.a differ
diff --git a/tools/sdk/esp32/lib/libpthread.a b/tools/sdk/esp32/lib/libpthread.a
index 1329ba601b3..cb1da0a5633 100644
Binary files a/tools/sdk/esp32/lib/libpthread.a and b/tools/sdk/esp32/lib/libpthread.a differ
diff --git a/tools/sdk/esp32/lib/libsdmmc.a b/tools/sdk/esp32/lib/libsdmmc.a
index 30b5bbaa4e5..09a42278610 100644
Binary files a/tools/sdk/esp32/lib/libsdmmc.a and b/tools/sdk/esp32/lib/libsdmmc.a differ
diff --git a/tools/sdk/esp32/lib/libsmartconfig.a b/tools/sdk/esp32/lib/libsmartconfig.a
index 2ae36360fa1..e5fb258bcbe 100644
Binary files a/tools/sdk/esp32/lib/libsmartconfig.a and b/tools/sdk/esp32/lib/libsmartconfig.a differ
diff --git a/tools/sdk/esp32/lib/libsoc.a b/tools/sdk/esp32/lib/libsoc.a
index 5b5287c04be..ad5bde888c3 100644
Binary files a/tools/sdk/esp32/lib/libsoc.a and b/tools/sdk/esp32/lib/libsoc.a differ
diff --git a/tools/sdk/esp32/lib/libspi_flash.a b/tools/sdk/esp32/lib/libspi_flash.a
index f9a11137071..7ba660971db 100644
Binary files a/tools/sdk/esp32/lib/libspi_flash.a and b/tools/sdk/esp32/lib/libspi_flash.a differ
diff --git a/tools/sdk/esp32/lib/libspiffs.a b/tools/sdk/esp32/lib/libspiffs.a
index 56eea5993cf..c3890f81830 100644
Binary files a/tools/sdk/esp32/lib/libspiffs.a and b/tools/sdk/esp32/lib/libspiffs.a differ
diff --git a/tools/sdk/esp32/lib/libtcp_transport.a b/tools/sdk/esp32/lib/libtcp_transport.a
index 83f2ad57d38..d26d2dc2277 100644
Binary files a/tools/sdk/esp32/lib/libtcp_transport.a and b/tools/sdk/esp32/lib/libtcp_transport.a differ
diff --git a/tools/sdk/esp32/lib/libtcpip_adapter.a b/tools/sdk/esp32/lib/libtcpip_adapter.a
index 20f3e0cb843..bcbad81866e 100644
Binary files a/tools/sdk/esp32/lib/libtcpip_adapter.a and b/tools/sdk/esp32/lib/libtcpip_adapter.a differ
diff --git a/tools/sdk/esp32/lib/libvfs.a b/tools/sdk/esp32/lib/libvfs.a
index 51230fadaf3..3b6596e43be 100644
Binary files a/tools/sdk/esp32/lib/libvfs.a and b/tools/sdk/esp32/lib/libvfs.a differ
diff --git a/tools/sdk/esp32/lib/libwapi.a b/tools/sdk/esp32/lib/libwapi.a
index 4d0200ad4b0..53e0fe9d410 100644
Binary files a/tools/sdk/esp32/lib/libwapi.a and b/tools/sdk/esp32/lib/libwapi.a differ
diff --git a/tools/sdk/esp32/lib/libwear_levelling.a b/tools/sdk/esp32/lib/libwear_levelling.a
index e6793f3b3c9..1d853f8b096 100644
Binary files a/tools/sdk/esp32/lib/libwear_levelling.a and b/tools/sdk/esp32/lib/libwear_levelling.a differ
diff --git a/tools/sdk/esp32/lib/libwifi_provisioning.a b/tools/sdk/esp32/lib/libwifi_provisioning.a
index b0764129217..5ae2dd1647e 100644
Binary files a/tools/sdk/esp32/lib/libwifi_provisioning.a and b/tools/sdk/esp32/lib/libwifi_provisioning.a differ
diff --git a/tools/sdk/esp32/lib/libwpa_supplicant.a b/tools/sdk/esp32/lib/libwpa_supplicant.a
index 011380db2c8..a65d59f66e6 100644
Binary files a/tools/sdk/esp32/lib/libwpa_supplicant.a and b/tools/sdk/esp32/lib/libwpa_supplicant.a differ
diff --git a/tools/sdk/esp32/lib/libxtensa.a b/tools/sdk/esp32/lib/libxtensa.a
index 37ff7464dc0..2e9b6d979ce 100644
Binary files a/tools/sdk/esp32/lib/libxtensa.a and b/tools/sdk/esp32/lib/libxtensa.a differ
diff --git a/tools/sdk/esp32/sdkconfig b/tools/sdk/esp32/sdkconfig
index 79ed5b40ca2..20f7d268c66 100644
--- a/tools/sdk/esp32/sdkconfig
+++ b/tools/sdk/esp32/sdkconfig
@@ -83,6 +83,7 @@ CONFIG_ESPTOOLPY_BAUD_OTHER_VAL=115200
 # CONFIG_ESPTOOLPY_FLASHMODE_QOUT is not set
 CONFIG_ESPTOOLPY_FLASHMODE_DIO=y
 # CONFIG_ESPTOOLPY_FLASHMODE_DOUT is not set
+CONFIG_ESPTOOLPY_FLASH_SAMPLE_MODE_STR=y
 CONFIG_ESPTOOLPY_FLASHMODE="dio"
 # CONFIG_ESPTOOLPY_FLASHFREQ_120M is not set
 # CONFIG_ESPTOOLPY_FLASHFREQ_80M is not set
@@ -1026,6 +1027,7 @@ CONFIG_LWIP_TCPIP_RECVMBOX_SIZE=32
 # CONFIG_LWIP_DHCP_DOES_ARP_CHECK is not set
 # CONFIG_LWIP_DHCP_DISABLE_CLIENT_ID is not set
 CONFIG_LWIP_DHCP_RESTORE_LAST_IP=y
+CONFIG_LWIP_DHCP_OPTIONS_LEN=68
 
 #
 # DHCP server
@@ -1282,6 +1284,7 @@ CONFIG_MDNS_SERVICE_ADD_TIMEOUT_MS=2000
 # CONFIG_MDNS_STRICT_MODE is not set
 CONFIG_MDNS_TIMER_PERIOD_MS=100
 # CONFIG_MDNS_NETWORKING_SOCKET is not set
+CONFIG_MDNS_MULTIPLE_INSTANCE=y
 # end of mDNS
 
 #
@@ -1466,6 +1469,7 @@ CONFIG_WL_SECTOR_SIZE=4096
 #
 CONFIG_WIFI_PROV_SCAN_MAX_ENTRIES=16
 CONFIG_WIFI_PROV_AUTOSTOP_TIMEOUT=30
+# CONFIG_WIFI_PROV_BLE_BONDING is not set
 # end of Wi-Fi Provisioning Manager
 
 #
diff --git a/tools/sdk/esp32c3/bin/bootloader_dio_40m.bin b/tools/sdk/esp32c3/bin/bootloader_dio_40m.bin
index 78c947604fd..de9f4505aad 100644
Binary files a/tools/sdk/esp32c3/bin/bootloader_dio_40m.bin and b/tools/sdk/esp32c3/bin/bootloader_dio_40m.bin differ
diff --git a/tools/sdk/esp32c3/bin/bootloader_dio_80m.bin b/tools/sdk/esp32c3/bin/bootloader_dio_80m.bin
index 5fdaf4eed6a..bf8998d52d3 100644
Binary files a/tools/sdk/esp32c3/bin/bootloader_dio_80m.bin and b/tools/sdk/esp32c3/bin/bootloader_dio_80m.bin differ
diff --git a/tools/sdk/esp32c3/bin/bootloader_dout_40m.bin b/tools/sdk/esp32c3/bin/bootloader_dout_40m.bin
index e369d99e862..dfe29b31b31 100644
Binary files a/tools/sdk/esp32c3/bin/bootloader_dout_40m.bin and b/tools/sdk/esp32c3/bin/bootloader_dout_40m.bin differ
diff --git a/tools/sdk/esp32c3/bin/bootloader_dout_80m.bin b/tools/sdk/esp32c3/bin/bootloader_dout_80m.bin
index 2b9800fc7e7..0534d487951 100644
Binary files a/tools/sdk/esp32c3/bin/bootloader_dout_80m.bin and b/tools/sdk/esp32c3/bin/bootloader_dout_80m.bin differ
diff --git a/tools/sdk/esp32c3/bin/bootloader_qio_40m.bin b/tools/sdk/esp32c3/bin/bootloader_qio_40m.bin
index 22880d8cf71..d9b1d9e775a 100644
Binary files a/tools/sdk/esp32c3/bin/bootloader_qio_40m.bin and b/tools/sdk/esp32c3/bin/bootloader_qio_40m.bin differ
diff --git a/tools/sdk/esp32c3/bin/bootloader_qio_80m.bin b/tools/sdk/esp32c3/bin/bootloader_qio_80m.bin
index d630c8774ee..f3ce6940c07 100644
Binary files a/tools/sdk/esp32c3/bin/bootloader_qio_80m.bin and b/tools/sdk/esp32c3/bin/bootloader_qio_80m.bin differ
diff --git a/tools/sdk/esp32c3/bin/bootloader_qout_40m.bin b/tools/sdk/esp32c3/bin/bootloader_qout_40m.bin
index 9566026d6b4..960af55a46f 100644
Binary files a/tools/sdk/esp32c3/bin/bootloader_qout_40m.bin and b/tools/sdk/esp32c3/bin/bootloader_qout_40m.bin differ
diff --git a/tools/sdk/esp32c3/bin/bootloader_qout_80m.bin b/tools/sdk/esp32c3/bin/bootloader_qout_80m.bin
index d8a35e03d63..cdc77df22bb 100644
Binary files a/tools/sdk/esp32c3/bin/bootloader_qout_80m.bin and b/tools/sdk/esp32c3/bin/bootloader_qout_80m.bin differ
diff --git a/tools/sdk/esp32c3/include/asio/port/include/esp_asio_config.h b/tools/sdk/esp32c3/include/asio/port/include/esp_asio_config.h
index cba316527e6..3f3a9b03ed4 100644
--- a/tools/sdk/esp32c3/include/asio/port/include/esp_asio_config.h
+++ b/tools/sdk/esp32c3/include/asio/port/include/esp_asio_config.h
@@ -18,6 +18,11 @@
 #  define ASIO_NO_TYPEID
 # endif   // CONFIG_COMPILER_RTTI
 
+//
+// Supress OpenSSL deprecation warning, when building ASIO
+//
+#define ESP_OPENSSL_SUPPRESS_LEGACY_WARNING
+
 //
 // LWIP compatibility inet and address macros/functions
 //
diff --git a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_block_internal.h b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_block_internal.h
index 9abe81557fa..b7ad0a554cc 100644
--- a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_block_internal.h
+++ b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_block_internal.h
@@ -191,6 +191,9 @@ int coap_handle_response_get_block(coap_context_t *context,
 void coap_block_delete_lg_xmit(coap_session_t *session,
                                coap_lg_xmit_t *lg_xmit);
 
+coap_tick_t coap_block_check_lg_xmit_timeouts(coap_session_t *session,
+                                              coap_tick_t now);
+
 /**
  * The function that does all the work for the coap_add_data_large*()
  * functions.
diff --git a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_dtls.h b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_dtls.h
index cbd369dfce6..fc30445431d 100644
--- a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_dtls.h
+++ b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_dtls.h
@@ -27,6 +27,12 @@ typedef struct coap_dtls_pki_t coap_dtls_pki_t;
 #ifndef COAP_DTLS_HINT_LENGTH
 #define COAP_DTLS_HINT_LENGTH 128
 #endif
+#ifndef COAP_DTLS_MAX_PSK_IDENTITY
+#define COAP_DTLS_MAX_PSK_IDENTITY 64
+#endif
+#ifndef COAP_DTLS_MAX_PSK
+#define COAP_DTLS_MAX_PSK 64
+#endif
 
 typedef enum coap_dtls_role_t {
   COAP_DTLS_ROLE_CLIENT, /**< Internal function invoked for client */
diff --git a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_event.h b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_event.h
index 89b2a63967c..b6ea60524a2 100644
--- a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_event.h
+++ b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_event.h
@@ -24,34 +24,34 @@
  * Scalar type to represent different events, e.g. DTLS events or
  * retransmission timeouts.
  */
- typedef unsigned int coap_event_t;
-
+typedef enum coap_event_t {
 /**
  * (D)TLS events for COAP_PROTO_DTLS and COAP_PROTO_TLS
  */
-#define COAP_EVENT_DTLS_CLOSED        0x0000
-#define COAP_EVENT_DTLS_CONNECTED     0x01DE
-#define COAP_EVENT_DTLS_RENEGOTIATE   0x01DF
-#define COAP_EVENT_DTLS_ERROR         0x0200
+  COAP_EVENT_DTLS_CLOSED       = 0x0000,
+  COAP_EVENT_DTLS_CONNECTED    = 0x01DE,
+  COAP_EVENT_DTLS_RENEGOTIATE  = 0x01DF,
+  COAP_EVENT_DTLS_ERROR        = 0x0200,
 
 /**
  * TCP events for COAP_PROTO_TCP and COAP_PROTO_TLS
  */
-#define COAP_EVENT_TCP_CONNECTED      0x1001
-#define COAP_EVENT_TCP_CLOSED         0x1002
-#define COAP_EVENT_TCP_FAILED         0x1003
+  COAP_EVENT_TCP_CONNECTED     = 0x1001,
+  COAP_EVENT_TCP_CLOSED        = 0x1002,
+  COAP_EVENT_TCP_FAILED        = 0x1003,
 
 /**
  * CSM exchange events for reliable protocols only
  */
-#define COAP_EVENT_SESSION_CONNECTED  0x2001
-#define COAP_EVENT_SESSION_CLOSED     0x2002
-#define COAP_EVENT_SESSION_FAILED     0x2003
+  COAP_EVENT_SESSION_CONNECTED = 0x2001,
+  COAP_EVENT_SESSION_CLOSED    = 0x2002,
+  COAP_EVENT_SESSION_FAILED    = 0x2003,
 
 /**
- * BLOCK2 receive errors
+ * (Q-)BLOCK receive errors
  */
-#define COAP_EVENT_PARTIAL_BLOCK      0x3001
+  COAP_EVENT_PARTIAL_BLOCK     = 0x3001
+} coap_event_t;
 
 /**
  * Type for event handler functions that can be registered with a CoAP
diff --git a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_time.h b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_time.h
index baa8650eaff..99d117a4eb7 100644
--- a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_time.h
+++ b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/coap_time.h
@@ -88,7 +88,11 @@ COAP_STATIC_INLINE uint64_t coap_ticks_to_rt_us(coap_tick_t t) {
 #elif defined(RIOT_VERSION)
 #include <xtimer.h>
 
+#ifdef XTIMER_HZ
 #define COAP_TICKS_PER_SECOND (XTIMER_HZ)
+#else /* XTIMER_HZ */
+#define COAP_TICKS_PER_SECOND (XTIMER_HZ_BASE)
+#endif /* XTIMER_HZ */
 
 typedef uint64_t coap_tick_t;
 typedef int64_t coap_tick_diff_t;
diff --git a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/net.h b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/net.h
index 577a0b5d5a5..ea5a2cba1bb 100644
--- a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/net.h
+++ b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/net.h
@@ -15,6 +15,7 @@
 #include <stdlib.h>
 #include <string.h>
 #ifndef _WIN32
+#include <sys/select.h>
 #include <sys/time.h>
 #endif
 #include <time.h>
diff --git a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/pdu.h b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/pdu.h
index a22869b69ed..8031a1c2c40 100644
--- a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/pdu.h
+++ b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/pdu.h
@@ -299,7 +299,6 @@ typedef enum coap_pdu_code_t {
   COAP_REQUEST_CODE_PATCH  = COAP_REQUEST_PATCH,
   COAP_REQUEST_CODE_IPATCH = COAP_REQUEST_IPATCH,
 
-  COAP_RESPONSE_CODE_OK                         = COAP_RESPONSE_CODE(200),
   COAP_RESPONSE_CODE_CREATED                    = COAP_RESPONSE_CODE(201),
   COAP_RESPONSE_CODE_DELETED                    = COAP_RESPONSE_CODE(202),
   COAP_RESPONSE_CODE_VALID                      = COAP_RESPONSE_CODE(203),
diff --git a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/resource.h b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/resource.h
index 2cd9aea48fa..5cf7751f67e 100644
--- a/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/resource.h
+++ b/tools/sdk/esp32c3/include/coap/libcoap/include/coap3/resource.h
@@ -83,7 +83,8 @@ typedef void (*coap_method_handler_t)
  * variable of coap_str_const_t has to point to constant text, or point to data
  * within the allocated coap_str_const_t parameter.
  *
- * @param uri_path The string URI path of the new resource.
+ * @param uri_path The string URI path of the new resource. The leading '/' is
+ *                 not normally required - e.g. just "full/path/for/resource".
  * @param flags    Flags for memory management (in particular release of
  *                 memory). Possible values:@n
  *
diff --git a/tools/sdk/esp32c3/include/config/sdkconfig.h b/tools/sdk/esp32c3/include/config/sdkconfig.h
index f1f31fdfd3b..c07e2b52f40 100644
--- a/tools/sdk/esp32c3/include/config/sdkconfig.h
+++ b/tools/sdk/esp32c3/include/config/sdkconfig.h
@@ -29,6 +29,7 @@
 #define CONFIG_BOOT_ROM_LOG_ALWAYS_ON 1
 #define CONFIG_ESPTOOLPY_BAUD_OTHER_VAL 115200
 #define CONFIG_ESPTOOLPY_FLASHMODE_DIO 1
+#define CONFIG_ESPTOOLPY_FLASH_SAMPLE_MODE_STR 1
 #define CONFIG_ESPTOOLPY_FLASHMODE "dio"
 #define CONFIG_ESPTOOLPY_FLASHFREQ_80M 1
 #define CONFIG_ESPTOOLPY_FLASHFREQ "80m"
@@ -340,6 +341,7 @@
 #define CONFIG_LWIP_GARP_TMR_INTERVAL 60
 #define CONFIG_LWIP_TCPIP_RECVMBOX_SIZE 32
 #define CONFIG_LWIP_DHCP_DOES_ARP_CHECK 1
+#define CONFIG_LWIP_DHCP_OPTIONS_LEN 68
 #define CONFIG_LWIP_DHCPS 1
 #define CONFIG_LWIP_DHCPS_LEASE_UNIT 60
 #define CONFIG_LWIP_DHCPS_MAX_STATION_NUM 8
@@ -450,6 +452,7 @@
 #define CONFIG_MDNS_TASK_AFFINITY 0x0
 #define CONFIG_MDNS_SERVICE_ADD_TIMEOUT_MS 2000
 #define CONFIG_MDNS_TIMER_PERIOD_MS 100
+#define CONFIG_MDNS_MULTIPLE_INSTANCE 1
 #define CONFIG_MQTT_PROTOCOL_311 1
 #define CONFIG_MQTT_TRANSPORT_SSL 1
 #define CONFIG_MQTT_TRANSPORT_WEBSOCKET 1
@@ -622,5 +625,5 @@
 #define CONFIG_TIMER_TASK_STACK_SIZE CONFIG_ESP_TIMER_TASK_STACK_SIZE
 #define CONFIG_TOOLPREFIX CONFIG_SDK_TOOLPREFIX
 #define CONFIG_UDP_RECVMBOX_SIZE CONFIG_LWIP_UDP_RECVMBOX_SIZE
-#define CONFIG_ARDUINO_IDF_COMMIT "3e370c4296"
+#define CONFIG_ARDUINO_IDF_COMMIT "b86fe0c66c"
 #define CONFIG_ARDUINO_IDF_BRANCH "master"
diff --git a/tools/sdk/esp32c3/include/driver/include/driver/rmt.h b/tools/sdk/esp32c3/include/driver/include/driver/rmt.h
index a7e2aad5f24..bc07954a080 100644
--- a/tools/sdk/esp32c3/include/driver/include/driver/rmt.h
+++ b/tools/sdk/esp32c3/include/driver/include/driver/rmt.h
@@ -856,16 +856,35 @@ esp_err_t rmt_remove_channel_from_group(rmt_channel_t channel);
 
 #if SOC_RMT_SUPPORT_TX_LOOP_COUNT
 /**
- * @brief Set loop count for RMT TX channel
+ * @brief Set loop count threshold value for RMT TX channel
+ *
+ * When tx loop count reaches this value, an ISR callback will notify user
  *
  * @param channel RMT channel
- * @param count loop count
+ * @param count loop count, 1 ~ 1023
  * @return
  *      - ESP_ERR_INVALID_ARG Parameter error
  *      - ESP_OK Success
  */
 esp_err_t rmt_set_tx_loop_count(rmt_channel_t channel, uint32_t count);
-#endif
+
+/**
+ * @brief Enable or disable the feature that when loop count reaches the threshold, RMT will stop transmitting.
+ *
+ * - When the loop auto-stop feature is enabled will halt RMT transmission after the loop count reaches a certain threshold
+ * - When disabled, the RMT transmission continue indefinitely until halted by the users
+ *
+ * @note The auto-stop feature is implemented in hardware on particular targets (i.e. those with SOC_RMT_SUPPORT_TX_LOOP_AUTOSTOP defined).
+ *       Otherwise, the auto-stop feature is implemented in software via the interrupt.
+ *
+ * @param channel RMT channel
+ * @param en enable bit
+ * @return
+ *      - ESP_ERR_INVALID_ARG Parameter error
+ *      - ESP_OK Success
+ */
+esp_err_t rmt_enable_tx_loop_autostop(rmt_channel_t channel, bool en);
+#endif // SOC_RMT_SUPPORT_TX_LOOP_COUNT
 
 /**
 * @brief Reset RMT TX/RX memory index.
diff --git a/tools/sdk/esp32c3/include/esp-face/include/detect/dl_detect_define.hpp b/tools/sdk/esp32c3/include/esp-face/include/detect/dl_detect_define.hpp
new file mode 100644
index 00000000000..cc53050cb9b
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/detect/dl_detect_define.hpp
@@ -0,0 +1,17 @@
+#pragma once
+
+#include <vector>
+
+namespace dl
+{
+    namespace detect
+    {
+        typedef struct
+        {
+            int category;              /*<! category index */
+            float score;               /*<! score of box */
+            std::vector<int> box;      /*<! [left_up_x, left_up_y, right_down_x, right_down_y] */
+            std::vector<int> keypoint; /*<! [x1, y1, x2, y2, ...] */
+        } result_t;
+    }
+}
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/dl_define.hpp b/tools/sdk/esp32c3/include/esp-face/include/dl_define.hpp
new file mode 100644
index 00000000000..7809768a5bc
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/dl_define.hpp
@@ -0,0 +1,99 @@
+#pragma once
+
+#include <climits>
+#include "sdkconfig.h"
+
+#define DL_LOG_LATENCY_UNIT 0  /*<! - 1: cycle */
+                               /*<! - 0: us */
+#define DL_LOG_NN_LATENCY 0    /*<! - 1: print the latency of each parts of nn */
+                               /*<! - 0: mute */
+#define DL_LOG_LAYER_LATENCY 0 /*<! - 1: print the latency of each parts of layer */
+                               /*<! - 0: mute */
+
+#if CONFIG_SPIRAM_SUPPORT || CONFIG_ESP32_SPIRAM_SUPPORT || CONFIG_ESP32S2_SPIRAM_SUPPORT || CONFIG_ESP32S3_SPIRAM_SUPPORT
+#define DL_SPIRAM_SUPPORT 1
+#else
+#define DL_SPIRAM_SUPPORT 0
+#endif
+
+#if CONFIG_IDF_TARGET_ESP32
+#define CONFIG_DEFAULT_ASSIGN_CORE \
+    {                              \
+    } // TODO: 多核 task 完成时，改成默认 0,1
+#elif CONFIG_IDF_TARGET_ESP32S2
+#define CONFIG_DEFAULT_ASSIGN_CORE \
+    {                              \
+    }
+#elif CONFIG_IDF_TARGET_ESP32S3
+#define CONFIG_DEFAULT_ASSIGN_CORE \
+    {                              \
+    } // TODO: 多核 task 完成时，改成默认 0,1
+#elif CONFIG_IDF_TARGET_ESP32C3
+#define CONFIG_DEFAULT_ASSIGN_CORE \
+    {                              \
+    }
+#else
+#define CONFIG_DEFAULT_ASSIGN_CORE \
+    {                              \
+    }
+#endif
+
+#define DL_Q16_MIN (-32768)
+#define DL_Q16_MAX (32767)
+#define DL_Q8_MIN (-128)
+#define DL_Q8_MAX (127)
+
+#ifndef DL_MAX
+#define DL_MAX(x, y) (((x) < (y)) ? (y) : (x))
+#endif
+
+#ifndef DL_MIN
+#define DL_MIN(x, y) (((x) < (y)) ? (x) : (y))
+#endif
+
+#ifndef DL_CLIP
+#define DL_CLIP(x, low, high) ((x) < (low)) ? (low) : (((x) > (high)) ? (high) : (x))
+#endif
+
+#ifndef DL_ABS
+#define DL_ABS(x) ((x) < 0 ? (-(x)) : (x))
+#endif
+
+#ifndef DL_RIGHT_SHIFT
+#define DL_RIGHT_SHIFT(x, shift) ((shift) > 0) ? ((x) >> (shift)) : ((x) << -(shift))
+#endif
+
+#ifndef DL_LEFT_SHIFT
+#define DL_LEFT_SHIFT(x, shift) ((shift) > 0) ? ((x) << (shift)) : ((x) >> -(shift))
+#endif
+
+namespace dl
+{
+    typedef enum
+    {
+        Linear,    /*<! Linear >*/
+        ReLU,      /*<! ReLU >*/
+        LeakyReLU, /*<! LeakyReLU >*/
+        PReLU,     /*<! PReLU >*/
+        // TODO: Sigmoid,   /*<! Sigmoid >*/
+        // TODO: Softmax,    /*<! Softmax*/
+        // TODO: TanH,
+        // TODO: ReLU6
+    } activation_type_t;
+
+    typedef enum
+    {
+        PADDING_NOT_SET,
+        PADDING_VALID,      /*<! no padding >*/
+        PADDING_SAME_BEGIN,  /*<! SAME in MXNET style >*/
+        PADDING_SAME_END,   /*<! SAME in TensorFlow style >*/
+    } padding_type_t;
+    
+    typedef enum
+    {
+        CONSTANT,
+        EDGE, 
+        REFLECT,
+        SYMMETRIC,
+    } padding_mode_t;
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/image/dl_image.hpp b/tools/sdk/esp32c3/include/esp-face/include/image/dl_image.hpp
new file mode 100644
index 00000000000..4a974df063a
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/image/dl_image.hpp
@@ -0,0 +1,439 @@
+#pragma once
+
+#include <stdint.h>
+#include <stdlib.h>
+#include <math.h>
+#include <vector>
+#include "dl_define.hpp"
+#include "dl_variable.hpp"
+#include "dl_math_matrix.hpp"
+
+namespace dl
+{
+    namespace image
+    {
+        typedef enum
+        {
+            IMAGE_RESIZE_BILINEAR = 0, /*<! Resize image by taking bilinear of four pixels */
+            IMAGE_RESIZE_MEAN = 1,     /*<! Resize image by taking mean of four pixels */
+            IMAGE_RESIZE_NEAREST = 2   /*<! Resize image by taking the nearest pixel */
+        } resize_type_t;
+
+        /**
+         * @brief Convert RGB888 pixel to Gray.
+         * 
+         * @param red   red value
+         * @param green green value
+         * @param blue  blue value
+         * @return gray value
+         */
+        inline uint8_t convert_pixel_rgb888_to_gray(int red, int green, int blue)
+        {
+            int temp = (red * 38 + green * 75 + blue * 15) >> 7;
+            return DL_CLIP(temp, 0, 255);
+        }
+
+        /**
+         * @brief Convert RGB565 pixel to RGB888.
+         * 
+         * @tparam T supports all integer types
+         * @param input  pixel value in RGB565
+         * @param output pixel value in RGB888
+         */
+        template <typename T>
+        inline void convert_pixel_rgb565_to_rgb888(uint16_t input, T *output)
+        {
+            output[0] = (input & 0x1F00) >> 5;                           // blue
+            output[1] = ((input & 0x7) << 5) | ((input & 0xE000) >> 11); // green
+            output[2] = input & 0xF8;                                    // red
+        }
+
+        /**
+         * @brief Convert RGB565 pixel to Gray.
+         * 
+         * @param input pixel value in RGB565
+         * @return pixel value in Gray
+         */
+        inline uint8_t convert_pixel_rgb565_to_gray(uint16_t input)
+        {
+            int blue = (input & 0x1F00) >> 5;                            // blue
+            int green = ((input & 0x7) << 5) | ((input & 0xE000) >> 11); // green
+            int red = input & 0xF8;                                      // red
+
+            return convert_pixel_rgb888_to_gray(red, green, blue);
+        }
+
+        /**
+         * @brief Crop a patch from image and resize and store to destination image.
+         * If the cropping box is out of image, destination image will be padded with edge.
+         * 
+         * The outer rectangle is the entire output image.
+         * The inner rectangle is where the resized image will be stored.
+         * In other world, this function could help you do padding while resize image.
+         *               ___________________________(dst_w)__________________
+         *              |         ___________________________                |
+         *              |        |(x_start, y_start)         |               | 
+         *              |        |                           |               | 
+         *              |        |                           |               | 
+         *       (dst_h)|        |                           |               | 
+         *              |        |                           |               | 
+         *              |        |                           |               | 
+         *              |        |___________________________|(x_end, y_end) | 
+         *              |____________________________________________________| 
+         * 
+         * @tparam T suppot all integer types
+         * @param dst_image     pointer of destination(output) image
+         * @param dst_width     destination image width
+         * @param dst_channel   destination image channel number
+         * @param dst_y_start   start y of resized image in destination image
+         * @param dst_y_end     end y of resized image in destination image
+         * @param dst_x_start   start x of resized image in destination image
+         * @param dst_x_end     end x of resized image in destination image
+         * @param src_image     pointer of source image
+         * @param src_height    source image height
+         * @param src_width     source image width
+         * @param src_channel   source image channel
+         * @param src_y_start   start y of resized image in source image
+         * @param src_y_end     end y of resized image in source image
+         * @param src_x_start   start x of resized image in source image
+         * @param src_x_end     end x of resized image in source image
+         * @param resize_type   one of IMAGE_RESIZE_BILINEAR or IMAGE_RESIZE_MEAN or IMAGE_RESIZE_NEAREST
+         * @param shift_left    bit left shift number implemented on output
+         */
+        template <typename T>
+        void crop_and_resize(T *dst_image,
+                             int dst_width,
+                             int dst_channel,
+                             int dst_y_start, int dst_y_end,
+                             int dst_x_start, int dst_x_end,
+                             uint16_t *src_image,
+                             int src_height,
+                             int src_width,
+                             int src_channel,
+                             int src_y_start, int src_y_end,
+                             int src_x_start, int src_x_end,
+                             resize_type_t resize_type = IMAGE_RESIZE_NEAREST,
+                             int shift_left = 0);
+
+        /**
+         * @brief Crop a patch from image and resize and store to destination image.
+         * If the cropping box is out of image, destination image will be padded with edge.
+         * 
+         * The outer rectangle is the entire output image.
+         * The inner rectangle is where the resized image will be stored.
+         * In other world, this function could help you do padding while resize image.
+         *               ___________________________(dst_w)__________________
+         *              |         ___________________________                |
+         *              |        |(x_start, y_start)         |               | 
+         *              |        |                           |               | 
+         *              |        |                           |               | 
+         *       (dst_h)|        |                           |               | 
+         *              |        |                           |               | 
+         *              |        |                           |               | 
+         *              |        |___________________________|(x_end, y_end) | 
+         *              |____________________________________________________| 
+         * 
+         * @tparam T suppot all integer types
+         * @param dst_image     pointer of destination(output) image
+         * @param dst_width     destination image width
+         * @param dst_channel   destination image channel number
+         * @param dst_y_start   start y of resized image in destination image
+         * @param dst_y_end     end y of resized image in destination image
+         * @param dst_x_start   start x of resized image in destination image
+         * @param dst_x_end     end x of resized image in destination image
+         * @param src_image     pointer of source image
+         * @param src_height    source image height
+         * @param src_width     source image width
+         * @param src_channel   source image channel
+         * @param src_y_start   start y of resized image in source image
+         * @param src_y_end     end y of resized image in source image
+         * @param src_x_start   start x of resized image in source image
+         * @param src_x_end     end x of resized image in source image
+         * @param resize_type   one of IMAGE_RESIZE_BILINEAR or IMAGE_RESIZE_MEAN or IMAGE_RESIZE_NEAREST
+         * @param shift_left    bit left shift number implemented on output
+         */
+        template <typename T>
+        void crop_and_resize(T *dst_image,
+                             int dst_width,
+                             int dst_channel,
+                             int dst_y_start, int dst_y_end,
+                             int dst_x_start, int dst_x_end,
+                             uint8_t *src_image,
+                             int src_height,
+                             int src_width,
+                             int src_channel,
+                             int src_y_start, int src_y_end,
+                             int src_x_start, int src_x_end,
+                             resize_type_t resize_type = IMAGE_RESIZE_NEAREST,
+                             int shift_left = 0);
+
+        /**
+         * @brief Draw a filled rectangle on RGB888 image.
+         * 
+         * @param image        pointer of input image
+         * @param image_height height of input image
+         * @param image_width  width of input image
+         * @param x1           left up corner x
+         * @param y1           left up corner y
+         * @param x2           right bottom corner x
+         * @param y2           right bottom corner y
+         * @param color        0x    00|       00|       00|       00
+         *                     reserved|channel 0|channel 1|channel 2 
+         */
+        void draw_filled_rectangle(uint8_t *image, const uint32_t image_height, const uint32_t image_width,
+                                   uint32_t x1, uint32_t y1, uint32_t x2, uint32_t y2,
+                                   const uint32_t color = 0x00FF0000);
+
+        /**
+         * @brief Draw a filled rectangle on RGB565 image.
+         * 
+         * @param image        pointer of input image
+         * @param image_height height of input image
+         * @param image_width  width of input image
+         * @param x1           left up corner x
+         * @param y1           left up corner y
+         * @param x2           right bottom corner x
+         * @param y2           right bottom corner y
+         * @param color        0b         000|    00000|    00000|           000
+         *                     channel 1[2:0]|channel 0|channel 2|channel 1[5:3] 
+         */
+        void draw_filled_rectangle(uint16_t *image, const uint32_t image_height, const uint32_t image_width,
+                                   uint32_t x1, uint32_t y1, uint32_t x2, uint32_t y2,
+                                   const uint16_t color = 0b0001111100000000);
+
+        /**
+         * @brief Draw a point on RGB888 image.
+         * 
+         * @param image        pointer of input image
+         * @param image_height height of input image
+         * @param image_width  width of input image
+         * @param x            point x
+         * @param y            point y
+         * @param size         size of point
+         * @param color        0x    00|       00|       00|       00
+         *                     reserved|channel 0|channel 1|channel 2 
+         */
+        void draw_point(uint8_t *image, const uint32_t image_height, const uint32_t image_width,
+                        const uint32_t x, const uint32_t y, const uint32_t size,
+                        const uint32_t color = 0x00FF0000);
+
+        /**
+         * @brief Draw a point on RGB565 image.
+         * 
+         * @param image        pointer of input image
+         * @param image_height height of input image
+         * @param image_width  width of input image
+         * @param x            point x
+         * @param y            point y
+         * @param size         size of point
+         * @param color        0b         000|    00000|    00000|           000
+         *                     channel 1[2:0]|channel 0|channel 2|channel 1[5:3] 
+         */
+        void draw_point(uint16_t *image, const uint32_t image_height, const uint32_t image_width,
+                        const uint32_t x, const uint32_t y, const uint32_t size,
+                        uint16_t color = 0b0001111100000000);
+
+        /**
+         * @brief Draw a hollow rectangle on RGB888 image.
+         * 
+         * @param image        pointer of input image
+         * @param image_height height of input image
+         * @param image_width  width of input image
+         * @param x1           left up corner x
+         * @param y1           left up corner y
+         * @param x2           right bottom corner x
+         * @param y2           right bottom corner y
+         * @param color        0x    00|       00|       00|       00
+         *                     reserved|channel 0|channel 1|channel 2 
+         */
+        void draw_hollow_rectangle(uint8_t *image, const uint32_t image_height, const uint32_t image_width,
+                                   uint32_t x1, uint32_t y1, uint32_t x2, uint32_t y2,
+                                   uint32_t color = 0x00FF0000);
+
+        /**
+         * @brief Draw a hollow rectangle on RGB565 image.
+         * 
+         * @param image        pointer of input image
+         * @param image_height height of input image
+         * @param image_width  width of input image
+         * @param x1           left up corner x
+         * @param y1           left up corner y
+         * @param x2           right bottom corner x
+         * @param y2           right bottom corner y
+         * @param color        0b         000|    00000|    00000|           000
+         *                     channel 1[2:0]|channel 0|channel 2|channel 1[5:3] 
+         */
+        void draw_hollow_rectangle(uint16_t *image, const uint32_t image_height, const uint32_t image_width,
+                                   uint32_t x1, uint32_t y1, uint32_t x2, uint32_t y2,
+                                   const uint16_t color = 0b0001111100000000);
+
+        /**
+         * @brief Detect target moving by activated detection point number. Each cross in the figure below is a detection point.
+         * Once abs(frame_1_detection_point[i] - frame_2_detection_point[i]) > threshold, this detection point is activated.
+         * This function will return the number of activated detection point.
+         * 
+         *         __stride__________________________
+         *         |        |        |        |   |
+         *  stride |        |        |        |   |
+         *         |        |        |        |   |
+         *         |________|________|________|   |
+         *         |        |        |        |   |
+         *         |        |        |        |   |
+         *         |        |        |        |   |
+         *         |________|________|________| height
+         *         |        |        |        |   |
+         *         |        |        |        |   |
+         *         |        |        |        |   |
+         *         |________|________|________|   |
+         *         |        |        |        |   |
+         *         |        |        |        |   |
+         *         |        |        |        |   |
+         *         |________|________|________|___|___
+         *         |                          |
+         *         |__________width___________|
+         *         |                          |
+         * 
+         * Time consumption:
+         * Frame shape = (240, 240)
+         * Both frame are in PSRAM
+         * On ESP32-S3 with CPU 240MHz, QSPI 80MHz
+         * 
+         * stride  latency
+         *      1  28316us
+         *      2   8770us
+         *      4   3622us
+         *      8   1990us
+         *     16    880us
+         *     32    260us
+         * 
+         * 
+         * In a application, outside this function, threshold of activated detection point number is needed.
+         * Once activated detection point number > number_threshold, this two frame are judged target moved.
+         * How to determine the number_threshold?
+         * Let's assume that the minimize shape of target is (target_min_height, target_max_width).
+         * Then, the number_threshold = [target_min_height / stride] * [target_max_width / stride] * ratio,
+         * where ratio is in (0, 1), the smaller the ratio is, the more sensitive the detector is, the more false detected.
+         * 
+         * 
+         * @param f1        one frame in RGB565
+         * @param f2        another frame in RGB565
+         * @param height    height of frame
+         * @param width     width of frame
+         * @param stride    stride of detection point, the smaller the stride is, the more reliable the detector is.
+         * @param threshold activation threshold of each detection point
+         * @return activated detection point number 
+         */
+        uint32_t get_moving_point_number(uint16_t *f1, uint16_t *f2, const uint32_t height, const uint32_t width, const uint32_t stride, const uint32_t threshold = 5);
+
+        /**
+         * @brief Detect target moving by activated detection point number. Each cross in the figure below is a detection point.
+         * Once abs(frame_1_detection_point[i] - frame_2_detection_point[i]) > threshold, this detection point is activated.
+         * This function will return the number of activated detection point.
+         * 
+         *         __stride__________________________
+         *         |        |        |        |   |
+         *  stride |        |        |        |   |
+         *         |        |        |        |   |
+         *         |________|________|________|   |
+         *         |        |        |        |   |
+         *         |        |        |        |   |
+         *         |        |        |        |   |
+         *         |________|________|________| height
+         *         |        |        |        |   |
+         *         |        |        |        |   |
+         *         |        |        |        |   |
+         *         |________|________|________|   |
+         *         |        |        |        |   |
+         *         |        |        |        |   |
+         *         |        |        |        |   |
+         *         |________|________|________|___|___
+         *         |                          |
+         *         |__________width___________|
+         *         |                          |
+         * 
+         * 
+         * In a application, outside this function, threshold of activated detection point number is needed.
+         * Once activated detection point number > number_threshold, this two frame are judged target moved.
+         * How to determine the number_threshold?
+         * Let's assume that the minimize shape of target is (target_min_height, target_max_width).
+         * Then, the number_threshold = [target_min_height / stride] * [target_max_width / stride] * ratio,
+         * where ratio is in (0, 1), the smaller the ratio is, the more sensitive the detector is, the more false detected.
+         * 
+         * 
+         * @param f1        one frame in RGB888
+         * @param f2        another frame in RGB888
+         * @param height    height of frame
+         * @param width     width of frame
+         * @param stride    stride of detection point, the smaller the stride is, the more reliable the detector is.
+         * @param threshold activation threshold of each detection point
+         * @return activated detection point number 
+         */
+        uint32_t get_moving_point_number(uint8_t *f1, uint8_t *f2, const uint32_t height, const uint32_t width, const uint32_t stride, const uint32_t threshold = 5);
+
+        /**
+         * @brief Apply an affine transformation to an image.
+         * 
+         * @tparam T 
+         * @param input     the input image.
+         * @param output    the output image.
+         * @param M_inv     the inverse transformation matrix.
+         */
+        template <typename T>
+        void warp_affine(dl::Tensor<T> *input, dl::Tensor<T> *output, dl::math::Matrix<float> *M_inv);
+
+        /**
+         * @brief Apply an affine transformation to an image.
+         * 
+         * @tparam T 
+         * @param input    the pointer of the input image.
+         * @param shape    the shape of the input image.
+         * @param output   the output image.
+         * @param M_inv    the inverse transformation matrix.
+         */
+        template <typename T>
+        void warp_affine(uint16_t *input, std::vector<int> shape, dl::Tensor<T> *output, dl::math::Matrix<float> *M_inv);
+
+        /**
+         * @brief Get the otsu thresh object.
+         * 
+         * @param image  the gray image.
+         * @return uint8_t the otsu thresh.
+         */
+        uint8_t get_otsu_thresh(Tensor<uint8_t> &image);
+
+        /**
+         * @brief Convert RGB image to gray image
+         * 
+         * @param image  input image
+         * @param bgr    true: the image is in BGR format
+         *               false: the image is in RGB format
+         * @return Tensor<uint8_t>* output image in gray format
+         */
+        Tensor<uint8_t> *rgb2gray(Tensor<uint8_t> &image, bool bgr = false);
+
+        /**
+         * @brief Convert RGB image to LAB image
+         * 
+         * @param image  input image
+         * @param bgr    true: the image is in BGR format
+         *               false: the image is in RGB format 
+         * @param fast   true: use the fast alogrithm， but the accuracy will be reduced
+         *               false: do not use the fast alogrithm
+         * @return Tensor<uint8_t>* output image in LAB foramt
+         */
+        Tensor<uint8_t> *rgb2lab(Tensor<uint8_t> &image, bool bgr = false, bool fast = true);
+
+        /**
+         * @brief Convert RGB image to HSV image
+         * 
+         * @param image   input image
+         * @param bgr     true: the image is in BGR format
+         *                false: the image is in RGB format 
+         * @param fast    true: use the fast alogrithm， but the accuracy will be reduced
+         *                false: do not use the fast alogrithm
+         * @return Tensor<uint8_t>* output image in HSV format
+         */
+        Tensor<uint8_t> *rgb2hsv(Tensor<uint8_t> &image, bool bgr = false, bool fast = true);
+
+    } // namespace image
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_add2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_add2d.hpp
new file mode 100644
index 00000000000..c43282b42de
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_add2d.hpp
@@ -0,0 +1,145 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn_add2d.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Activation(Add2D(input0, input1)).
+         * NOTE: addition is element-wise, i.e., output[i,j,k] = input0[i,j,k] + input1[i,j,k]
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class Add2D : public Layer
+        {
+        private:
+            const Activation<feature_t> *activation; /*<! activation of add2d, if you don't specify anything, no activation is applied >*/
+            const int output_exponent;               /*<! exponent of output >*/
+            Tensor<feature_t> *output;               /*<! output ptr of add2d >*/
+            bool inplace;                            /*<! true: the output will store to input0
+                                                          false: the output will store to a separate memory >*/
+            std::vector<int> output_shape;           /*<! output shape of add2d >*/
+
+        public:
+            /**
+             * @brief Construct a new Add2D object.
+             * 
+             * @param output_exponent exponent of output
+             * @param activation      activation of add2d, if you don't specify anything, no activation is applied
+             * @param name            name of add2d
+             * @param inplace         true: the output will store to input0
+             *                        false: the output will store to a separate memory
+             */
+            Add2D(const int output_exponent, const Activation<feature_t> *activation = NULL, const char *name = "Add2D", bool inplace = false) : Layer(name),
+                                                                                                                                                 activation(activation),
+                                                                                                                                                 output_exponent(output_exponent),
+                                                                                                                                                 output(NULL),
+                                                                                                                                                 inplace(inplace),
+                                                                                                                                                 output_shape({}) {}
+
+            /**
+             * @brief Destroy the Add2D object
+             */
+            ~Add2D()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape.
+             * NOTE: input0.shape must equal to input1.shape.
+             * 
+             * @param input0 as one input
+             * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
+            {
+                assert(input0.is_same_shape(input1));
+                this->output_shape = input0.shape;
+
+                if (!this->inplace)
+                {
+                    if (this->output == NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(input0.shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input0;
+                }
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+            * @brief Get the output
+            * 
+            * @return Tensor<feature_t>& Add2D result
+            */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Add2D operation.
+             * 
+             * @param input0      as one input
+             * @param input1      as another input
+             * @param assign_core not effective yet
+             * @return Tensor<feature_t>& added result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input0, Tensor<feature_t> &input1, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
+                    this->output->set_exponent(this->output_exponent);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                    DL_LOG_LAYER_LATENCY_START();
+                    nn::add2d(*this->output, input0, input1, this->activation, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "add2d");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    nn::add2d(*this->output, input0, input1, this->activation, assign_core, this->output_exponent);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "add2d");
+                }
+
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_avg_pool2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_avg_pool2d.hpp
new file mode 100644
index 00000000000..8a9aaa8dfbe
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_avg_pool2d.hpp
@@ -0,0 +1,161 @@
+#pragma once
+
+#include <vector>
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn_avg_pool2d.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief AvgPool2D(input).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class AvgPool2D : public Layer
+        {
+        private:
+            const int output_exponent;         /*<! exponent of output >*/
+            std::vector<int> filter_shape;     /*<! filter shape in [filter_height, filter_width] >*/
+            const int stride_y;                /*<! stride in height >*/
+            const int stride_x;                /*<! stride in width >*/
+            const padding_type_t padding_type; /*<! one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN >*/
+            std::vector<int> padding;          /*<! padding size needed in [top, bottom, left, right] of this operation >*/
+            Tensor<feature_t> *output;         /*<! output ptr of AvgPool2D >*/
+            std::vector<int> output_shape;     /*<! output shape of AvgPool2D >*/
+
+        public:
+            /**
+             * @brief Construct a new AvgPool2D object.
+             * 
+             * @param output_exponent exponent of output
+             * @param filter_shape    filter shape in [filter_height, filter_width]
+             * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN or PADDING_NOT_SET,
+             *                        - PADDING_VALID means no padding
+             *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
+             *                        such that output has the same height/width dimension as the input,
+             *                        - PADDING_SAME_END results padding in TensorFlow style
+             *                        - PADDING_SAME_BEGIN results padding in MXNET style
+             *                        - PADDING_NOT_SET means padding with the specific "padding" value below. 
+             * @param padding         if padding_type is PADDING_NOT_SET, this value will be used as padding size. 
+             *                        the shape must be 4, the value of each position is: [padding top, padding bottom, padding left, padding right]
+             * @param stride_y        stride in height
+             * @param stride_x        stride in width
+             * @param name            name of layer
+             */
+            AvgPool2D(const int output_exponent,
+                      const std::vector<int> filter_shape,
+                      const padding_type_t padding_type = PADDING_VALID,
+                      std::vector<int> padding = {},
+                      const int stride_y = 1,
+                      const int stride_x = 1,
+                      const char *name = "AvgPool2D") : Layer(name),
+                                                        output_exponent(output_exponent),
+                                                        filter_shape(filter_shape),
+                                                        padding_type(padding_type),
+                                                        padding(padding),
+                                                        stride_y(stride_y),
+                                                        stride_x(stride_x),
+                                                        output_shape({})
+            {
+                this->output = new Tensor<feature_t>;
+                if (this->padding_type == PADDING_NOT_SET)
+                {
+                    assert(this->padding.size() == 4);
+                }
+            }
+
+            /**
+            * @brief Destroy the AvgPool2D object.
+            * 
+            */
+            ~AvgPool2D()
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape and padding.
+             * 
+             * @param input as an input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                assert(input.shape[0] > 0);
+                assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
+
+                this->output_shape = nn::get_output_shape(input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type, false, this->padding);
+                this->output->set_shape(this->output_shape);
+                this->output->set_exponent(this->output_exponent);
+
+                if (this->padding_type != PADDING_NOT_SET)
+                {
+                    this->padding = nn::get_pad_size(this->output_shape, input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
+                }
+
+                this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& AvgPool2D result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call AvgPool2D operation
+             * 
+             * @param input           as an input
+             * @param autoload_enable one of true or false, 
+             *                        - true: load input and output from PSRAM to CACHE automatically
+             *                        - false: do not
+             * @return AvgPool2D result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input, uint8_t autoload_enable = 0)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
+                this->output->set_exponent(this->output_exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                if (autoload_enable)
+                {
+                    dl::tool::cache::autoload_func((uint32_t)(this->output->element), this->output->get_size() * sizeof(feature_t),
+                                                   (uint32_t)(input.element), input.get_size() * sizeof(feature_t));
+                }
+
+                DL_LOG_LAYER_LATENCY_START();
+                nn::avg_pool2d(*this->output, input, this->padding, this->filter_shape, this->stride_y, this->stride_x);
+                DL_LOG_LAYER_LATENCY_END(this->name, "avg_pool2d");
+
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_base.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_base.hpp
new file mode 100644
index 00000000000..b265b454538
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_base.hpp
@@ -0,0 +1,56 @@
+#pragma once
+#include "dl_tool.hpp"
+#include "dl_tool_cache.hpp"
+#include <iostream>
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Base class for layer.
+         * 
+         */
+        class Layer
+        {
+        public:
+            char *name; /*<! name of layer >*/
+
+            /**
+             * @brief Construct a new Layer object.
+             * 
+             * @param name name of layer.
+             */
+            Layer(const char *name = NULL);
+
+            /**
+             * @brief Destroy the Layer object. Return resource.
+             * 
+             */
+            ~Layer();
+        };
+    } // namespace layer
+} // namespace dl
+
+#if DL_LOG_LAYER_LATENCY
+/**
+ * @brief Initialize.
+ */
+#define DL_LOG_LAYER_LATENCY_INIT() dl::tool::Latency latency
+
+/**
+ * @brief Time starts.
+ */
+#define DL_LOG_LAYER_LATENCY_START() latency.start()
+
+/**
+ * @brief Time ends and printed.
+ */
+#define DL_LOG_LAYER_LATENCY_END(prefix, key) \
+    latency.end();                            \
+    latency.print(prefix, key)
+#else
+#define DL_LOG_LAYER_LATENCY_INIT()
+#define DL_LOG_LAYER_LATENCY_START()
+#define DL_LOG_LAYER_LATENCY_END(prefix, key)
+#endif
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_concat.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_concat.hpp
new file mode 100644
index 00000000000..35ebe652e53
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_concat.hpp
@@ -0,0 +1,139 @@
+#pragma once
+
+#include <assert.h>
+#include <vector>
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+#include "dl_nn_concat.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Concat(input1, input2, input3, ...).
+         * 
+         * @tparam feature_t support all kinds of integer and float data type
+         */
+        template <typename feature_t>
+        class Concat : Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            int axis;                      /*<! The axis along which the Tensor will be concatenated. >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Concat >*/
+            std::vector<int> output_shape; /*<! output shape of Concat >*/
+        public:
+            /**
+             * @brief Construct a new Concat object.
+             * 
+             * @param name name of layer
+             * @param axis The axis along which the Tensor will be concatenated.
+             */
+            Concat(int axis, const char *name = "Concat") : Layer(name), axis(axis), output_shape({})
+            {
+                this->output = new Tensor<feature_t>;
+            }
+
+            /**
+             * @brief Destroy the Concat object
+             */
+            ~Concat()
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Collect inputs' channel and memory offset, called in Model.build().
+             * 
+             * @param args pointers of concatenated Tensor
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(std::vector<Tensor<feature_t> *> args, bool print_shape = false)
+            {
+                assert(args.size() > 1);
+                int shape_size = args[0]->shape.size();
+
+                if (this->axis < 0)
+                {
+                    this->axis = shape_size + this->axis;
+                }
+                assert((this->axis < shape_size) && (this->axis > -1));
+
+                int output_shape_axis = args[0]->shape[this->axis];
+
+                for (int i = 1; i < args.size(); i++)
+                {
+                    assert(shape_size == args[i]->shape.size());
+                    assert(args[i]->exponent == args[i - 1]->exponent);
+                    output_shape_axis += args[i]->shape[this->axis];
+
+                    for (int j = 0; j < shape_size; j++)
+                    {
+                        if (j != this->axis)
+                        {
+                            assert(args[i]->shape[j] == args[i - 1]->shape[j]);
+                        }
+                    }
+                }
+
+                this->output_exponent = args[0]->exponent;
+                this->output_shape = args[0]->shape;
+                this->output_shape[this->axis] = output_shape_axis;
+
+                this->output->set_shape(this->output_shape);
+                this->output->set_exponent(this->output_exponent);
+                this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Call Concat operation
+             * 
+             * @param inputs                the pointers of inputs 
+             * @param free_inputs           true: free the inputs after call
+             *                              false: do not free inputs
+             * @return Tensor<feature_t>&   concat result
+             */
+            Tensor<feature_t> &call(std::vector<Tensor<feature_t> *> inputs, bool free_inputs = false)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
+                this->output->set_exponent(this->output_exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                DL_LOG_LAYER_LATENCY_START();
+                nn::concat(*this->output, inputs, this->axis, free_inputs);
+                DL_LOG_LAYER_LATENCY_END(this->name, "concat");
+                return *this->output;
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>&  Concat result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_concat2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_concat2d.hpp
new file mode 100644
index 00000000000..a086f1ccfde
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_concat2d.hpp
@@ -0,0 +1,179 @@
+#pragma once
+
+#include <assert.h>
+#include <vector>
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Concat2D(input1, input2, input3, ...).
+         * 
+         * @tparam feature_t support all kinds of integer and float data type
+         */
+        template <typename feature_t>
+        class Concat2D : Layer
+        {
+        private:
+            std::vector<Tensor<feature_t> *> output_vec; /*<! pointers of concatenated inputs >*/
+            std::vector<int> offset;                     /*<! memory offset of each concatenated inputs in entire element >*/
+            std::vector<int> channel;                    /*<! channel of concatenated inputs >*/
+            Tensor<feature_t> *output;                  /*<! output ptr of Concat2D >*/
+            int output_exponent;                        /*<! exponent of output >*/
+        public: 
+
+            /**
+             * @brief Construct a new Concat2D object.
+             * 
+             * @param name name of layer
+             */
+            Concat2D(const char *name = NULL) : Layer(name) {
+                this->output = new Tensor<feature_t>;
+            }
+
+            /**
+             * @brief Destroy the Concat2D object
+             */
+            ~Concat2D() 
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Collect inputs' channel and memory offset, called in Model.build().
+             * 
+             * @param args pointers of concatenated Tensor
+             */
+            void build(std::vector<Tensor<feature_t> *> args)
+            {
+                assert(args.size() > 0);
+
+                this->output_vec = args;
+
+                this->offset = std::vector<int>(args.size());
+                this->channel = std::vector<int>(args.size());
+
+                this->output_exponent = args[0]->exponent;
+                this->offset[0] = 0;
+                this->channel[0] = args[0]->shape[2];
+                std::vector<int> output_shape = args[0]->shape;
+
+                for (int i = 1; i < args.size(); i++)
+                {
+                    assert(output_shape[0] == args[i]->shape[0]); // height
+                    assert(output_shape[1] == args[i]->shape[1]); // width
+                    // assert(this->output_exponent == args[i]->exponent); // exponent
+
+                    this->offset[i] = output_shape[2];
+                    this->channel[i] = args[i]->shape[2];
+                    output_shape[2] += args[i]->shape[2];
+                }
+                this->output->set_shape(output_shape);
+                this->output->set_exponent(this->output_exponent);
+                this->output->free_element();
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>&  Concat2d result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Get the maximum padding among inputs and output-> Then, set to this->output. Called at the end of Model.build().
+             * NOTE: Some special situations like C = Concat2D_1(A, B), E = Concat2D_2(C, D), where A, B, C, D, E are Tensor.
+             *         For avoiding memory copy, we apply an entire element for E, and take it apart for A, B, D. 
+             *         A, B, C, D and E will become other layer's inputs so that result different size of padding.
+             *         For get the maximum padding, we should call at the end of Model.build(),
+             *              Concat2D_1.backward();  // max_padding_temp = get_max_padding(A, B, C), padding of A, B and C are set to max_padding_temp.
+             *              Concat2D_2.backward();  // max_padding = get_max_padding(max_padding_temp, get_max_padding(D, E)) , padding of C, D and E are set to max_padding.
+             *                                         However, padding of A and B is still max_padding_temp.
+             *              Concat2D_1.backward();  // padding of A and B are set to max_padding.
+             *         Or,
+             *              Concat2D_2.backward();
+             *              Concat2D_1.backward();
+             *              Concat2D_2.backward();
+             */
+            void backward()
+            {
+                std::vector<int> max_padding = this->output->padding;
+                int max_channel_with_padding = this->output->shape_with_padding[2];
+                for (int i = 0; i < this->output_vec.size(); i++)
+                {
+                    for (int j = 0; j < max_padding.size(); j++)
+                    {
+                        max_padding[j] = DL_MAX(max_padding[j], this->output_vec[i]->padding[j]);
+                    }
+                    max_channel_with_padding = DL_MAX(max_channel_with_padding, this->output_vec[i]->shape_with_padding[2]);
+                }
+
+                this->output->set_padding_size(max_padding);
+                this->output->shape_with_padding[2] = max_channel_with_padding;
+                for (int i = 0; i < this->output_vec.size(); i++)
+                {
+                    this->output_vec[i]->set_padding_size(max_padding);
+                    this->output_vec[i]->shape_with_padding[2] = max_channel_with_padding;
+#if CONFIG_DEBUG_MODE
+                    assert(this->output->shape_with_padding[0] == this->output_vec[i]->shape_with_padding[0]);
+                    assert(this->output->shape_with_padding[1] == this->output_vec[i]->shape_with_padding[1]);
+                    assert(this->output->shape_with_padding[2] == this->output_vec[i]->shape_with_padding[2]);
+#endif
+                }
+            }
+
+            /**
+             * @brief Calloc an entire element for concatnate result. Take the entire element apart and deliver element pointers to concatenated layer.
+             * NOTE: For example, C = Concat2D(A, B). We apply an entire element for C and deliver two element pointers to A and B.
+             *       Let's assume that A result is produced first. We should call Concat2D.calloc_element() just before A result is produced
+             *       to make sure the element of A is ready and could be filled.
+             */
+            void calloc_element()
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                this->output->calloc_element();
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                DL_LOG_LAYER_LATENCY_START();
+                for (int i = 0; i < this->offset.size(); i++)
+                {
+                    this->output_vec[i]->element = this->output->element + this->offset[i];
+                    this->output_vec[i]->set_auto_free(false);
+                }
+                DL_LOG_LAYER_LATENCY_END(this->name, "deliver");
+            }
+
+            void apply_element()
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                this->output->apply_element();
+                this->output->set_exponent(this->output_exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                DL_LOG_LAYER_LATENCY_START();
+                for (int i = 0; i < this->offset.size(); i++)
+                {
+                    this->output_vec[i]->element = this->output->element + this->offset[i];
+                    this->output_vec[i]->set_auto_free(false);
+                }
+                DL_LOG_LAYER_LATENCY_END(this->name, "deliver");
+            }
+        };
+    } // namespace layer
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_conv2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_conv2d.hpp
new file mode 100644
index 00000000000..038dd6cd79f
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_conv2d.hpp
@@ -0,0 +1,186 @@
+#pragma once
+
+#include "dl_nn_conv2d.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Activation(Conv2D(input, filter) + bias).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @tparam bias_t supports int16_t and int8_t, must specify when using int8 per-channel quantization
+         *         - int16_t: for int16 quantization and int8 per-channel quantization
+         *         - int8_t: for int8 per-tensor quantization
+         */
+        template <typename feature_t, typename bias_t = feature_t>
+        class Conv2D : public Layer
+        {
+        private:
+            const int output_exponent;               /*<! exponent of output >*/
+            const Filter<feature_t> *filter;         /*<! filter of Conv2D >*/
+            const int stride_y;                      /*<! stride in height >*/
+            const int stride_x;                      /*<! stride in width >*/
+            const padding_type_t padding_type;       /*<! one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN >*/
+            const Bias<bias_t> *bias;                /*<! bias of Conv2D, if you don't specify anything, no bias is added >*/
+            const Activation<feature_t> *activation; /*<! activation of Conv2D, if you don't specify anything, no activation is applied >*/
+            std::vector<int> padding;                /*<! padding size needed in [top, bottom, left, right] of this operation >*/
+            Tensor<feature_t> *output;               /*<! output ptr of Conv2D >*/
+            std::vector<int> output_shape;           /*<! output shape of Conv2D >*/
+
+        public:
+            /**
+             * @brief Construct a new Conv2D object.
+             * 
+             * @param output_exponent exponent of output
+             * @param filter          filter of Conv2D
+             * @param bias            bias of Conv2D, if you don't specify anything, no bias is added
+             * @param activation      activation of Conv2D, if you don't specify anything, no activation is applied
+             * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN or PADDING_NOT_SET,
+             *                        - PADDING_VALID means no padding
+             *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
+             *                        such that output has the same height/width dimension as the input,
+             *                        - PADDING_SAME_END results padding in TensorFlow style
+             *                        - PADDING_SAME_BEGIN results padding in MXNET style
+             *                        - PADDING_NOT_SET means padding with the specific "padding" value below. 
+             * @param padding         if padding_type is PADDING_NOT_SET, this value will be used as padding size. 
+             *                        the shape must be 4, the value of each position is: [padding top, padding bottom, padding left, padding right]
+             * @param stride_y        stride in height
+             * @param stride_x        stride in width
+             * @param name            name of layer
+             */
+            Conv2D(const int output_exponent,
+                   const Filter<feature_t> *filter,
+                   const Bias<bias_t> *bias = NULL,
+                   const Activation<feature_t> *activation = NULL,
+                   const padding_type_t padding_type = PADDING_VALID,
+                   std::vector<int> padding = {},
+                   const int stride_y = 1,
+                   const int stride_x = 1,
+                   const char *name = "Conv2D") : Layer(name),
+                                                  output_exponent(output_exponent),
+                                                  filter(filter),
+                                                  stride_y(stride_y),
+                                                  stride_x(stride_x),
+                                                  padding_type(padding_type),
+                                                  bias(bias),
+                                                  activation(activation),
+                                                  padding(padding),
+                                                  output_shape({})
+            {
+                this->output = new Tensor<feature_t>;
+                if (this->padding_type == PADDING_NOT_SET)
+                {
+                    assert(this->padding.size() == 4);
+                }
+            }
+
+            /**
+             * @brief Destroy the Conv2D object.
+             * 
+             */
+            ~Conv2D()
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output padding and input padding.
+             * 
+             * @param input as an input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                assert(input.shape[0] > 0);
+                assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
+                assert(this->filter->shape.size() == 4);
+                assert(input.shape[2] == this->filter->shape[2]);
+
+                this->output_shape = nn::get_output_shape(input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type, true, this->padding);
+                this->output->set_shape(this->output_shape);
+                this->output->set_exponent(this->output_exponent);
+                this->output->free_element();
+                if (this->padding_type != PADDING_NOT_SET)
+                {
+                    this->padding = nn::get_pad_size(this->output_shape, input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type);
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Conv2D result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Conv2D operation
+             * 
+             * @param input           as an input.
+             * @param autoload_enable one of true or false, 
+             *                        - true: load input and output from PSRAM to CACHE automatically
+             *                        - false: do not
+             * @param assign_core     not effective yet
+             * @return Conv2D result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input, bool autoload_enable = false, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
+                this->output->set_exponent(this->output_exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                if (autoload_enable)
+                {
+                    dl::tool::cache::autoload_func((uint32_t)(this->output->element), this->output->get_size() * sizeof(feature_t),
+                                                   (uint32_t)(input.element), input.get_size() * sizeof(feature_t));
+                }
+
+                DL_LOG_LAYER_LATENCY_START();
+                nn::conv2d(*this->output, input, this->padding, *(this->filter), this->stride_y, this->stride_x, this->bias, this->activation, assign_core);
+                DL_LOG_LAYER_LATENCY_END(this->name, "conv2d");
+                return *this->output;
+            }
+
+            /**
+             * @brief Preload the filter to Cache.
+             * NOTE: Call this layer's preload() before previous layer's call() such that filter could be loaded while previous layer is doing calculation.
+             */
+            void preload()
+            {
+                size_t size = sizeof(feature_t);
+                int shape_size = this->filter->shape.size();
+                for (int i = 0; i < shape_size; ++i)
+                {
+                    size *= filter->shape[i];
+                }
+                dl::tool::cache::preload_func((uint32_t)(this->filter->element), size);
+            }
+        };
+
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_depthwise_conv2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_depthwise_conv2d.hpp
new file mode 100644
index 00000000000..30b2c2a6c77
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_depthwise_conv2d.hpp
@@ -0,0 +1,188 @@
+#pragma once
+
+#include "dl_nn_depthwise_conv2d.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Activation(DepthwiseConv2D(filter, input) + bias).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @tparam bias_t supports int16_t and int8_t, must specify when using int8 per-channel quantization
+         *         - int16_t: for int16 quantization and int8 per-channel quantization
+         *         - int8_t: for int8 per-tensor quantization
+         */
+        template <typename feature_t, typename bias_t = feature_t>
+        class DepthwiseConv2D : public Layer
+        {
+        private:
+            const int output_exponent;               /*<! exponent of output >*/
+            const Filter<feature_t> *filter;         /*<! filter of DepthwiseConv2D >*/
+            const int stride_y;                      /*<! stride in height >*/
+            const int stride_x;                      /*<! stride in width >*/
+            const padding_type_t padding_type;       /*<! one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN >*/
+            const Bias<bias_t> *bias;                /*<! bias of DepthwiseConv2D, if you don't specify anything, no bias is added >*/
+            const Activation<feature_t> *activation; /*<! activation of DepthwiseConv2D, if you don't specify anything, no activation is applied >*/
+            std::vector<int> padding;                /*<! padding size needed in [top, bottom, left, right] of this operation >*/
+            Tensor<feature_t> *output;               /*<! output ptr of DepthwiseConv2D >*/
+            std::vector<int> output_shape;           /*<! output shape of DepthwiseConv2D >*/
+
+        public:
+            /**
+             * @brief Construct a new DepthwiseConv2D object.
+             * 
+             * @param output_exponent exponent of output
+             * @param filter          filter of DepthwiseConv2D
+             * @param bias            bias of DepthwiseConv2D, if you don't specify anything, no bias is added
+             * @param activation      activation of DepthwiseConv2D, if you don't specify anything, no activation is applied
+             * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN or PADDING_NOT_SET,
+             *                        - PADDING_VALID means no padding
+             *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
+             *                        such that output has the same height/width dimension as the input,
+             *                        - PADDING_SAME_END results padding in TensorFlow style
+             *                        - PADDING_SAME_BEGIN results padding in MXNET style
+             *                        - PADDING_NOT_SET means padding with the specific "padding" value below. 
+             * @param padding         if padding_type is PADDING_NOT_SET, this value will be used as padding size. 
+             *                        the shape must be 4, the value of each position is: [padding top, padding bottom, padding left, padding right]
+             * @param stride_y        - stride in height
+             * @param stride_x        - stride in width
+             * @param name            name of layer
+             */
+            DepthwiseConv2D(const int output_exponent,
+                            const Filter<feature_t> *filter,
+                            const Bias<bias_t> *bias = NULL,
+                            const Activation<feature_t> *activation = NULL,
+                            const padding_type_t padding_type = PADDING_VALID,
+                            std::vector<int> padding = {},
+                            const int stride_y = 1,
+                            const int stride_x = 1,
+                            const char *name = "DepthwiseConv2D") : Layer(name),
+                                                                    output_exponent(output_exponent),
+                                                                    filter(filter),
+                                                                    stride_y(stride_y),
+                                                                    stride_x(stride_x),
+                                                                    padding_type(padding_type),
+                                                                    bias(bias),
+                                                                    activation(activation),
+                                                                    padding(padding),
+                                                                    output_shape({})
+            {
+                this->output = new Tensor<feature_t>;
+                if (this->padding_type == PADDING_NOT_SET)
+                {
+                    assert(this->padding.size() == 4);
+                }
+            }
+
+            /**
+             * @brief Destroy the DepthwiseConv2D object.
+             * 
+             */
+            ~DepthwiseConv2D()
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape and padding.
+             * 
+             * @param input as an input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                assert(input.shape[0] > 0);
+                assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
+                assert(this->filter->shape.size() == 4);
+                assert(input.shape[2] == this->filter->shape[2]);
+
+                this->output_shape = nn::get_output_shape(input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type, false, this->padding);
+                this->output->set_shape(this->output_shape);
+                this->output->set_exponent(this->output_exponent);
+
+                if (this->padding_type != PADDING_NOT_SET)
+                {
+                    this->padding = nn::get_pad_size(this->output_shape, input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type);
+                }
+                this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& DepthwiseConv2D result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call DepthwiseConv2D operation.
+             * 
+             * @param input           as an input
+             * @param autoload_enable one of true or false,
+             *                        - true: load input and output from PSRAM to CACHE automatically
+             *                        - false: do not 
+             * @param assign_core     not effective yet
+             * @return DepthwiseConv2D result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input, bool autoload_enable = false, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+
+                this->output->malloc_element();
+                this->output->set_exponent(this->output_exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                if (autoload_enable)
+                {
+                    dl::tool::cache::autoload_func((uint32_t)(this->output->element), this->output->get_size() * sizeof(feature_t),
+                                                   (uint32_t)(input.element), input.get_size() * sizeof(feature_t));
+                }
+
+                DL_LOG_LAYER_LATENCY_START();
+                nn::depthwise_conv2d(*this->output, input, this->padding, *(this->filter), this->stride_y, this->stride_x, this->bias, this->activation, assign_core);
+                DL_LOG_LAYER_LATENCY_END(this->name, "depthwise_conv2d");
+
+                return *this->output;
+            }
+
+            /**
+             * @brief Preload the filter to Cache. 
+             * NOTE: Call this layer's preload() before previous layer's call() such that filter could be loaded while previous layer is calculating.
+             */
+            void preload()
+            {
+                size_t size = sizeof(feature_t);
+                int shape_size = this->filter->shape.size();
+                for (int i = 0; i < shape_size; ++i)
+                {
+                    size *= filter->shape[i];
+                }
+                dl::tool::cache::preload_func((uint32_t)(this->filter->element), size);
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_expand_dims.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_expand_dims.hpp
new file mode 100644
index 00000000000..a59bed183fb
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_expand_dims.hpp
@@ -0,0 +1,128 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief 
+         * 
+         * @tparam feature_t 
+         */
+        template <typename feature_t>
+        class ExpandDims : public Layer
+        {
+        private:
+            std::vector<int> output_shape; /*<! output shape of ExpandDims >*/
+            std::vector<int> axis;         /*<! position where the new axis is placed. >*/
+            Tensor<feature_t> *output;     /*<! output ptr of ExpandDims >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+
+        public:
+            int output_exponent;
+
+            /**
+             * @brief Construct a new ExpandDims object
+             * 
+             * @param axis position where the new axis is placed.
+             * @param name name of layer
+             * @param inplace true: the output will store to input
+             *                false: the output will store to a separate memory
+             */
+            ExpandDims(std::vector<int> axis, const char *name = "ExpandDims", bool inplace = false) : Layer(name),
+                                                                                                       axis(axis), inplace(inplace), output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the ExpandDims object
+             * 
+             */
+            ~ExpandDims()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief  Update output shape.
+            * 
+            * @param input as an input.
+            * @param print_shape  whether to print the output shape.
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->expand_dims(this->axis);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                    this->output->expand_dims(this->axis);
+                }
+                this->output_shape = this->output->shape;
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& ExpandDims result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief call ExpandDims opeartion
+             * 
+             * @param input 
+             * @return Tensor<feature_t>& ExpandDims result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->copy_element(input, true);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "ExpandDims");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_shape(this->output_shape);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "ExpandDims");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_flatten.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_flatten.hpp
new file mode 100644
index 00000000000..70ae483a07f
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_flatten.hpp
@@ -0,0 +1,120 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief 
+         * 
+         * @tparam feature_t 
+         */
+        template <typename feature_t>
+        class Flatten : public Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Flatten >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of Flatten >*/
+
+        public:
+            /**
+             * @brief Construct a new Flatten object
+             * 
+             * @param name name of layer
+             * @param inplace true: the output will store to input0
+             *                false: the output will store to a separate memory
+             */
+            Flatten(const char *name = "Flatten", bool inplace = false) : Layer(name), inplace(inplace), output_shape({})
+            {}
+
+            /**
+             * @brief Destroy the Flatten object
+             * 
+             */
+            ~Flatten()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief  Update output shape.
+            * 
+            * @param input as an input
+            * @param print_shape  whether to print the output shape. 
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                this->output_shape = {input.get_size()};
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                }
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Flatten result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Flatten operation.
+             * 
+             * @param input as an input
+             * @return Tensor<feature_t>& Flatten result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->flatten();
+                    this->output->copy_element(input, true);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "flatten");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->flatten();
+                    DL_LOG_LAYER_LATENCY_END(this->name, "flatten");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_fullyconnected.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_fullyconnected.hpp
new file mode 100644
index 00000000000..afa7e5befba
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_fullyconnected.hpp
@@ -0,0 +1,167 @@
+#pragma once
+
+#include "dl_nn_fully_connected.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Activation(FullyConnected(input, filter) + bias).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @tparam bias_t supports int16_t and int8_t, must specify when using int8 per-channel quantization
+         *         - int16_t: for int16 quantization and int8 per-channel quantization
+         *         - int8_t: for int8 per-tensor quantization
+         */
+        template <typename feature_t, typename bias_t = feature_t>
+        class FullyConnected : public Layer
+        {
+        private:
+            const int output_exponent;               /*<! exponent of output >*/
+            const bool flatten;                      /*<! true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+                                                          false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim] >*/
+            const Filter<feature_t> *filter;         /*<! filter of FullyConnected >*/
+            const Bias<bias_t> *bias;                /*<! bias of FullyConnected, if you don't specify anything, no bias is added >*/
+            const Activation<feature_t> *activation; /*<! activation of FullyConnected, if you don't specify anything, no activation is applied >*/
+            Tensor<feature_t> *output;               /*<! output ptr of FullyConnected >*/
+            std::vector<int> output_shape;           /*<! output shape of FullyConnected >*/
+
+        public:
+            /**
+             * @brief Construct a new FullyConnected object.
+             * 
+             * @param output_exponent exponent of output
+             * @param filter          filter of FullyConnected
+             * @param bias            bias of FullyConnected, if you don't specify anything, no bias is added
+             * @param activation      activation of FullyConnected, if you don't specify anything, no activation is applied
+             * @param flatten         true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+                                      false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+             * @param name            name of layer
+             */
+            FullyConnected(const int output_exponent,
+                           const Filter<feature_t> *filter,
+                           const Bias<bias_t> *bias = NULL,
+                           const Activation<feature_t> *activation = NULL,
+                           const bool flatten = true,
+                           const char *name = "FullyConnected") : Layer(name),
+                                                                  output_exponent(output_exponent),
+                                                                  flatten(flatten),
+                                                                  filter(filter),
+                                                                  bias(bias),
+                                                                  activation(activation),
+                                                                  output_shape({})
+            {
+                this->output = new Tensor<feature_t>;
+            }
+
+            /**
+             * @brief Destroy the FullyConnected object.
+             * 
+             */
+            ~FullyConnected()
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output padding and input padding.
+             * 
+             * @param input as an input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                assert(this->filter->shape.size() == 4);
+                assert(this->filter->shape[0] == 1);
+                assert(this->filter->shape[1] == 1);
+                if (this->flatten)
+                {
+                    assert(input.get_size() == this->filter->shape[2]);
+                    this->output_shape = {this->filter->shape[3]};
+                }
+                else
+                {
+                    assert(input.shape.back() == this->filter->shape[2]);
+                    this->output_shape = input.shape;
+                    this->output_shape[this->output_shape.size() - 1] = this->filter->shape[3];
+                }
+                this->output->set_shape(this->output_shape);
+                this->output->set_exponent(this->output_exponent);
+                this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& FullyConnected result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call FullyConnected operation
+             * 
+             * @param input           as an input.
+             * @param autoload_enable one of true or false, 
+             *                        - true: load input and output from PSRAM to CACHE automatically
+             *                        - false: do not
+             * @param assign_core     not effective yet
+             * @return FullyConnected result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input, bool autoload_enable = false, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
+                this->output->set_exponent(this->output_exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                if (autoload_enable)
+                {
+                    dl::tool::cache::autoload_func((uint32_t)(this->output->element), this->output->get_size() * sizeof(feature_t),
+                                                   (uint32_t)(input.element), input.get_size() * sizeof(feature_t));
+                }
+
+                DL_LOG_LAYER_LATENCY_START();
+                nn::fully_connected(*this->output, input, *(this->filter), this->bias, this->activation, this->flatten, assign_core);
+                DL_LOG_LAYER_LATENCY_END(this->name, "fully_connected");
+                return *this->output;
+            }
+
+            /**
+             * @brief Preload the filter to Cache.
+             * NOTE: Call this layer's preload() before previous layer's call() such that filter could be loaded while previous layer is doing calculation.
+             */
+            void preload()
+            {
+                size_t size = sizeof(feature_t);
+                int shape_size = this->filter->shape.size();
+                for (int i = 0; i < shape_size; ++i)
+                {
+                    size *= filter->shape[i];
+                }
+                dl::tool::cache::preload_func((uint32_t)(this->filter->element), size);
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_global_avg_pool2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_global_avg_pool2d.hpp
new file mode 100644
index 00000000000..93f2d30ac89
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_global_avg_pool2d.hpp
@@ -0,0 +1,126 @@
+#pragma once
+
+#include <vector>
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn_global_avg_pool2d.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief GlobalAveragePool2D(input).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class GlobalAveragePool2D : public Layer
+        {
+        private:
+            const int output_exponent;     /*<! exponent of output >*/
+            std::vector<int> output_shape; /*<! output shape of GlobalAveragePool2D >*/
+            Tensor<feature_t> *output;     /*<! output ptr of GlobalAveragePool2D >*/
+        public:
+            /**
+             * @brief Construct a new GlobalAveragePool2D object.
+             * 
+             * @param output_exponent exponent of output
+             * @param name            name of layer
+             */
+            GlobalAveragePool2D(const int output_exponent, const char *name = "GlobalAveragePool2D") : Layer(name),
+                                                                                                       output_exponent(output_exponent),
+                                                                                                       output_shape({})
+
+            {
+                this->output = new Tensor<feature_t>;
+            }
+
+            /**
+            * @brief Destroy the GlobalAveragePool2D object.
+            * 
+            */
+            ~GlobalAveragePool2D()
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape.
+             * 
+             * @param input as an input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                assert(input.shape[0] > 0);
+                assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
+
+                std::vector<int> output_shape(input.shape.size(), 1);
+                output_shape[2] = input.shape[2];
+                this->output_shape = output_shape;
+                this->output->set_shape(this->output_shape);
+                this->output->set_exponent(this->output_exponent);
+                this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& GlobalAveragePool2D result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call GlobalAveragePool2D operation
+             * 
+             * @param input           as an input
+             * @param autoload_enable one of true or false, 
+             *                        - true: load input and output from PSRAM to CACHE automatically
+             *                        - false: do not
+             * @param assign_core     not effective yet
+             * @return GlobalAveragePool2D result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input, uint8_t autoload_enable = 0)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
+                this->output->set_exponent(this->output_exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                if (autoload_enable)
+                {
+                    dl::tool::cache::autoload_func((uint32_t)(this->output->element), this->output->get_size() * sizeof(feature_t),
+                                                   (uint32_t)(input.element), input.get_size() * sizeof(feature_t));
+                }
+
+                DL_LOG_LAYER_LATENCY_START();
+                nn::global_avg_pool2d(*this->output, input);
+                DL_LOG_LAYER_LATENCY_END(this->name, "global_avg_pool2d");
+
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_global_max_pool2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_global_max_pool2d.hpp
new file mode 100644
index 00000000000..f9b7f73ff8c
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_global_max_pool2d.hpp
@@ -0,0 +1,121 @@
+#pragma once
+
+#include <vector>
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn_global_max_pool2d.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief GlobalMaxPool2D(input).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class GlobalMaxPool2D : public Layer
+        {
+        private:
+            Tensor<feature_t> *output;     /*<! output ptr of GlobalMaxPool2D >*/
+            std::vector<int> output_shape; /*<! output shape of GlobalMaxPool2D >*/
+        public:
+            /**
+             * @brief Construct a new GlobalMaxPool2D object.
+             * 
+             * @param name         name of layer
+             */
+            GlobalMaxPool2D(const char *name = "GlobalMaxPool2D") : Layer(name), output_shape({})
+            {
+                this->output = new Tensor<feature_t>;
+            }
+
+            /**
+             * @brief Destroy the GlobalMaxPool2D object.
+             * 
+             */
+            ~GlobalMaxPool2D()
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape and exponent.
+             * 
+             * @param input as an input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                assert(input.shape[0] > 0);
+                assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
+                this->output->set_exponent(input.exponent);
+
+                std::vector<int> output_shape(input.shape.size(), 1);
+                output_shape[2] = input.shape[2];
+                this->output_shape = output_shape;
+                this->output->set_shape(this->output_shape);
+                this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output 
+             * 
+             * @return Tensor<feature_t>& GlobalMaxPool2D result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call GlobalMaxPool2D operation
+             * 
+             * @param input           as an input
+             * @param autoload_enable one of true or false, 
+             *                        - true: load input and output from PSRAM to CACHE automatically
+             *                        - false: do not
+             * @param assign_core     not effective yet
+             * @return GlobalMaxPool2D result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input, uint8_t autoload_enable = 0)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
+                this->output->set_exponent(input.exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                if (autoload_enable)
+                {
+                    dl::tool::cache::autoload_func((uint32_t)(this->output->element), this->output->get_size() * sizeof(feature_t),
+                                                   (uint32_t)(input.element), input.get_size() * sizeof(feature_t));
+                }
+
+                DL_LOG_LAYER_LATENCY_START();
+                nn::global_max_pool2d(*this->output, input);
+                DL_LOG_LAYER_LATENCY_END(this->name, "global_max_pool2d");
+
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_leakyrelu.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_leakyrelu.hpp
new file mode 100644
index 00000000000..a972e135006
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_leakyrelu.hpp
@@ -0,0 +1,141 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn_leakyrelu.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief LeakyReLU(input).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class LeakyReLU : public Layer
+        {
+        private:
+            feature_t activation_alpha;    /*<! quantized alpha >*/
+            int activation_exponent;       /*<! exponent of quantized alpha >*/
+            Tensor<feature_t> *output;     /*<! output ptr of leakyrelu>*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of leakyrelu >*/
+        public:
+            /**
+             * @brief Construct a new LeakyReLU object
+             * 
+             * @param activation_alpha     quantized alpha
+             * @param activation_exponent  exponent of quantized alpha
+             * @param name                 name of leakyrelu
+             * @param inplace              true: the output will store to input0
+             *                             false: the output will store to a separate memory
+             */
+            LeakyReLU(const int activation_alpha, const int activation_exponent, const char *name = "LeakyReLU", bool inplace = false) : Layer(name), output(NULL), output_shape({})
+            {
+                this->activation_alpha = activation_alpha;
+                this->activation_exponent = activation_exponent;
+                this->inplace = inplace;
+            }
+
+            /**
+             * @brief Destroy the LeakyReLU object
+             * 
+             */
+            ~LeakyReLU()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape and exponent
+             * 
+             * @param input       as an input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_shape = input.shape;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_shape(this->output_shape);
+                    this->output->set_exponent(input.exponent);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output 
+             * 
+             * @return Tensor<feature_t>& LeakyReLU result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call LeakyReLU operation.
+             * 
+             * @param input       as an input
+             * @param assign_core not effective yet
+             * @return LeakyReLU result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
+                    this->output->set_exponent(input.exponent);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                    DL_LOG_LAYER_LATENCY_START();
+                    nn::leakyrelu(*this->output, input, this->activation_alpha, this->activation_exponent, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "leakyrelu");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    nn::leakyrelu<true>(*this->output, input, this->activation_alpha, this->activation_exponent, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "leakyrelu");
+                }
+
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_max2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_max2d.hpp
new file mode 100644
index 00000000000..93bc5899443
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_max2d.hpp
@@ -0,0 +1,143 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_nn_max2d.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Max2D(input0, input1).
+         * NOTE: maximum is element-wise, i.e., output[i,j,k] = max(input0[i,j,k], input1[i,j,k])
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class Max2D : public Layer
+        {
+        private:
+            Tensor<feature_t> *output;     /*<! output ptr of max2d >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of max2d >*/
+        public:
+            /**
+             * @brief Construct a new Max2D object.
+             * 
+             * @param name            name of max2d
+             * @param inplace         true: the output will store to input0
+             *                        false: the output will store to a separate memory
+             */
+            Max2D(const char *name = "Max2D", bool inplace = false) : Layer(name),
+                                                                      output(NULL), inplace(inplace), output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the Max2D object
+             * 
+             */
+            ~Max2D()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape and exponent
+             * NOTE: input0.shape must equal to input1.shape.
+             *       input0.exponent must equal to input1.exponent.
+             * 
+             * @param input0 as one input
+             * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
+            {
+                assert(input0.is_same_shape(input1));
+                assert(input0.exponent == input1.exponent);
+                this->output_shape = input0.shape;
+
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input0;
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Max2D result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Max2D operation.
+             * 
+             * @param input0      as one input
+             * @param input1      as another input
+             * @param assign_core not effective yet
+             * @return Max2D result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input0, Tensor<feature_t> &input1, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
+                    this->output->set_exponent(input0.exponent);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                    DL_LOG_LAYER_LATENCY_START();
+                    nn::max2d(*this->output, input0, input1, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "max2d");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    nn::max2d<true>(*this->output, input0, input1, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "max2d");
+                }
+
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_max_pool2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_max_pool2d.hpp
new file mode 100644
index 00000000000..629aa87f515
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_max_pool2d.hpp
@@ -0,0 +1,157 @@
+#pragma once
+
+#include <vector>
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn_max_pool2d.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief MaxPool2D(input).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class MaxPool2D : public Layer
+        {
+        private:
+            std::vector<int> filter_shape;     /*<! filter shape in [filter_height, filter_width] >*/
+            const int stride_y;                /*<! stride in height >*/
+            const int stride_x;                /*<! stride in width >*/
+            const padding_type_t padding_type; /*<! one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN >*/
+            std::vector<int> padding;          /*<! padding size needed in [top, bottom, left, right] of this operation >*/
+            Tensor<feature_t> *output;         /*<! output ptr of MaxPool2D >*/
+            std::vector<int> output_shape;     /*<! output shape of MaxPool2D >*/
+
+        public:
+            /**
+             * @brief Construct a new MaxPool2D object.
+             * 
+             * @param filter_shape filter shape in [filter_height, filter_width]
+             * @param padding_type one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN or PADDING_NOT_SET,
+             *                     - PADDING_VALID means no padding
+             *                     PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
+             *                     such that output has the same height/width dimension as the input,
+             *                     - PADDING_SAME_END results padding in TensorFlow style
+             *                     - PADDING_SAME_BEGIN results padding in MXNET style
+             *                     - PADDING_NOT_SET means padding with the specific "padding" value below. 
+             * @param padding      if padding_type is PADDING_NOT_SET, this value will be used as padding size. 
+             *                     the shape must be 4, the value of each position is: [padding top, padding bottom, padding left, padding right]
+             * @param stride_y     stride in height
+             * @param stride_x     stride in width
+             * @param name         name of layer
+             */
+            MaxPool2D(const std::vector<int> filter_shape,
+                      const padding_type_t padding_type = PADDING_VALID,
+                      std::vector<int> padding = {},
+                      const int stride_y = 1,
+                      const int stride_x = 1,
+                      const char *name = "MaxPool2D") : Layer(name),
+                                                        filter_shape(filter_shape),
+                                                        padding_type(padding_type),
+                                                        padding(padding),
+                                                        stride_y(stride_y),
+                                                        stride_x(stride_x),
+                                                        output_shape({})
+            {
+                this->output = new Tensor<feature_t>;
+                if (this->padding_type == PADDING_NOT_SET)
+                {
+                    assert(this->padding.size() == 4);
+                }
+            }
+
+            /**
+             * @brief Destroy the MaxPool2D object.
+             * 
+             */
+            ~MaxPool2D()
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape and padding.
+             * 
+             * @param input as an input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                assert(input.shape[0] > 0);
+                assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
+
+                this->output->set_exponent(input.exponent);
+                this->output_shape = nn::get_output_shape(input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type, false, this->padding);
+                this->output->set_shape(this->output_shape);
+
+                if (this->padding_type != PADDING_NOT_SET)
+                {
+                    this->padding = nn::get_pad_size(this->output_shape, input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
+                }
+                this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& MaxPool2D result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call MaxPool2D operation
+             * 
+             * @param input           as an input
+             * @param autoload_enable one of true or false, 
+             *                        - true: load input and output from PSRAM to CACHE automatically
+             *                        - false: do not
+             * @param assign_core     not effective yet
+             * @return MaxPool2D result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input, uint8_t autoload_enable = 0)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
+                this->output->set_exponent(input.exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                if (autoload_enable)
+                {
+                    dl::tool::cache::autoload_func((uint32_t)(this->output->element), this->output->get_size() * sizeof(feature_t),
+                                                   (uint32_t)(input.element), input.get_size() * sizeof(feature_t));
+                }
+
+                DL_LOG_LAYER_LATENCY_START();
+                nn::max_pool2d(*this->output, input, this->padding, this->filter_shape, this->stride_y, this->stride_x);
+                DL_LOG_LAYER_LATENCY_END(this->name, "max_pool2d");
+
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_min2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_min2d.hpp
new file mode 100644
index 00000000000..e38fbf3d0d2
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_min2d.hpp
@@ -0,0 +1,143 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_nn_min2d.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Min2D(input0, input1).
+         * NOTE: minimum is element-wise, i.e., output[i,j,k] = min(input0[i,j,k], input1[i,j,k])
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class Min2D : public Layer
+        {
+        private:
+            Tensor<feature_t> *output;     /*<! output of ptr min2d>*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of min2d >*/
+        public:
+            /**
+             * @brief Construct a new Min2D object
+             * 
+             * @param name            name of min2d
+             * @param inplace         true: the output will store to input0
+             *                        false: the output will store to a separate memory
+             */
+            Min2D(const char *name = "Min2D", bool inplace = false) : Layer(name),
+                                                                      output(NULL),
+                                                                      inplace(inplace),
+                                                                      output_shape({}) {}
+
+            /**
+             * @brief Destroy the Min2D object
+             * 
+             */
+            ~Min2D()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape and exponent
+             * NOTE: input0.shape must equal to input1.shape.
+             *       input0.exponent must equal to input1.exponent.
+             * 
+             * @param input0 as one input
+             * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
+            {
+                assert(input0.is_same_shape(input1));
+                assert(input0.exponent == input1.exponent);
+                this->output_shape = input0.shape;
+
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_shape(this->output_shape);
+                    this->output->set_exponent(input0.exponent);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input0;
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Min2D result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Min2D operation
+             * 
+             * @param input0      as one input
+             * @param input1      as another input
+             * @param assign_core not effective yet
+             * @return Min2D result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input0, Tensor<feature_t> &input1, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
+                    this->output->set_exponent(input0.exponent);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                    DL_LOG_LAYER_LATENCY_START();
+                    nn::min2d(*this->output, input0, input1, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "min2d");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    nn::min2d<true>(*this->output, input0, input1, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "min2d");
+                }
+
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_model.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_model.hpp
new file mode 100644
index 00000000000..2064ef2d33a
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_model.hpp
@@ -0,0 +1,52 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Neural Network Model.
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class Model
+        {
+        private:
+            std::vector<int> input_shape; /*<! input shape in [height, width, channel] >*/
+
+        public:
+            /**
+             * @brief Destroy the Model object.
+             * 
+             */
+            virtual ~Model() {}
+
+            /**
+             * @brief Build a model including update output shape and input padding of each layer.
+             * 
+             * @param input as an input
+             */
+            virtual void build(Tensor<feature_t> &input) = 0;
+
+            /**
+             * @brief Call the model layer by layer.
+             * 
+             * @param input as an input.
+             */
+            virtual void call(Tensor<feature_t> &input) = 0;
+
+            /**
+             * @brief If input.shape changes, call Model.build(), otherwise, do not. Then call Model.call().
+             * 
+             * @param input as an input
+             */
+            void forward(Tensor<feature_t> &input);
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_mul2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_mul2d.hpp
new file mode 100644
index 00000000000..21bcca7a81e
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_mul2d.hpp
@@ -0,0 +1,151 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn_mul2d.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Activation(Multiply2D(input0, input1)).
+         * NOTE: multiplication is element-wise, i.e., output[i,j,k] = input0[i,j,k] * input1[i,j,k]
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class Mul2D : public Layer
+        {
+        private:
+            const int output_exponent;               /*<! exponent of output >*/
+            const Activation<feature_t> *activation; /*<! activation of Mul2D, if you don't specify anything, no activation is applied >*/
+            Tensor<feature_t> *output;               /*<! output ptr of Mul2D >*/
+            bool inplace;                            /*<! true: the output will store to input0
+                                                            false: the output will store to a separate memory >*/
+            std::vector<int> output_shape;           /*<! output shape of Mul2D >*/
+        public:
+            /**
+             * @brief Construct a new Mul2D object.
+             * 
+             * @param output_exponent exponent of output
+             * @param activation      activation of Mul2D, if you don't specify anything, no activation is applied
+             * @param name            name of layer
+             * @param inplace         true: the output will store to input0
+             *                        false: the output will store to a separate memory
+             */
+            Mul2D(const int output_exponent,
+                  const Activation<feature_t> *activation = NULL,
+                  const char *name = "Mul2D",
+                  bool inplace = false) : Layer(name),
+                                          output_exponent(output_exponent),
+                                          activation(activation),
+                                          output(NULL),
+                                          inplace(inplace),
+                                          output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the Multiply2D object.
+             */
+            ~Mul2D()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape.
+             * NOTE: input0.shape must equal to input1.shape.
+             * 
+             * @param input0 as one input
+             * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
+            {
+                assert(input0.is_same_shape(input1));
+                this->output_shape = input0.shape;
+
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+
+                else
+                {
+                    this->output = &input0;
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Mul2D result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Mul2D operation.
+             * 
+             * @param input0      as one input
+             * @param input1      as another input
+             * @param assign_core not effective yet
+             * @return Mul2D result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input0, Tensor<feature_t> &input1, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
+                    this->output->set_exponent(this->output_exponent);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                    DL_LOG_LAYER_LATENCY_START();
+                    nn::mul2d(*this->output, input0, input1, this->activation, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "mul2d");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    nn::mul2d<true>(*this->output, input0, input1, this->activation, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "mul2d");
+                }
+
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_prelu.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_prelu.hpp
new file mode 100644
index 00000000000..96168a783b1
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_prelu.hpp
@@ -0,0 +1,145 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn_prelu.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief PReLU(input).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class PReLU : public Layer
+        {
+        private:
+            feature_t *activation_element; /*<! quantized alpha elements along channel axis >*/
+            int activation_exponent;       /*<! exponent of quantized alpha elements >*/
+            Tensor<feature_t> *output;     /*<! output ptr of prelu >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of prelu >*/
+        public:
+            /**
+             * @brief Construct a new PReLU object
+             * 
+             * @param activation_element   quantized alpha elements along channel axis
+             * @param activation_exponent  exponent of quantized alpha elements
+             * @param name                 name of prelu
+             * @param inplace              true: the output will store to input0
+             *                             false: the output will store to a separate memory
+             */
+            PReLU(const feature_t *activation_element,
+                  const int activation_exponent = 0,
+                  const char *name = NULL,
+                  bool inplace = "PReLU") : Layer(name),
+                                            activation_element(activation_element),
+                                            activation_exponent(activation_exponent),
+                                            output(NULL),
+                                            inplace(inplace),
+                                            output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the PReLU object
+             * 
+             */
+            ~PReLU()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape and exponent
+             * 
+             * @param input       as an input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_shape = input.shape;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(input.exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& PReLU result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call PReLU operation.
+             * 
+             * @param input       as an input
+             * @param assign_core not effective yet
+             * @return PReLU result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->set_exponent(input.exponent);
+                    this->output->malloc_element();
+                    DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                    DL_LOG_LAYER_LATENCY_START();
+                    nn::prelu(*this->output, input, this->activation_element, this->activation_exponent, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "leakyrelu");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    nn::prelu(*this->output, input, this->activation_element, this->activation_exponent, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "leakyrelu");
+                }
+
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_relu.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_relu.hpp
new file mode 100644
index 00000000000..7dd29d4a178
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_relu.hpp
@@ -0,0 +1,135 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_nn_relu.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief ReLU(input).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class ReLU : public Layer
+        {
+        private:
+            Tensor<feature_t> *output;     /*<! output ptr of relu >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of relu >*/
+        public:
+            /**
+             * @brief Construct a new ReLU object
+             * 
+             * @param name            name of relu
+             * @param inplace         true: the output will store to input0
+             *                        false: the output will store to a separate memory
+             */
+            ReLU(const char *name = "ReLU", bool inplace = false) : Layer(name),
+                                                                    output(NULL), inplace(inplace), output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the ReLU object
+             * 
+             */
+            ~ReLU()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape and exponent
+             * 
+             * @param input       as an input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_shape = input.shape;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(input.exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& ReLU result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call ReLU operation.
+             * 
+             * @param input       as an input
+             * @param assign_core not effective yet
+             * @return ReLU result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
+                    this->output->set_exponent(input.exponent);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                    DL_LOG_LAYER_LATENCY_START();
+                    nn::relu(*this->output, input, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "relu");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    nn::relu(*this->output, input, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "relu");
+                }
+
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_reshape.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_reshape.hpp
new file mode 100644
index 00000000000..3f2ed72b6e0
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_reshape.hpp
@@ -0,0 +1,124 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief  Reshape(input)
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class Reshape : public Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Reshape >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of Reshape >*/
+        public:
+            /**
+             * @brief Construct a new Reshape object
+             * 
+             * @param shape      the target shape
+             * @param name       name of Reshape layer
+             * @param inplace    true: the output will store to input0
+             *                   false: the output will store to a separate memory
+             */
+            Reshape(std::vector<int> shape, const char *name = "Reshape", bool inplace = false) : Layer(name),
+                                                                                                  output_shape(shape), inplace(inplace)
+            {
+            }
+
+            /**
+             * @brief Destroy the Reshape object
+             * 
+             */
+            ~Reshape()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief             Update output shape and exponent
+            * 
+            * @param input        as an input
+            * @param print_shape  whether to print the output shape.
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Reshape result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Reshape operation.
+             * 
+             * @param input  as an input
+             * @return Tensor<feature_t>& Reshape result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->reshape(this->output_shape);
+                    this->output->copy_element(input, true);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "reshape");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->reshape(this->output_shape);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "reshape");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_squeeze.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_squeeze.hpp
new file mode 100644
index 00000000000..cee92f22764
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_squeeze.hpp
@@ -0,0 +1,127 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief 
+         * 
+         * @tparam feature_t 
+         */
+        template <typename feature_t>
+        class Squeeze : public Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Squeeze >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                 false: the output will store to a separate memory >*/
+            int axis;                      /*<! the dim to to be remove. make sure the length of the dim is equal to 1.
+                                            if axis == INT32_MAX, all the dims with length==1 will be removed. >*/
+            std::vector<int> output_shape; /*<! output shape of AvgPool2D >*/
+        public:
+            /**
+             * @brief Construct a new Squeeze object
+             * 
+             * @param axis the dim to to be remove. make sure the length of the dim is equal to 1.
+             *             if axis == INT32_MAX, all the dims with length==1 will be removed.
+             * @param name      name of Squeeze layer
+             * @param inplace   true: the output will store to input0
+             *                  false: the output will store to a separate memory
+             */
+            Squeeze(int axis = INT32_MAX, const char *name = "Squeeze", bool inplace = false) : Layer(name), axis(axis), inplace(inplace), output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the Squeeze object
+             * 
+             */
+            ~Squeeze()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief Update output shape and exponent
+            * 
+            * @param input as an input
+            * @param print_shape  whether to print the output shape.
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(input.shape);
+                    this->output->squeeze(this->axis);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(input.shape);
+                    this->output->squeeze(this->axis);
+                }
+                this->output_shape = this->output->shape;
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Squeeze result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief  Call Squeeze operation.
+             * 
+             * @param input as an input
+             * @return Tensor<feature_t>& Squeeze result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->copy_element(input, true);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "Squeeze");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_shape(this->output_shape);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "Squeeze");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_sub2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_sub2d.hpp
new file mode 100644
index 00000000000..da03b4aad85
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_sub2d.hpp
@@ -0,0 +1,141 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn_sub2d.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Activation(Sub2D(input0, input1)).
+         * NOTE: subtraction is element-wise, i.e., output[i,j,k] = input0[i,j,k] - input1[i,j,k]
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class Sub2D : public Layer
+        {
+        private:
+            const int output_exponent;               /*<! exponent of output >*/
+            const Activation<feature_t> *activation; /*<! activation of Sub2D, if you don't specify anything, no activation is applied >*/
+            Tensor<feature_t> *output;               /*<! output ptr of Sub2D >*/
+            bool inplace;                            /*<! true: the output will store to input0
+                                                            false: the output will store to a separate memory >*/
+            std::vector<int> output_shape;           /*<! output shape of Sub2D >*/
+        public:
+            /**
+             * @brief Construct a new Sub2D object.
+             * 
+             * @param output_exponent exponent of output
+             * @param activation      activation of Mul2D, if you don't specify anything, no activation is applied
+             * @param name            name of layer
+             * @param inplace         true: the output will store to input0
+             *                        false: the output will store to a separate memory
+             */
+            Sub2D(const int output_exponent, const Activation<feature_t> *activation = NULL, const char *name = "Sub2D", bool inplace = false) : Layer(name),
+                                                                                                                                                 output_exponent(output_exponent), activation(activation), output(NULL), inplace(inplace), output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the Sub2D object.
+             */
+            ~Sub2D()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output shape.
+             * NOTE: input0.shape must equal to input1.shape.
+             * 
+             * @param input0 as one input
+             * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
+            {
+                assert(input0.is_same_shape(input1));
+                this->output_shape = input0.shape;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input0;
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Sub2D result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Sub2D operation.
+             * 
+             * @param input0      as one input
+             * @param input1      as another input
+             * @param assign_core not effective yet
+             * @return Sub2D result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input0, Tensor<feature_t> &input1, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output.malloc_element();
+                    this->output->set_exponent(input0.exponent);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                    DL_LOG_LAYER_LATENCY_START();
+                    nn::sub2d(this->output, input0, input1, this->activation, assign_core);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "sub2d");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    nn::sub2d<true>(this->output, input0, input1, this->activation, assign_core, this->output_exponent);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "sub2d");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_transpose.hpp b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_transpose.hpp
new file mode 100644
index 00000000000..d89ba8daed5
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/layer/dl_layer_transpose.hpp
@@ -0,0 +1,126 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief 
+         * 
+         * @tparam feature_t 
+         */
+        template <typename feature_t>
+        class Transpose : public Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Transpose >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> perm;         /*<! the new arangement of the dims. if perm == {}, the dims arangement will be reversed. >*/
+            std::vector<int> output_shape; /*<! output shape of Transpose >*/
+        public:
+            /**
+             * @brief Construct a new Transpose object
+             * 
+             * @param perm          the new arangement of the dims. if perm == {}, the dims arangement will be reversed.
+             * @param name          name of Transpose layer
+             * @param inplace       true: the output will store to input
+             *                      false: the output will store to a separate memory
+             */
+            Transpose(std::vector<int> perm = {}, const char *name = "Transpose", bool inplace = false) : Layer(name), perm(perm), inplace(inplace), output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the Transpose object
+             * 
+             */
+            ~Transpose()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief Update output shape and exponent
+            * 
+            * @param input as an input
+            * @param print_shape  whether to print the output shape.
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                this->output_shape = input.shape;
+                for (int i = 0; i < this->perm.size(); i++)
+                {
+                    this->output_shape[i] = input.shape[this->perm[i]];
+                }
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Transpose result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Transpose operation.
+             * 
+             * @param input as an input.
+             * @return Tensor<feature_t>& Transpose result.
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->transpose(input, this->perm);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "transpose");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->transpose(this->perm);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "transpose");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/math/dl_math.hpp b/tools/sdk/esp32c3/include/esp-face/include/math/dl_math.hpp
new file mode 100644
index 00000000000..d3f2b94d3de
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/math/dl_math.hpp
@@ -0,0 +1,188 @@
+#pragma once
+
+#include "dl_define.hpp"
+
+namespace dl
+{
+    namespace math
+    {
+        /**
+         * @brief x^a.
+         * 
+         * @param x as a base
+         * @param a as an exponent
+         * @return x^a
+         */
+        inline float power(float x, int a)
+        {
+            if (a > 0)
+            {
+                return x * power(x, a - 1);
+            }
+            else if (a < 0)
+            {
+                return 1 / (x * power(x, -a - 1));
+            }
+            else
+            {
+                return 1.f;
+            }
+        }
+
+        /**
+         * @brief sqrt(x).
+         * 
+         * @param x as a base
+         * @return sqrt(x)
+         */
+        inline float sqrt_quick(float x)
+        {
+            const int result = 0x1fbb4000 + (*(int *)&x >> 1);
+            return *(float *)&result;
+        }
+
+        /**
+         * @brief 1/sqrt(x).
+         * 
+         * @param x as a base
+         * @return 1/sqrt(x)
+         */
+        inline float sqrt_reciprocal_quick(float x)
+        {
+            float xhalf = 0.5f * x;
+            int i = *(int *)&x;             // get bits for floating value
+            i = 0x5f375a86 - (i >> 1);      // gives initial guess y0
+            x = *(float *)&i;               // convert bits back to float
+            x = x * (1.5f - xhalf * x * x); // Newton step, repeating increases accuracy
+            return x;
+        }
+
+        static const float EN = 0.00001f;
+
+        /**
+         * @brief sqrt(x).
+         * 
+         * @param x as a base
+         * @return sqrt(x)
+         */
+        inline float sqrt_newton(float x)
+        {
+            /**
+            * Use Newton iteration method to find the square root
+            * */
+            if (x == 0.f)
+                return 0.f;
+            float result = x;
+            float last_value;
+            do
+            {
+                last_value = result;
+                result = (last_value + x / last_value) * 0.5;
+            } while (DL_ABS(result - last_value) > EN);
+            return result;
+        }
+
+        /**
+         * @brief n-th root of x.
+         * 
+         * @param x as a base
+         * @param n root times
+         * @return n-th root of x
+         */
+        inline float root_newton(float x, int n)
+        {
+            if (n == 2)
+                return sqrt_newton(x);
+            if (n == 0)
+                return 1.f;
+            if (n == 1)
+                return x;
+            if (x == 0.f)
+                return 0.f;
+            float result = x;
+            float last_value;
+            float _n = (float)((n - 1) * n);
+            do
+            {
+                last_value = result;
+                result = _n * last_value + x / (n * power(last_value, n - 1));
+            } while (DL_ABS(result - last_value) > EN);
+            return result;
+        }
+
+        /**
+         * @brief atan(x).
+         * 
+         * @param x as an input
+         * @return atan(x) in range [-pi/2, pi/2]
+         */
+        inline float atan(float x)
+        {
+            return x * (0.78539816 - (DL_ABS(x) - 1) * (0.2447 + 0.0663 * DL_ABS(x)));
+            // float s = x*x;
+            // return ((-0.0464964749 * s + 0.15931422) * s - 0.327622764) * s * x + x;
+        }
+
+        // TODO:@yuanjiong
+        /**
+         * @brief 
+         * 
+         * @param x
+         * @param y 
+         * @return in range [-pi, pi]
+         */
+        inline float atan2(float x, float y)
+        {
+            float ax = DL_ABS(x);
+            float ay = DL_ABS(y);
+            float eps = 1e-8;
+            float a = DL_MIN(ax, ay) / (DL_MAX(ax, ay) + eps);
+            float r = atan(a); //[0, pi/2]
+            if (ay > ax)
+                r = 1.57079633 - r;
+            if (x < 0)
+                r = 3.14159265 - r;
+            if (y < 0)
+                r = -r;
+
+            return r;
+        }
+
+        /**
+         * @brief acos(x).
+         * 
+         * @param x as an input
+         * @return acos(x) in range [-pi/2, pi/2]
+         */
+        inline float acos(float x)
+        {
+            return atan2(x, sqrt_newton(1.0 - x * x));
+        }
+
+        /**
+         * @brief asin(x).
+         * 
+         * @param x as an input
+         * @return asin(x) in range [0, pi]
+         */
+        inline float asin(float x)
+        {
+            return atan2(sqrt_newton(1.0 - x * x), x);
+        }
+
+        /**
+         * @brief e^x
+         * 
+         * @param x     exponent
+         * @param steps iteration steps
+         * @return e^x
+         */
+        inline float exp_fast(double x, int steps)
+        {
+            x = 1.0 + x / (1 << steps);
+            for (int i = 0; i < steps; i++)
+                x *= x;
+            return x;
+        }
+    }
+}
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/math/dl_math_matrix.hpp b/tools/sdk/esp32c3/include/esp-face/include/math/dl_math_matrix.hpp
new file mode 100644
index 00000000000..30718f85c32
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/math/dl_math_matrix.hpp
@@ -0,0 +1,397 @@
+#pragma once
+
+#include <stdint.h>
+#include <stdio.h>
+#include <math.h>
+#include <vector>
+#include "dl_define.hpp"
+#include "dl_tool.hpp"
+#include "dl_variable.hpp"
+#include "esp_timer.h"
+
+namespace dl
+{
+    namespace math
+    {
+        /**
+         * @brief the Matrix class
+         * 
+         * @tparam T 
+         */
+        template <typename T>
+        class Matrix
+        {
+        public:
+            T **array;
+            int h;
+            int w;
+            Matrix() : h(0), w(0)
+            {
+                this->array = NULL;
+            }
+
+            Matrix(int h, int w) : h(h), w(w)
+            {
+                this->calloc_element();
+            }
+
+            Matrix(int h, int w, T s) : h(h), w(w)
+            {
+                this->calloc_element();
+                this->set_value(s);
+            }
+
+            Matrix(const Matrix<T> &mat) : h(mat.h), w(mat.w)
+            {
+                this->calloc_element();
+                this->set_value(mat);
+            }
+            virtual ~Matrix()
+            {
+                if (this->array != NULL)
+                {
+                    for (int i = 0; i < this->h; i++)
+                    {
+                        free(this->array[i]);
+                    }
+                    free(this->array);
+                    this->array = NULL;
+                }
+            }
+
+            /**
+             * @brief calloc the matrix element 
+             * 
+             */
+            void calloc_element()
+            {
+                if ((this->h > 0) && (this->w > 0))
+                {
+                    this->array = (T **)calloc(this->h, sizeof(T *));
+                    for (int i = 0; i < this->h; i++)
+                    {
+                        this->array[i] = (T *)calloc(this->w, sizeof(T));
+                    }
+                }
+                else
+                {
+                    this->array = NULL;
+                }
+            }
+
+            /**
+             * @brief Set the matrix element to random number.
+             * 
+             * @param thresh the max abs value of the element. 
+             */
+            void set_random(T thresh = 1)
+            {
+                unsigned int seed = esp_timer_get_time();
+                srand(seed);
+                for (int i = 0; i < this->h; i++)
+                {
+                    for (int j = 0; j < this->w; j++)
+                    {
+                        this->array[i][j] = ((T)rand()) / (T)(RAND_MAX)*thresh;
+                    }
+                }
+            }
+
+            /**
+             * @brief Set the small value to zero
+             * 
+             * @param thresh the threshold of small value 
+             */
+            void set_zero(T thresh = 1e-8)
+            {
+                for (int i = 0; i < this->h; i++)
+                {
+                    for (int j = 0; j < this->w; j++)
+                    {
+                        if (DL_ABS(this->array[i][j]) < thresh)
+                        {
+                            this->array[i][j] = 0;
+                        }
+                    }
+                }
+            }
+
+            /**
+             * @brief Set the matrix value from a vector
+             * 
+             * @tparam TT 
+             * @param mat   the input vector
+             */
+            template <typename TT>
+            void set_value(std::vector<TT> mat)
+            {
+                int area = this->w * this->h;
+                assert(area == mat.size());
+                int index = 0;
+                for (int i = 0; i < this->h; i++)
+                {
+                    for (int j = 0; j < this->w; j++)
+                    {
+                        this->array[i][j] = (T)(mat[index++]);
+                    }
+                }
+            }
+
+            /**
+             * @brief Set the matrix value from another matrix.
+             * 
+             * @tparam TT 
+             * @param mat the input matrix.
+             */
+            template <typename TT>
+            void set_value(const Matrix<TT> &mat)
+            {
+                assert((this->h == mat.h) && (this->w == mat.w));
+                for (int i = 0; i < this->h; i++)
+                {
+                    for (int j = 0; j < this->w; j++)
+                    {
+                        this->array[i][j] = (T)(mat.array[i][j]);
+                    }
+                }
+            }
+
+            /**
+             * @brief Set a part of the matrix value from another matrix.
+             * 
+             * @param h_start    the start index of height
+             * @param h_end      the end index of height
+             * @param w_start    the start index of width
+             * @param w_end      the end index of width
+             * @param mat        the input matrix
+             */
+            void set_value(int h_start, int h_end, int w_start, int w_end, const Matrix<T> &mat)
+            {
+                int h = h_end - h_start;
+                int w = w_end - w_start;
+
+                assert((h == mat.h) && (w == mat.w));
+                assert((h_end <= this->h) && (w_end <= this->w) && (h_start >= 0) && (w_start >= 0));
+                for (int i = 0; i < h; i++)
+                {
+                    for (int j = 0; j < w; j++)
+                    {
+                        this->array[i + h_start][j + w_start] = mat.array[i][j];
+                    }
+                }
+            }
+
+            /**
+             * @brief Set the matrix value to a constant.
+             * 
+             * @tparam TT 
+             * @param s  the input value.
+             */
+            template <typename TT>
+            void set_value(TT s)
+            {
+                for (int i = 0; i < this->h; i++)
+                {
+                    for (int j = 0; j < this->w; j++)
+                    {
+                        this->array[i][j] = (T)s;
+                    }
+                }
+            }
+
+            /**
+             * @brief print the matrix element.
+             * 
+             */
+            void print_value() const
+            {
+                printf("h: %d, w: %d\n", this->h, this->w);
+                for (int i = 0; i < this->h; i++)
+                {
+                    for (int j = 0; j < this->w; j++)
+                    {
+                        printf("%f ", (float)(this->array[i][j]));
+                    }
+                    printf("\n");
+                }
+            }
+
+            /**
+             * @brief  do matrix multiply
+             * 
+             * @param input     the input matrix
+             * @return Matrix<T> the output matrix
+             */
+            Matrix<T> matmul(const Matrix<T> &input) const;
+
+            /**
+             * @brief transpose the matrix
+             * 
+             * @return Matrix<T> the transposed matrix
+             */
+            Matrix<T> transpose() const;
+
+            /**
+             * @brief get the inverse matrix
+             * 
+             * @return Matrix<T> the output matrix
+             */
+            Matrix<T> inverse() const;
+
+            /**
+             * @brief get the diagonal of the matrix
+             * 
+             * @return Matrix<T> the diagonal
+             */
+            Matrix<T> diagonal() const;
+
+            /**
+             * @brief slice the matrix
+             * 
+             * @param h_start   the start index of height
+             * @param h_end     the end index of height
+             * @param w_start   the start index of width
+             * @param w_end     the end index of width
+             * @return Matrix<T> the output.
+             */
+            Matrix<T> slice(int h_start, int h_end, int w_start, int w_end) const;
+
+            /**
+             * @brief get an identity matrix
+             * 
+             * @param n     the dim of the identity matrix
+             * @return Matrix<T>  the output
+             */
+            static Matrix<T> identity(int n)
+            {
+                Matrix<T> A(n, n);
+                for (int i = 0; i < n; ++i)
+                {
+                    A.array[i][i] = 1;
+                }
+                return A;
+            }
+
+            /**
+             * @brief get a diag matrix
+             * 
+             * @param d  the diagonal value.
+             * @return Matrix<T>  the output
+             */
+            static Matrix<T> diag(const Matrix<T> &d)
+            {
+                assert(d.h == 1);
+                Matrix<T> A(d.w, d.w);
+                for (int i = 0; i < d.w; ++i)
+                {
+                    A.array[i][i] = d.array[0][i];
+                }
+                return A;
+            }
+
+
+            static Matrix<T> arange(uint32_t n)
+            {
+                Matrix<T> A(1, n);
+                for (int i = 0; i < n; ++i)
+                {
+                    A.array[0][i] = i;
+                }
+                return A;
+            }
+
+            static Matrix<T> arange(uint32_t n1, uint32_t n2)
+            {
+                int len = n2 - n1;
+                assert(len > 0);
+                Matrix<T> A(1, len);
+                for (int i = 0; i < len; ++i)
+                {
+                    A.array[0][i] = n1 + i;
+                }
+
+                return A;
+            }
+
+            /**
+             * @brief get the F_norm of the matrix
+             * 
+             * @return T the output F_norm
+             */
+            T F_norm() const
+            {
+                T f_n = 0.0;
+                for (int i = 0; i < this->h; ++i)
+                {
+                    for (int j = 0; j < this->w; ++j)
+                    {
+                        f_n += (this->array[i][j] * this->array[i][j]);
+                    }
+                }
+                f_n = sqrt_newton(f_n);
+                return f_n;
+            }
+
+            Matrix<T> &operator=(const Matrix<T> &A)
+            {
+                if ((A.h == this->h) && (A.w == this->w))
+                {
+                    for (int i = 0; i < A.h; ++i)
+                    {
+                        for (int j = 0; j < A.w; ++j)
+                        {
+                            this->array[i][j] = A.array[i][j];
+                        }
+                    }
+                }
+                else
+                {
+                    if (this->array != NULL)
+                    {
+                        for (int i = 0; i < this->h; ++i)
+                        {
+                            free(this->array[i]);
+                        }
+                        free(this->array);
+                        this->array = NULL;
+                    }
+                    this->h = A.h;
+                    this->w = A.w;
+                    if ((A.h > 0) && (A.w > 0))
+                    {
+                        this->calloc_element();
+                        this->set_value(A);
+                    }
+                }
+                return *this;
+            }
+        };
+
+        /**
+         * @brief Get the affine transform matrix
+         * 
+         * @param source_coord the source coordinates
+         * @param dest_coord   the target coordinates
+         * @return Matrix<float> the output matrix
+         */
+        Matrix<float> get_affine_transform(Matrix<float> &source_coord, Matrix<float> &dest_coord);
+        
+        /**
+         * @brief Get the similarity transform matrix
+         * 
+         * @param source_coord the source coordinates
+         * @param dest_coord the target coordinates
+         * @return Matrix<float> the output matrix
+         */
+        Matrix<float> get_similarity_transform(Matrix<float> &source_coord, Matrix<float> &dest_coord);
+        
+        /**
+         * @brief Get the perspective transform matrix
+         * 
+         * @param source_coord the source coordinates
+         * @param dest_coord   the target coordinates
+         * @return Matrix<float> the output matrix
+         */
+        Matrix<float> get_perspective_transform(Matrix<float> &source_coord, Matrix<float> &dest_coord);
+    } // namespace math
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/model_zoo/cat_face_detect_mn03.hpp b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/cat_face_detect_mn03.hpp
new file mode 100644
index 00000000000..6d952dc44ab
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/cat_face_detect_mn03.hpp
@@ -0,0 +1,47 @@
+#pragma once
+
+#include <stdint.h>
+#include <vector>
+#include <list>
+#include "dl_detect_define.hpp"
+
+/**
+ * @brief Hardware Requirement.
+ *        - flash 310kB
+ */
+
+class CatFaceDetectMN03
+{
+private:
+    void *model;
+
+public:
+    /**
+     * @brief Construct a new Cat Face Detect MN03 object.
+     * 
+     * @param score_threshold predicted boxes with score lower than the threshold will be filtered out
+     * @param nms_threshold   predicted boxes with IoU higher than the threshold will be filtered out
+     * @param top_k           first k highest score boxes will be remained
+     * @param resize_scale    resize scale to implement on input image
+     */
+    CatFaceDetectMN03(const float score_threshold, const float nms_threshold, const int top_k, const float resize_scale);
+
+    /**
+     * @brief Destroy the Cat Face Detect MN03 object.
+     * 
+     */
+    ~CatFaceDetectMN03();
+
+    /**
+     * @brief Inference.
+     * 
+     * @tparam T supports uint8_t and uint16_t
+     *         - uint8_t: input image is RGB888
+     *         - uint16_t: input image is RGB565
+     * @param input_element pointer of input image
+     * @param input_shape   shape of input image
+     * @return detection result
+     */
+    template <typename T>
+    std::list<dl::detect::result_t> &infer(T *input_element, std::vector<int> input_shape);
+};
diff --git a/tools/sdk/esp32c3/include/esp-face/include/model_zoo/color_detector.hpp b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/color_detector.hpp
new file mode 100644
index 00000000000..063ab20b34a
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/color_detector.hpp
@@ -0,0 +1,68 @@
+#pragma once
+
+#include "dl_image.hpp"
+
+typedef struct
+{
+    int area;                /*!< Area of connected domains >*/
+    std::vector<int> center; /*<! centroid of connected domains [x, y] >*/
+    std::vector<int> box;    /*<! [left_up_x, left_up_y, right_down_x, right_down_y] >*/
+} components_stats_t;
+
+class ColorDetector
+{
+private:
+    std::vector<std::vector<components_stats_t>> results; /*!< detection results >*/
+
+public:
+    std::vector<std::vector<uint8_t>> color_thresh; /*!< threshold of colors, The threshold of each color is composed of 6 numbers >*/
+    std::vector<int> area_thresh;                   /*!< the area threshold of each color, 
+                                                                the area that is smaller than the threshold is filtered >*/
+    bool bgr;                                       /*!< true: the input image is in BGR format
+                                                                false: the input image is in RGB format >*/
+
+    /**
+     * @brief get the color threshold of rectangular region in the image
+     * 
+     * @param image the input image
+     * @param box   the coordinates of the rectanglar region : [left_up_x, left_up_y, right_down_x, right_down_y]
+     * @return std::vector<uint8_t> the threshold.
+     */
+    std::vector<uint8_t> cal_color_thresh(dl::Tensor<uint8_t> &image, std::vector<int> box);
+
+    /**
+     * @brief detect the colors based on the color thresholds
+     * 
+     * @param image the input image.
+     * @return std::vector<std::vector<components_stats_t>>&  detection result.
+     */
+    std::vector<std::vector<components_stats_t>> &detect(dl::Tensor<uint8_t> &image);
+
+    /**
+     * @brief Construct a new Color Detector object
+     * 
+     * @param color_thresh  threshold of colors, The threshold of each color is composed of 6 numbers
+     * @param area_thresh   the area threshold of each color,the area that is smaller than the threshold is filtered
+     * @param bgr           true: the input image is in BGR format
+     *                      false: the input image is in RGB format
+     */
+    ColorDetector(std::vector<std::vector<uint8_t>> color_thresh, std::vector<int> area_thresh, bool bgr = false) : color_thresh(color_thresh), area_thresh(area_thresh), bgr(bgr)
+    {
+    }
+
+    /**
+     * @brief Destroy the Color Detector object
+     * 
+     */
+    ~ColorDetector() {}
+
+    /**
+     * @brief Get the results object
+     * 
+     * @return std::vector<std::vector<components_stats_t>>& the detection result.
+     */
+    std::vector<std::vector<components_stats_t>> &get_results()
+    {
+        return this->results;
+    }
+};
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/model_zoo/face_recognition_112_v1_s16.hpp b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/face_recognition_112_v1_s16.hpp
new file mode 100644
index 00000000000..91f7747fe3f
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/face_recognition_112_v1_s16.hpp
@@ -0,0 +1,30 @@
+#pragma once
+
+#include "dl_variable.hpp"
+#include "face_recognition_tool.hpp"
+#include "face_recognizer.hpp"
+#include <vector>
+
+using namespace dl;
+
+/**
+ * @brief face recognition model v1
+ * input size: 112 x 112 x 3
+ * quantization mode: S16
+ * 
+ */
+class FaceRecognition112V1S16 : public FaceRecognizer<int16_t>
+{       
+    public:
+        /**
+         * @brief Construct a new Face_Recognition_112_V1_S16 object
+         * 
+         */
+        FaceRecognition112V1S16();
+        
+        /**
+         * @brief Destroy the Face_Recognition_112_V1_S16 object
+         * 
+         */
+        ~FaceRecognition112V1S16();
+};
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/model_zoo/face_recognition_112_v1_s8.hpp b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/face_recognition_112_v1_s8.hpp
new file mode 100644
index 00000000000..ba377713ca2
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/face_recognition_112_v1_s8.hpp
@@ -0,0 +1,30 @@
+#pragma once
+
+#include "dl_variable.hpp"
+#include "face_recognition_tool.hpp"
+#include "face_recognizer.hpp"
+#include <vector>
+
+using namespace dl;
+
+/**
+ * @brief face recognition model v1
+ * input size: 112 x 112 x 3
+ * quantization mode: S8
+ * 
+ */
+class FaceRecognition112V1S8 : public FaceRecognizer<int8_t>
+{       
+    public:
+        /**
+         * @brief Construct a new Face_Recognition_112_V1_S8 object
+         * 
+         */
+        FaceRecognition112V1S8();
+
+        /**
+         * @brief Destroy the Face Recognition_112_V1_S8 object
+         * 
+         */
+        ~FaceRecognition112V1S8();
+};
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/model_zoo/face_recognition_tool.hpp b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/face_recognition_tool.hpp
new file mode 100644
index 00000000000..2226d32daf9
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/face_recognition_tool.hpp
@@ -0,0 +1,162 @@
+#pragma once
+
+#include "dl_variable.hpp"
+#include "dl_define.hpp"
+#include "dl_tool.hpp"
+#include "dl_math.hpp"
+#include "dl_math_matrix.hpp"
+#include <vector>
+#include <list>
+#include <algorithm>
+#include <math.h>
+#include <string>
+
+/**
+ * @brief struct of face similarity
+ * 
+ */
+typedef struct
+{
+    int id;
+    std::string name;
+    float similarity;
+} face_info_t;
+
+
+/**
+ * @brief Face ID
+ * 
+ * @tparam feature_t 
+ */
+template <typename feature_t>
+class FaceID
+{
+public:
+    int id;     /*<! id index >*/
+    dl::Tensor<feature_t> id_emb;   /*<! id embedding >*/
+    std::string name;                /*<! id name >*/
+
+    /**
+     * @brief Construct a new Face ID object
+     * 
+     * @param id        id index
+     * @param id_emb    id embedding
+     * @param name      id name
+     */
+    FaceID(int id, dl::Tensor<feature_t> &id_emb, std::string name = "");
+
+    /**
+     * @brief Destroy the Face ID object
+     * 
+     */
+    ~FaceID() {}
+
+    /**
+     * @brief print the face id information
+     * 
+     */
+    void print();
+};
+
+namespace face_recognition_tool
+{
+    /**
+     * @brief l2 normalize the feautre
+     * 
+     * @param feature 
+     */
+    void l2_norm(dl::Tensor<float> &feature);
+
+    /**
+     * @brief calculate the cosine distance of the input ids
+     * 
+     * @param id_1  id 1
+     * @param id_2  id 2
+     * @param normalized_ids true: the input ids have been normalized.
+     *                       false: the input ids have not been normlized 
+     * @param type           0: cos dist: [-1, 1]
+     *                       1: normalzied cos dist: [0, 1]
+     * @return float  the cosine distance
+     */
+    float cos_distance(dl::Tensor<float> &id_1, dl::Tensor<float> &id_2, bool normalized_ids = true, int8_t type = 0);
+
+    /**
+     * @brief transform the image to the input of a mfn model 
+     * 
+     * @tparam T 
+     * @param image             the input image.
+     * @param free_input        true: free the input image.
+     *                          false: do not free the input image.
+     * @param do_padding        true: pad the result.
+     *                          false: do not pad the result. 
+     * @return dl::Tensor<T>* 
+     */
+    template <typename T>
+    dl::Tensor<T> *transform_mfn_input(dl::Tensor<uint8_t> &image, bool free_input = false);
+
+    /**
+     * @brief  transform the image to the input of a mfn model 
+     * 
+     * @tparam T 
+     * @param image                the input image.
+     * @param output               the preprocessed image.
+     * @param free_input           true: free the input image.
+     *                             false: do not free the input image.
+     * @param do_padding           true: pad the result.
+     *                             false: do not pad the result
+     */
+    template <typename T>
+    void transform_mfn_input(dl::Tensor<uint8_t> &image, dl::Tensor<T> &output, bool free_input = false);
+
+    /**
+     * @brief transform the mfn output embedding to a floating embedding
+     * 
+     * @tparam T 
+     * @param input the input embedding. 
+     * @param norm   true: normalize the output embedding.
+     *               false: do not normalize the output embedding.
+     * @param free_input true: free the input embedding.
+     *                   false: do not free the input embedding.
+     * @return dl::Tensor<float>* 
+     */
+    template <typename T>
+    dl::Tensor<float> *transform_mfn_output(dl::Tensor<T> &input, bool norm = true, bool free_input = false);
+
+    /**
+     * @brief transform the mfn output embedding to a floating embedding
+     * 
+     * @tparam T 
+     * @param input         the input embedding. 
+     * @param output        the output embedding.
+     * @param norm          true: normalize the output embedding.
+     *                      false: do not normalize the output embedding.
+     * @param free_input    true: free the input embedding.
+     *                      false: do not free the input embedding.
+     */
+    template <typename T>
+    void transform_mfn_output(dl::Tensor<T> &input, dl::Tensor<float> &output, bool norm = true, bool free_input = false);
+
+    /**
+     * @brief get the aligned face.
+     * 
+     * @tparam T 
+     * @param input     input tensor 
+     * @param output    the output aligned face.
+     * @param landmarks the landmarks of the face.
+     */
+    template <typename T>
+    void align_face(dl::Tensor<T> *input, dl::Tensor<T> *output, std::vector<int> &landmarks);
+
+    /**
+     * @brief get the aligned face.
+     * 
+     * @tparam T 
+     * @param input     input image with rgb565 format.
+     * @param shape     the shape of the input image.
+     * @param output    the output aligned face.
+     * @param landmarks the landmarks of the face.
+     */
+    template <typename T>
+    void align_face(uint16_t *input, std::vector<int> shape, dl::Tensor<T> *output, std::vector<int> &landmarks);
+
+} // namespace face_recognition_tool
diff --git a/tools/sdk/esp32c3/include/esp-face/include/model_zoo/face_recognizer.hpp b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/face_recognizer.hpp
new file mode 100644
index 00000000000..864881c984a
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/face_recognizer.hpp
@@ -0,0 +1,220 @@
+#pragma once
+
+#include "dl_variable.hpp"
+#include "face_recognition_tool.hpp"
+#include <vector>
+
+using namespace dl;
+
+/**
+ * @brief 
+ * 
+ * @tparam feature_t 
+ */
+template<typename feature_t>
+class FaceRecognizer
+{
+    public:
+        /**
+         * @brief Construct a new Face Recognizer object
+         * 
+         */
+        FaceRecognizer();
+
+        /**
+         * @brief Destroy the Face Recognizer object
+         * 
+         */
+        virtual ~FaceRecognizer();
+
+        void *model;
+
+        /**
+         * @brief Set the face recognition threshold [-1, 1], default thresh: 0.55
+         * Note: If the similarity of two faces is greater than the threshold, they will be judged as the same person
+         * 
+         * @param thresh 
+         */
+        void set_thresh(float thresh);
+
+        /**
+         * @brief Get the current threshold of recognizer.
+         * 
+         * @return float current threshold.
+         */
+        float get_thresh();
+
+        /**
+         * @brief Get the input shape of the recognizer.
+         * 
+         * @return std::vector<int> the input shape of the recognizer. 
+         */
+        std::vector<int> get_input_shape();
+
+        /**
+         * @brief do forward
+         * 
+         * @param model_input            the input data of the face recognition model. 
+         * Note: the input data should have been preprocessed.
+         * @return Tensor<feature_t>&    the output of the face recognition model.
+         */
+        Tensor<feature_t> &forward(Tensor<feature_t> &model_input);
+
+        /**
+         * @brief recognize face
+         * 
+         * @param image_input       the pointer of the input image with format bgr565.
+         * @param shape             the shape of the input image
+         * @param landmarks         face landmarks coordinates
+         * @return face_info_t      the recognition result.
+         */
+        face_info_t recognize(uint16_t *image_input, std::vector<int> shape, std::vector<int> &landmarks);
+        
+        /**
+         * @brief recognize face
+         * 
+         * @param image_input        the pointer of the input image with format bgr565.
+         * @param shape              the shape of the input image  
+         * @param aligned_face       the Tensor to store the intermeidate aligned face.
+         * @param landmarks          face landmarks coordinates
+         * @return face_info_t       the recognition result.
+         */
+        face_info_t recognize(uint16_t *image_input, std::vector<int> shape, Tensor<uint8_t> &aligned_face, std::vector<int> &landmarks);
+
+        /**
+         * @brief recognize face
+         * 
+         * @param image_input         the Tensor of input image with format bgr888.
+         * @param landmarks           face landmarks coordinates
+         * @return face_info_t        the recognition result.
+         */
+        face_info_t recognize(Tensor<uint8_t> &image_input, std::vector<int> &landmarks);
+
+        /**
+         * @brief recognize face
+         * 
+         * @param image_input           the Tensor of input image with format bgr888.
+         * @param aligned_face          the Tensor to store the intermeidate aligned face.
+         * @param landmarks             face landmarks coordinates
+         * @return face_info_t          the recognition result.
+         */
+        face_info_t recognize(Tensor<uint8_t> &image_input, Tensor<uint8_t> &aligned_face, std::vector<int> &landmarks);
+
+        /**
+         * @brief recognize face
+         * 
+         * @param aligned_face          the Tensor of the input aligned face with format bgr888.
+         * @return face_info_t          the recognition result.
+         */
+        face_info_t recognize(Tensor<uint8_t> &aligned_face);
+
+        /**
+         * @brief recognize the face embedding.
+         * 
+         * @param emb  the normalized face embbeding.
+         * @return face_info_t  the recognition result.
+         */
+        face_info_t recognize(Tensor<float> &emb);
+
+        /**
+         * @brief Get the index of the enrolled ids 
+         * 
+         * @return std::vector<int> a vector of face ids index
+         */
+        std::vector<face_info_t> get_enrolled_ids();
+
+        /**
+         * @brief Get the face embedding 
+         * 
+         * @param id the face id index
+         * @return Tensor<float>  the face embedding of the face id index. 
+         *                        if there is no matched id return the embedding of last input image.
+         */
+        Tensor<float> &get_face_emb(int id=-1);
+
+        /**
+         * @brief Get the number of enrolled id
+         * 
+         * @return int the number of enrolled id
+         */
+        int get_enrolled_id_num();
+
+        /**
+         * @brief enroll face id
+         * 
+         * @param image_input       the pointer of the input image with format bgr565.
+         * @param shape             the shape of the input image
+         * @param landmarks         face landmarks coordinates
+         * @param name              name of the face id.
+         * @return  int             the face id index of the enrolled embedding.
+         */
+        int enroll_id(uint16_t *image_input, std::vector<int> shape, std::vector<int> &landmarks, std::string name="");
+        
+        /**
+         * @brief enroll face id
+         * 
+         * @param image_input        the pointer of the input image with format bgr565.
+         * @param shape              the shape of the input image  
+         * @param aligned_face       the Tensor to store the intermeidate aligned face.
+         * @param landmarks          face landmarks coordinates
+         * @param name               name of the face id.
+         * @return  int              the face id index of the enrolled embedding.
+         */
+        int enroll_id(uint16_t *image_input, std::vector<int> shape, Tensor<uint8_t> &aligned_face, std::vector<int> &landmarks, std::string name="");
+
+        /**
+         * @brief enroll face id
+         * 
+         * @param image_input         the Tensor of input image with format bgr888.
+         * @param landmarks           face landmarks coordinates
+         * @param name                name of the face id.
+         * @return  int               the face id index of the enrolled embedding.
+         */
+        int enroll_id(Tensor<uint8_t> &image_input, std::vector<int> &landmarks, std::string name="");
+
+        /**
+         * @brief enroll face id
+         * 
+         * @param image_input           the Tensor of input image with format bgr888.
+         * @param aligned_face          the Tensor to store the intermeidate aligned face.
+         * @param landmarks             face landmarks coordinates
+         * @param name                  name of the face id.
+         * @return  int                 the face id index of the enrolled embedding.
+         */
+        int enroll_id(Tensor<uint8_t> &image_input, Tensor<uint8_t> &aligned_face, std::vector<int> &landmarks, std::string name="");
+
+        /**
+         * @brief enroll face id
+         * 
+         * @param aligned_face          the Tensor of the input aligned face with format bgr888.
+         * @param name                  name of the face id.
+         * @return  int                 the face id index of the enrolled embedding.
+         */
+        int enroll_id(Tensor<uint8_t> &aligned_face, std::string name="");
+        
+        /**
+         * @brief       enroll the normalzied face embedding.
+         * 
+         * @param emb   the normalized face embbeding.
+         * @param name  name of the face id.
+         * @return int  the face id index of the enrolled embedding.
+         */
+        int enroll_id(Tensor<float> &emb, std::string name="");
+
+        /**
+         * @brief       delete the last enrolled face id.
+         * 
+         * @return int  the number of remained face ids.
+         *              if the face ids list is empty, return -1
+         */
+        int delete_id();
+
+        /**
+         * @brief       delete the face id with id index.
+         * 
+         * @param id    face id index.
+         * @return int  the number of remained face ids.
+         *              if there is no matched id return -1
+         */
+        int delete_id(int id);
+};
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/model_zoo/human_face_detect_mnp01.hpp b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/human_face_detect_mnp01.hpp
new file mode 100644
index 00000000000..62a25f0350c
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/human_face_detect_mnp01.hpp
@@ -0,0 +1,41 @@
+#pragma once
+
+#include <vector>
+#include <list>
+#include "dl_detect_define.hpp"
+
+class HumanFaceDetectMNP01
+{
+private:
+    void *model;
+
+public:
+    /**
+     * @brief Construct a new Human Face Detect MNP01 object.
+     * 
+     * @param score_threshold predicted boxes with score lower than the threshold will be filtered out
+     * @param nms_threshold   predicted boxes with IoU higher than the threshold will be filtered out
+     * @param top_k           first k highest score boxes will be remained
+     */
+    HumanFaceDetectMNP01(const float score_threshold, const float nms_threshold, const int top_k);
+
+    /**
+     * @brief Destroy the Human Face Detect MNP01 object.
+     * 
+     */
+    ~HumanFaceDetectMNP01();
+
+    /**
+     * @brief Inference.
+     * 
+     * @tparam T supports uint16_t and uint8_t,
+     *         - uint16_t: input image is RGB565
+     *         - uint8_t: input image is RGB888
+     * @param input_element pointer of input image
+     * @param input_shape   shape of input image
+     * @param candidates    candidate boxes on input image
+     * @return detection result
+     */
+    template <typename T>
+    std::list<dl::detect::result_t> &infer(T *input_element, std::vector<int> input_shape, std::list<dl::detect::result_t> &candidates);
+};
diff --git a/tools/sdk/esp32c3/include/esp-face/include/model_zoo/human_face_detect_msr01.hpp b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/human_face_detect_msr01.hpp
new file mode 100644
index 00000000000..94ba6ac909a
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/model_zoo/human_face_detect_msr01.hpp
@@ -0,0 +1,40 @@
+#pragma once
+
+#include <list>
+#include <vector>
+#include "dl_detect_define.hpp"
+
+class HumanFaceDetectMSR01
+{
+private:
+    void *model;
+
+public:
+    /**
+     * @brief Construct a new Human Face Detect MSR01 object
+     * 
+     * @param score_threshold   predicted boxes with score lower than the threshold will be filtered out
+     * @param nms_threshold     predicted boxes with IoU higher than the threshold will be filtered out
+     * @param top_k             first k highest score boxes will be remained
+     * @param resize_scale      resize scale to implement on input image
+     */
+    HumanFaceDetectMSR01(const float score_threshold, const float nms_threshold, const int top_k, float resize_scale);
+
+    /**
+     * @brief Destroy the Human Face Detect MSR01 object
+     */
+    ~HumanFaceDetectMSR01();
+
+    /**
+     * @brief Inference.
+     * 
+     * @tparam T supports uint8_t and uint16_t
+     *         - uint8_t: input image is RGB888
+     *         - uint16_t: input image is RGB565
+     * @param input_element pointer of input image
+     * @param input_shape   shape of input image
+     * @return detection result
+     */
+    template <typename T>
+    std::list<dl::detect::result_t> &infer(T *input_element, std::vector<int> input_shape);
+};
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn.hpp
new file mode 100644
index 00000000000..6c737c1d341
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn.hpp
@@ -0,0 +1,61 @@
+#pragma once
+#include <vector>
+#include "dl_define.hpp"
+#include "dl_tool.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief Get the output shape object
+         * 
+         * @param input_shape  input shape
+         * @param filter_shape filter shape with dilation
+         * @param stride_y     stride in height
+         * @param stride_x     stride in width
+         * @param pad_type     one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN
+         * @param is_conv2d    one of true or false,
+         *                     - true: serve for Conv2D
+         *                     - false: serve for other operations
+         * @return std::vector<int> 
+         */
+        std::vector<int> get_output_shape(const std::vector<int> &input_shape, const std::vector<int> &filter_shape, const int stride_y, const int stride_x, const padding_type_t pad_type, const bool is_conv2d = false, std::vector<int> padding = {});
+
+        /**
+         * @brief Get the pad size object
+         * 
+         * @param output_shape output shape
+         * @param input_shape  input shape
+         * @param filter_shape filter shape with dilation
+         * @param stride_y     stride in height
+         * @param stride_x     stride in width
+         * @param padding_type one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN
+         * @return padding size
+         */
+        std::vector<int> get_pad_size(const std::vector<int> &output_shape, const std::vector<int> &input_shape, const std::vector<int> &filter_shape, const int stride_y, const int stride_x, const padding_type_t padding_type);
+    } // namespace nn
+} // namespace dl
+
+#if DL_LOG_NN_LATENCY
+/**
+ * @brief Initialize.
+ */
+#define DL_LOG_NN_LATENCY_INIT() dl::tool::Latency latency
+
+/**
+ * @brief Time starts.
+ */
+#define DL_LOG_NN_LATENCY_START() latency.start()
+
+/**
+ * @brief Time ends and printed.
+ */
+#define DL_LOG_NN_LATENCY_END(key) \
+    latency.end();                 \
+    latency.print("nn", key)
+#else
+#define DL_LOG_NN_LATENCY_INIT()
+#define DL_LOG_NN_LATENCY_START()
+#define DL_LOG_NN_LATENCY_END(key)
+#endif
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_add2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_add2d.hpp
new file mode 100644
index 00000000000..4d4daaaef05
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_add2d.hpp
@@ -0,0 +1,91 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief activation(add2d(input0, input1)).
+         * 
+         * @param output            as an output
+         * @param input0            as one input
+         * @param input1            as another input
+         * @param activation        activation of add2d, if you don't specify anything, no activation is applied
+         * @param assign_core       not effective yet
+         * @param output_exponent   exponent of output, only and must specify if inplace operation happens 
+         */
+        void add2d(Tensor<int16_t> &output,
+                   Tensor<int16_t> &input0,
+                   Tensor<int16_t> &input1,
+                   const Activation<int16_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
+
+        /**
+         * @brief activation(add2d(input0, input1)).
+         * 
+         * @param output            as an output
+         * @param input0            as one input
+         * @param input1            as another input
+         * @param activation        activation of add2d, if you don't specify anything, no activation is applied
+         * @param assign_core       not effective yet
+         * @param output_exponent   exponent of output, only and must specify if inplace operation happens 
+         */
+        void add2d(Tensor<int8_t> &output,
+                   Tensor<int8_t> &input0,
+                   Tensor<int8_t> &input1,
+                   const Activation<int8_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE, const int output_exponent = INT_MIN);
+
+        /**
+         * @brief activation(add2d(input0, input1))
+         * 
+         * @tparam inplace: whether directly store the output to input0
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param output_exponent exponent of output
+         * @param input0          as one input
+         * @param input1          as another input
+         * @param activation      activation of add2d, if you don't specify anything, no activation is applied
+         * @param assign_core     not effective yet
+         * @param inplace         whether directly store the output to input0
+         * @return add2d result or no return(result store to input0)
+         */
+        template <bool inplace = false, typename feature_t>
+        auto add2d(const int output_exponent,
+                   Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const Activation<feature_t> *activation,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        {
+            assert(input0.is_same_shape(input1));
+
+            DL_LOG_NN_LATENCY_INIT();
+
+            Tensor<feature_t> output;
+            if constexpr (!inplace)
+            {
+                DL_LOG_NN_LATENCY_START();
+                output.set_exponent(output_exponent).set_shape(input0.shape).malloc_element();
+                DL_LOG_NN_LATENCY_END("apply");
+
+                DL_LOG_NN_LATENCY_START();
+                add2d(output, input0, input1, activation, assign_core);
+                DL_LOG_NN_LATENCY_END("add2d");
+                return output;
+            }
+            else
+            {
+                DL_LOG_NN_LATENCY_START();
+                add2d(input0, input0, input1, activation, assign_core, output_exponent);
+                input0.set_exponent(output_exponent);
+                DL_LOG_NN_LATENCY_END("add2d");
+            }
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_avg_pool2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_avg_pool2d.hpp
new file mode 100644
index 00000000000..6e7db6ed0e0
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_avg_pool2d.hpp
@@ -0,0 +1,102 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+#include <stdint.h>
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief avg_pool2d(input).
+         * 
+         * @param output        as an output
+         * @param input         as an input
+         * @param padding       padding size needed in [top, bottom, left, right] of this operation
+         * @param filter_shape  filter_shape in [filter_height, filter_width]
+         * @param stride_y      stride in height
+         * @param stride_x      stride in width
+         * @param assign_core   not effective yet
+         */
+        void avg_pool2d(Tensor<int16_t> &output,
+                        Tensor<int16_t> &input,
+                        std::vector<int> &padding,
+                        std::vector<int> &filter_shape,
+                        const int stride_y,
+                        const int stride_x,
+                        const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief avg_pool2d(input).
+         * 
+         * @param output        as an output
+         * @param input         as an input
+         * @param padding       padding size needed in [top, bottom, left, right] of this operation
+         * @param filter_shape  filter_shape in [filter_height, filter_width]
+         * @param stride_y      stride in height
+         * @param stride_x      stride in width
+         * @param assign_core   not effective yet
+         */
+        void avg_pool2d(Tensor<int8_t> &output,
+                        Tensor<int8_t> &input,
+                        std::vector<int> &padding,
+                        std::vector<int> &filter_shape,
+                        const int stride_y,
+                        const int stride_x,
+                        const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief avg_pool2d(input).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param output_exponent exponent of output
+         * @param input           as an input
+         * @param filter_shape    filter_shape in [filter_height, filter_width]
+         * @param stride_y        stride in height
+         * @param stride_x        stride in width
+         * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN,
+         *                        - PADDING_VALID: no padding
+         *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
+         *                        such that output has the same height/width dimension as the input,
+         *                        - PADDING_SAME_END results padding in TensorFlow style
+         *                        - PADDING_SAME_BEGIN results padding in MXNET style
+         * @param assign_core     not effective yet
+         * @return avg_pool2d result
+         */
+        template <typename feature_t>
+        Tensor<feature_t> avg_pool2d(const int output_exponent,
+                                     Tensor<feature_t> &input,
+                                     std::vector<int> filter_shape,
+                                     const int stride_y,
+                                     const int stride_x,
+                                     const padding_type_t padding_type,
+                                     const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+        {
+            DL_LOG_NN_LATENCY_INIT();
+
+            DL_LOG_NN_LATENCY_START();
+            std::vector<int> output_shape = get_output_shape(input.shape, filter_shape, stride_y, stride_x, padding_type);
+            Tensor<feature_t> output;
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
+            DL_LOG_NN_LATENCY_END("apply");
+            std::vector<int> padding(4, 0);
+
+            DL_LOG_NN_LATENCY_START();
+            if (padding_type == PADDING_SAME_END || padding_type == PADDING_SAME_BEGIN)
+            {
+                padding = get_pad_size(output_shape, input.shape, filter_shape, stride_y, stride_x, padding_type);
+            }
+            DL_LOG_NN_LATENCY_END("padding");
+
+            DL_LOG_NN_LATENCY_START();
+            avg_pool2d(output, input, padding, filter_shape, stride_y, stride_x, assign_core);
+            DL_LOG_NN_LATENCY_END("avg_pool2d");
+
+            return output;
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_concat.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_concat.hpp
new file mode 100644
index 00000000000..73ed1aae905
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_concat.hpp
@@ -0,0 +1,63 @@
+#pragma once
+
+#include <vector>
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        template <typename feature_t>
+        void concat(Tensor<feature_t> &output, std::vector<Tensor<feature_t> *> &inputs, int axis, bool free_inputs = false);
+
+        template <typename feature_t>
+        Tensor<feature_t> concat(std::vector<Tensor<feature_t> *> &inputs, int axis, bool free_inputs = false)
+        {
+            DL_LOG_NN_LATENCY_INIT();
+
+            DL_LOG_NN_LATENCY_START();
+            assert(inputs.size() > 1);
+            int shape_size = inputs[0]->shape.size();
+
+            if (axis < 0)
+            {
+                axis = shape_size + axis;
+            }
+
+            assert((axis < shape_size) && (axis > -1));
+
+            int output_shape_axis = inputs[0]->shape[axis];
+
+            for (int i = 1; i < inputs.size(); i++)
+            {
+                assert(shape_size == inputs[i]->shape.size());
+                assert(inputs[i]->exponent == inputs[i - 1]->exponent);
+                output_shape_axis += inputs[i]->shape[axis];
+
+                for (int j = 0; j < shape_size; j++)
+                {
+                    if (j != axis)
+                    {
+                        assert(inputs[i]->shape[j] == inputs[i - 1]->shape[j]);
+                    }
+                }
+            }
+            DL_LOG_NN_LATENCY_END("assert");
+
+            DL_LOG_NN_LATENCY_START();
+            Tensor<feature_t> output;
+            std::vector<int> output_shape = inputs[0]->shape;
+            output_shape[axis] = output_shape_axis;
+            output.set_shape(output_shape);
+            output.set_exponent(inputs[0]->exponent);
+            output.malloc_element();
+            DL_LOG_NN_LATENCY_END("malloc");
+
+            DL_LOG_NN_LATENCY_START();
+            concat(output, inputs, axis, free_inputs);
+            DL_LOG_NN_LATENCY_END("concat");
+            return output;
+        }
+    } // namespace nn
+} // namespace dl
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_concat2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_concat2d.hpp
new file mode 100644
index 00000000000..adcae1e7a7d
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_concat2d.hpp
@@ -0,0 +1,22 @@
+#pragma once
+
+#include <vector>
+#include "dl_variable.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief concat2d(input_1, input_2, ...)
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param output as an output
+         * @param inputs a bundle of inputs to be concatenated
+         */
+        template <typename feature_t>
+        void concat2d(Tensor<feature_t> &output, std::vector<Tensor<feature_t>> inputs);
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_conv2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_conv2d.hpp
new file mode 100644
index 00000000000..27ba0372730
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_conv2d.hpp
@@ -0,0 +1,136 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief activation(conv2d(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param padding     padding size needed in [top, bottom, left, right] of this operation
+         * @param filter      filter of conv2d
+         * @param stride_y    stride in height
+         * @param stride_x    stride in width
+         * @param bias        bias of conv2d, if you don't specify anything, no bias is added
+         * @param activation  activation of conv2d, if you don't specify anything, no activation is applied
+         * @param assign_core not effective yet
+         */
+        void conv2d(Tensor<int16_t> &output,
+                    Tensor<int16_t> &input,
+                    std::vector<int> &padding,
+                    const Filter<int16_t> &filter,
+                    const int stride_y,
+                    const int stride_x,
+                    const Bias<int16_t> *const bias = NULL,
+                    const Activation<int16_t> *const activation = NULL,
+                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(conv2d(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param padding     padding size needed in [top, bottom, left, right] of this operation
+         * @param filter      filter of conv2d
+         * @param stride_y    stride in height
+         * @param stride_x    stride in width
+         * @param bias        bias of conv2d, if you don't specify anything, no bias is added
+         * @param activation  activation of conv2d, if you don't specify anything, no activation is applied
+         * @param assign_core not effective yet
+         */
+        void conv2d(Tensor<int8_t> &output,
+                    Tensor<int8_t> &input,
+                    std::vector<int> &padding,
+                    const Filter<int8_t> &filter,
+                    const int stride_y,
+                    const int stride_x,
+                    const Bias<int8_t> *const bias = NULL,
+                    const Activation<int8_t> *const activation = NULL,
+                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(conv2d(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param padding     padding size needed in [top, bottom, left, right] of this operation
+         * @param filter      filter of conv2d
+         * @param stride_y    stride in height
+         * @param stride_x    stride in width
+         * @param bias        bias of conv2d, if you don't specify anything, no bias is added
+         * @param activation  activation of conv2d, if you don't specify anything, no activation is applied
+         * @param assign_core not effective yet
+         */
+        void conv2d(Tensor<int8_t> &output,
+                    Tensor<int8_t> &input,
+                    std::vector<int> &padding,
+                    const Filter<int8_t> &filter,
+                    const int stride_y,
+                    const int stride_x,
+                    const Bias<int16_t> *const bias = NULL,
+                    const Activation<int8_t> *const activation = NULL,
+                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(conv2d(input, filter) + bias).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param output_exponent exponent of output
+         * @param input           as an input
+         * @param filter          Filter of conv2d
+         * @param stride_y        stride in height
+         * @param stride_x        stride in width
+         * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN,
+         *                        - PADDING_VALID: no padding
+         *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
+         *                        such that output has the same height/width dimension as the input,
+         *                        - PADDING_SAME_END results padding in TensorFlow style
+         *                        - PADDING_SAME_BEGIN results padding in MXNET style
+         * @param bias            bias of conv2d, if you don't specify anything, no bias is added
+         * @param activation      activation of conv2d, if you don't specify anything, no activation is applied
+         * @param assign_core     not effective yet
+         * @return conv2d result
+         */
+        template <typename feature_t, typename bias_t>
+        Tensor<feature_t> conv2d(const int output_exponent,
+                                 Tensor<feature_t> &input,
+                                 const Filter<feature_t> &filter,
+                                 const int stride_y,
+                                 const int stride_x,
+                                 const padding_type_t padding_type,
+                                 const Bias<bias_t> *bias,
+                                 const Activation<feature_t> *activation,
+                                 const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+        {
+            DL_LOG_NN_LATENCY_INIT();
+
+            DL_LOG_NN_LATENCY_START();
+            std::vector<int> output_shape = get_output_shape(input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type, true);
+            Tensor<feature_t> output;
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
+            DL_LOG_NN_LATENCY_END("apply");
+
+            std::vector<int> padding(4, 0);
+            DL_LOG_NN_LATENCY_START();
+            if (padding_type == PADDING_SAME_END || padding_type == PADDING_SAME_BEGIN)
+            {
+                padding = get_pad_size(output_shape, input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
+            }
+            DL_LOG_NN_LATENCY_END("padding");
+
+            DL_LOG_NN_LATENCY_START();
+            conv2d(output, input, padding, filter, stride_y, stride_x, bias, activation, assign_core);
+            DL_LOG_NN_LATENCY_END("conv2d");
+
+            return output;
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_depthwise_conv2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_depthwise_conv2d.hpp
new file mode 100644
index 00000000000..135815a65cb
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_depthwise_conv2d.hpp
@@ -0,0 +1,137 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief activate(depthwise_conv2d(input, filter) + bias)
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param padding     padding size needed in [top, bottom, left, right] of this operation
+         * @param filter      Filter of depthwise_conv2d
+         * @param stride_y    stride in height
+         * @param stride_x    stride in width
+         * @param bias        bias of depthwise_conv2d, if you don't specify anything, no bias is added
+         * @param activation  activation of depthwise_conv2d, if you don't specify anything, no activation is applied
+         * @param assign_core not effective yet
+         */
+        void depthwise_conv2d(Tensor<int16_t> &output,
+                              Tensor<int16_t> &input,
+                              std::vector<int> &padding,
+                              const Filter<int16_t> &filter,
+                              const int stride_y,
+                              const int stride_x,
+                              const Bias<int16_t> *bias = NULL,
+                              const Activation<int16_t> *activation = NULL,
+                              const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activate(depthwise_conv2d(input, filter) + bias)
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param padding     padding size needed in [top, bottom, left, right] of this operation
+         * @param filter      filter of depthwise_conv2d
+         * @param stride_y    stride in height
+         * @param stride_x    stride in width
+         * @param bias        bias of depthwise_conv2d, if you don't specify anything, no bias is added
+         * @param activation  activation of depthwise_conv2d, if you don't specify anything, no activation is applied
+         * @param assign_core not effective yet
+         */
+        void depthwise_conv2d(Tensor<int8_t> &output,
+                              Tensor<int8_t> &input,
+                              std::vector<int> &padding,
+                              const Filter<int8_t> &filter,
+                              const int stride_y,
+                              const int stride_x,
+                              const Bias<int8_t> *bias = NULL,
+                              const Activation<int8_t> *activation = NULL,
+                              const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activate(depthwise_conv2d(input, filter) + bias)
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param padding     padding size needed in [top, bottom, left, right] of this operation
+         * @param filter      Filter of depthwise_conv2d
+         * @param stride_y    stride in height
+         * @param stride_x    stride in width
+         * @param bias        bias of depthwise_conv2d, if you don't specify anything, no bias is added
+         * @param activation  activation of depthwise_conv2d, if you don't specify anything, no activation is applied
+         * @param assign_core not effective yet
+         */
+        void depthwise_conv2d(Tensor<int8_t> &output,
+                              Tensor<int8_t> &input,
+                              std::vector<int> &padding,
+                              const Filter<int8_t> &filter,
+                              const int stride_y,
+                              const int stride_x,
+                              const Bias<int16_t> *bias = NULL,
+                              const Activation<int8_t> *activation = NULL,
+                              const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(depthwise_conv2d(input, filter) + bias)
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param output_exponent exponent of output
+         * @param input           as an input
+         * @param filter          filter of depthwise_conv2d
+         * @param stride_y        stride in height
+         * @param stride_x        stride in width
+         * @param pad_type        one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN,
+         *                        - PADDING_VALID means no padding
+         *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
+         *                        such that output has the same height/width dimension as the input,
+         *                        - PADDING_SAME_END results padding in TensorFlow style
+         *                        - PADDING_SAME_BEGIN results padding in MXNET style
+         * @param bias            bias of depthwise_conv2d, if you don't specify anything, no bias is added
+         * @param activation      activation of depthwise_conv2d, if you don't specify anything, no activation is applied
+         * @param assign_core     not effective yet
+         * @return depthwise_conv2d result
+         */
+        template <typename feature_t, typename bias_t>
+        Tensor<feature_t> depthwise_conv2d(const int output_exponent,
+                                           Tensor<feature_t> &input,
+                                           const Filter<feature_t> &filter,
+                                           const int stride_y,
+                                           const int stride_x,
+                                           const padding_type_t padding_type,
+                                           const Bias<bias_t> *bias,
+                                           const Activation<feature_t> *activation,
+                                           const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+        {
+            DL_LOG_NN_LATENCY_INIT();
+
+            DL_LOG_NN_LATENCY_START();
+            std::vector<int> output_shape = get_output_shape(input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
+            Tensor<feature_t> output;
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
+            DL_LOG_NN_LATENCY_END("apply");
+
+            std::vector<int> padding(4, 0);
+
+            DL_LOG_NN_LATENCY_START();
+            if (padding_type == PADDING_SAME_END || padding_type == PADDING_SAME_BEGIN)
+            {
+                padding = get_pad_size(output_shape, input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
+            }
+            DL_LOG_NN_LATENCY_END("padding");
+
+            DL_LOG_NN_LATENCY_START();
+            depthwise_conv2d(output, input, padding, filter, stride_y, stride_x, bias, activation, assign_core);
+            DL_LOG_NN_LATENCY_END("depthwise_conv2d");
+
+            return output;
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_fully_connected.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_fully_connected.hpp
new file mode 100644
index 00000000000..372c84825fd
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_fully_connected.hpp
@@ -0,0 +1,126 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief activation(FullyConnected(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param filter      filter of FullyConnected
+         * @param bias        bias of FullyConnected, if you don't specify anything, no bias is added
+         * @param activation  activation of FullyConnected, if you don't specify anything, no activation is applied
+         * @param flatten     true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+         *                    false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+         * @param assign_core not effective yet
+         */
+        void fully_connected(Tensor<int16_t> &output,
+                             Tensor<int16_t> &input,
+                             const Filter<int16_t> &filter,
+                             const Bias<int16_t> *const bias = NULL,
+                             const Activation<int16_t> *const activation = NULL,
+                             const bool flatten = true,
+                             const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(FullyConnected(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param filter      filter of FullyConnected
+         * @param bias        bias of FullyConnected, if you don't specify anything, no bias is added
+         * @param activation  activation of FullyConnected, if you don't specify anything, no activation is applied
+         * @param flatten     true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+         *                    false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+         * @param assign_core not effective yet
+         */
+        void fully_connected(Tensor<int8_t> &output,
+                             Tensor<int8_t> &input,
+                             const Filter<int8_t> &filter,
+                             const Bias<int8_t> *const bias = NULL,
+                             const Activation<int8_t> *const activation = NULL,
+                             const bool flatten = true,
+                             const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(FullyConnected(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param filter      filter of FullyConnected
+         * @param bias        bias of FullyConnected, if you don't specify anything, no bias is added
+         * @param activation  activation of FullyConnected, if you don't specify anything, no activation is applied
+         * @param flatten     true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+         *                    false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+         * @param assign_core not effective yet
+         */
+        void fully_connected(Tensor<int8_t> &output,
+                             Tensor<int8_t> &input,
+                             const Filter<int8_t> &filter,
+                             const Bias<int16_t> *const bias = NULL,
+                             const Activation<int8_t> *const activation = NULL,
+                             const bool flatten = true,
+                             const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(FullyConnected(input, filter) + bias).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param output_exponent exponent of output
+         * @param input           as an input
+         * @param filter          Filter of FullyConnected
+         * @param bias            bias of FullyConnected, if you don't specify anything, no bias is added
+         * @param activation      activation of FullyConnected, if you don't specify anything, no activation is applied
+         * @param flatten         true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+         *                        false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+         * @param assign_core     not effective yet
+         * @return FullyConnected result
+         */
+        template <typename feature_t>
+        Tensor<feature_t> fully_connected(const int output_exponent,
+                                          Tensor<feature_t> &input,
+                                          const Filter<feature_t> &filter,
+                                          const Bias<feature_t> *bias,
+                                          const Activation<feature_t> *activation,
+                                          const bool flatten,
+                                          const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+        {
+            DL_LOG_NN_LATENCY_INIT();
+
+            DL_LOG_NN_LATENCY_START();
+            assert(filter.shape.size() == 4);
+            assert(filter.shape[0] == 1);
+            assert(filter.shape[1] == 1);
+
+            std::vector<int> output_shape;
+            if (flatten)
+            {
+                assert(input.get_size() == filter.shape[2]);
+                output_shape = {filter.shape.back()};
+            }
+            else
+            {
+                assert(input.shape.back() == filter->shape[2]);
+                output_shape = input.shape;
+                output_shape[output_shape.size() - 1] = filter.shape.back();
+            }
+            Tensor<feature_t> output;
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
+            DL_LOG_NN_LATENCY_END("apply");
+
+            DL_LOG_NN_LATENCY_START();
+            fully_connected(output, input, filter, bias, activation, flatten, assign_core);
+            DL_LOG_NN_LATENCY_END("fully_connected");
+
+            return output;
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_global_avg_pool2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_global_avg_pool2d.hpp
new file mode 100644
index 00000000000..724d3ca0952
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_global_avg_pool2d.hpp
@@ -0,0 +1,66 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+#include <stdint.h>
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief global_avg_pool2d(input).
+         * 
+         * @param output        as an output
+         * @param input         as an input
+         * @param assign_core   not effective yet
+         */
+        void global_avg_pool2d(Tensor<int16_t> &output,
+                               Tensor<int16_t> &input,
+                               const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief global_avg_pool2d(input).
+         * 
+         * @param output        as an output
+         * @param input         as an input
+         * @param assign_core   not effective yet
+         */
+        void global_avg_pool2d(Tensor<int8_t> &output,
+                               Tensor<int8_t> &input,
+                               const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief global_avg_pool2d(input).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param output_exponent exponent of output
+         * @param input           as an input
+         * @param assign_core     not effective yet
+         * @return global_avg_pool2d result
+         */
+        template <typename feature_t>
+        Tensor<feature_t> global_avg_pool2d(const int output_exponent,
+                                            Tensor<feature_t> &input,
+                                            const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+        {
+            DL_LOG_NN_LATENCY_INIT();
+
+            DL_LOG_NN_LATENCY_START();
+            std::vector<int> output_shape(input.shape.size(), 1);
+            output_shape[2] = input.shape[2];
+            Tensor<feature_t> output;
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
+            DL_LOG_NN_LATENCY_END("apply");
+
+            DL_LOG_NN_LATENCY_START();
+            global_avg_pool2d(output, input, assign_core);
+            DL_LOG_NN_LATENCY_END("global_avg_pool2d");
+
+            return output;
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_global_max_pool2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_global_max_pool2d.hpp
new file mode 100644
index 00000000000..f6f15e91bd3
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_global_max_pool2d.hpp
@@ -0,0 +1,64 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+#include <stdint.h>
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief global_max_pool2d(input).
+         * 
+         * @param output       as an output
+         * @param input        as an input
+         * @param assign_core  not effective yet
+         */
+        void global_max_pool2d(Tensor<int16_t> &output,
+                               Tensor<int16_t> &input,
+                               const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief global_max_pool2d(input).
+         * 
+         * @param output       as an output
+         * @param input        as an input
+         * @param assign_core  not effective yet
+         */
+        void global_max_pool2d(Tensor<int8_t> &output,
+                               Tensor<int8_t> &input,
+                               const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief global_max_pool2d(input).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param input        as an input
+         * @param assign_core  not effective yet
+         * @return global_max_pool2d result
+         */
+        template <typename feature_t>
+        Tensor<feature_t> global_max_pool2d(Tensor<feature_t> &input,
+                                            const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+        {
+            DL_LOG_NN_LATENCY_INIT();
+
+            DL_LOG_NN_LATENCY_START();
+            std::vector<int> output_shape(input.shape.size(), 1);
+            output_shape[2] = input.shape[2];
+            Tensor<feature_t> output;
+            output.set_exponent(input.exponent).set_shape(output_shape).malloc_element();
+            DL_LOG_NN_LATENCY_END("apply");
+
+            DL_LOG_NN_LATENCY_START();
+            global_max_pool2d(output, input, assign_core);
+            DL_LOG_NN_LATENCY_END("global_max_pool2d");
+
+            return output;
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_leakyrelu.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_leakyrelu.hpp
new file mode 100644
index 00000000000..c41728b1ad5
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_leakyrelu.hpp
@@ -0,0 +1,82 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief leakyrelu(input).
+         * 
+         * @param output                as an output
+         * @param input                 as an input
+         * @param activation_alpha      quantized alpha
+         * @param activation_exponent   exponent of quantized alpha
+         * @param assign_core not effective yet
+         */
+        void leakyrelu(Tensor<int16_t> &output,
+                       Tensor<int16_t> &input,
+                       const int16_t activation_alpha,
+                       const int activation_exponent,
+                       const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief leakyrelu(input).
+         * 
+         * @param output                as an output
+         * @param input                 as an input
+         * @param activation_alpha      quantized alpha
+         * @param activation_exponent   exponent of quantized alpha
+         * @param assign_core not effective yet
+         */
+        void leakyrelu(Tensor<int8_t> &output,
+                       Tensor<int8_t> &input,
+                       const int8_t activation_alpha,
+                       const int activation_exponent,
+                       const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief leakyrelu(input)
+         * 
+         * @tparam inplace: whether directly store the output to input
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param input                 as an input
+         * @param activation_alpha      quantized alpha
+         * @param activation_exponent   exponent of quantized alpha
+         * @param assign_core           not effective yet
+         * @return leakyrelu result or no return(result store to input)
+         */
+        template <bool inplace = false, typename feature_t>
+        auto leakyrelu(Tensor<feature_t> &input,
+                       const int activation_alpha,
+                       const int activation_exponent,
+                       const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        {
+            DL_LOG_NN_LATENCY_INIT();
+            Tensor<feature_t> output;
+            if constexpr (!inplace)
+            {
+                DL_LOG_NN_LATENCY_START();
+                output.set_exponent(input.exponent).set_shape(input.shape).malloc_element();
+                DL_LOG_NN_LATENCY_END("apply");
+
+                DL_LOG_NN_LATENCY_START();
+                leakyrelu(output, input, activation_alpha, activation_exponent, assign_core);
+                DL_LOG_NN_LATENCY_END("leakyrelu");
+
+                return output;
+            }
+            else
+            {
+                DL_LOG_NN_LATENCY_START();
+                leakyrelu(input, input, activation_alpha, activation_exponent, assign_core);
+                DL_LOG_NN_LATENCY_END("leakyrelu");
+            }
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_max2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_max2d.hpp
new file mode 100644
index 00000000000..466089bc386
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_max2d.hpp
@@ -0,0 +1,81 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief max2d(input0, input1) 
+         * 
+         * @param output            as an output
+         * @param input0            as one input
+         * @param input1            as another input
+         * @param assign_core       not effective yet
+         */
+        void max2d(Tensor<int16_t> &output,
+                   Tensor<int16_t> &input0,
+                   Tensor<int16_t> &input1,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief max2d(input0, input1)  
+         * 
+         * @param output            as an output
+         * @param input0            as one input
+         * @param input1            as another input
+         * @param assign_core       not effective yet 
+         * @param output_exponent   exponent of output, only and must specify if inplace operation happens
+         */
+        void max2d(Tensor<int8_t> &output,
+                   Tensor<int8_t> &input0,
+                   Tensor<int8_t> &input1,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief max2d(input0, input1)
+         *
+         * @tparam inplace: whether directly store the output to input0 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param input0          as one input
+         * @param input1          as another input
+         * @param assign_core     not effective yet
+         * @return max2d result or no return(result store to input0)
+         */
+        template <bool inplace = false, typename feature_t>
+        auto max2d(Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        {
+            assert(input0.is_same_shape(input1));
+            assert(input0.exponent == input1.exponent);
+
+            DL_LOG_NN_LATENCY_INIT();
+            Tensor<feature_t> output;
+
+            if constexpr (!inplace)
+            {
+                DL_LOG_NN_LATENCY_START();
+                output.set_exponent(input0.exponent).set_shape(input0.shape).malloc_element();
+                DL_LOG_NN_LATENCY_END("apply");
+
+                DL_LOG_NN_LATENCY_START();
+                max2d(output, input0, input1, assign_core);
+                DL_LOG_NN_LATENCY_END("max2d");
+
+                return output;
+            }
+            else
+            {
+                DL_LOG_NN_LATENCY_START();
+                max2d(input0, input0, input1, assign_core);
+                DL_LOG_NN_LATENCY_END("max2d");
+            }
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_max_pool2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_max_pool2d.hpp
new file mode 100644
index 00000000000..50d51728ce0
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_max_pool2d.hpp
@@ -0,0 +1,101 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+#include <stdint.h>
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief max_pool2d(input).
+         * 
+         * @param output       as an output
+         * @param input        as an input
+         * @param padding      padding size needed in [top, bottom, left, right] of this operation
+         * @param filter_shape filter shape in [filter_height, filter_width]
+         * @param stride_y     stride in height
+         * @param stride_x     stride in width
+         * @param assign_core  not effective yet
+         */
+        void max_pool2d(Tensor<int16_t> &output,
+                        Tensor<int16_t> &input,
+                        std::vector<int> &padding,
+                        std::vector<int> &filter_shape,
+                        const int stride_y,
+                        const int stride_x,
+                        const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief max_pool2d(input).
+         * 
+         * @param output       as an output
+         * @param input        as an input
+         * @param padding      padding size needed in [top, bottom, left, right] of this operation
+         * @param filter_shape filter shape in [filter_height, filter_width]
+         * @param stride_y     stride in height
+         * @param stride_x     stride in width
+         * @param assign_core  not effective yet
+         */
+        void max_pool2d(Tensor<int8_t> &output,
+                        Tensor<int8_t> &input,
+                        std::vector<int> &padding,
+                        std::vector<int> &filter_shape,
+                        const int stride_y,
+                        const int stride_x,
+                        const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief max_pool2d(input).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param input        as an input
+         * @param filter_shape filter shape in [filter_height, filter_width]
+         * @param stride_y     stride in height
+         * @param stride_x     stride in width
+         * @param padding_type one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN,
+         *                     - PADDING_VALID: no padding
+         *                     PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
+         *                     such that output has the same height/width dimension as the input,
+         *                     - PADDING_SAME_END results padding in TensorFlow style
+         *                     - PADDING_SAME_BEGIN results padding in MXNET style
+         * @param assign_core  not effective yet
+         * @return max_pool2d result
+         */
+        template <typename feature_t>
+        Tensor<feature_t> max_pool2d(Tensor<feature_t> &input,
+                                     std::vector<int> filter_shape,
+                                     const int stride_y,
+                                     const int stride_x,
+                                     const padding_type_t padding_type,
+                                     const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+        {
+            DL_LOG_NN_LATENCY_INIT();
+
+            DL_LOG_NN_LATENCY_START();
+            std::vector<int> output_shape = get_output_shape(input.shape, filter_shape, stride_y, stride_x, padding_type);
+            Tensor<feature_t> output;
+            output.set_exponent(input.exponent).set_shape(output_shape).malloc_element();
+            DL_LOG_NN_LATENCY_END("apply");
+
+            std::vector<int> padding(4, 0);
+
+            DL_LOG_NN_LATENCY_START();
+            if (padding_type == PADDING_SAME_END || padding_type == PADDING_SAME_BEGIN)
+            {
+                padding = get_pad_size(output_shape, input.shape, filter_shape, stride_y, stride_x, padding_type);
+            }
+            DL_LOG_NN_LATENCY_END("padding");
+
+            DL_LOG_NN_LATENCY_START();
+            max_pool2d(output, input, padding, filter_shape, stride_y, stride_x, assign_core);
+            DL_LOG_NN_LATENCY_END("max_pool2d");
+
+            return output;
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_min2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_min2d.hpp
new file mode 100644
index 00000000000..8faddf3c228
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_min2d.hpp
@@ -0,0 +1,80 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief min2d(input0, input1) 
+         * 
+         * @param output      as an output
+         * @param input0      as one input
+         * @param input1      as another input
+         * @param assign_core 
+         */
+        void min2d(Tensor<int16_t> &output,
+                   Tensor<int16_t> &input0,
+                   Tensor<int16_t> &input1,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief min2d(input0, input1)  
+         * 
+         * @param output      as an output
+         * @param input0      as one input
+         * @param input1      as another input
+         * @param assign_core 
+         */
+        void min2d(Tensor<int8_t> &output,
+                   Tensor<int8_t> &input0,
+                   Tensor<int8_t> &input1,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief min2d(input0, input1)
+         * 
+         * @tparam inplace: whether directly store the output to input0
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param input0          as one input
+         * @param input1          as another input
+         * @param assign_core     not effective yet
+         * @return min2d result or no return(result store to input0)
+         */
+        template <bool inplace = false, typename feature_t>
+        auto min2d(Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        {
+            assert(input0.is_same_shape(input1));
+            assert(input0.exponent == input1.exponent);
+
+            DL_LOG_NN_LATENCY_INIT();
+            Tensor<feature_t> output;
+
+            if constexpr (!inplace)
+            {
+                DL_LOG_NN_LATENCY_START();
+                output.set_exponent(input0.exponent).set_shape(input0.shape).malloc_element();
+                DL_LOG_NN_LATENCY_END("apply");
+
+                DL_LOG_NN_LATENCY_START();
+                min2d(output, input0, input1, assign_core);
+                DL_LOG_NN_LATENCY_END("min2d");
+
+                return output;
+            }
+            else
+            {
+                DL_LOG_NN_LATENCY_START();
+                min2d(input0, input0, input1, assign_core);
+                DL_LOG_NN_LATENCY_END("min2d");
+            }
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_mul2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_mul2d.hpp
new file mode 100644
index 00000000000..909619a8767
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_mul2d.hpp
@@ -0,0 +1,91 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief activation(mul2d(input0, input1)).
+         * 
+         * @param output            as an output
+         * @param input0            as one input
+         * @param input1            as another input
+         * @param activation        activation of mul2d, if you don't specify anything, no activation is applied
+         * @param assign_core       not effective yet
+         * @param output_exponent   exponent of output, only and must specify if inplace operation happens
+         */
+        void mul2d(Tensor<int16_t> &output,
+                   Tensor<int16_t> &input0,
+                   Tensor<int16_t> &input1,
+                   const Activation<int16_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
+
+        /**
+         * @brief activation(mul2d(input0, input1)).
+         * 
+         * @param output            as an output
+         * @param input0            as one input
+         * @param input1            as another input
+         * @param activation        activation of mul2d, if you don't specify anything, no activation is applied
+         * @param assign_core       not effective yet
+         * @param output_exponent   exponent of output, only and must specify if inplace operation happens
+         */
+        void mul2d(Tensor<int8_t> &output,
+                   Tensor<int8_t> &input0,
+                   Tensor<int8_t> &input1,
+                   const Activation<int8_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
+
+        /**
+         * @brief activation(mul2d(input0, input1)).
+         * 
+         * @tparam inplace: whether directly store the output to input0
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param output_exponent exponent of output
+         * @param input0          as one input
+         * @param input1          as another input
+         * @param activation      activation of mul2d, if you don't specify anything, no activation is applied
+         * @param assign_core     not effective yet
+         * @return mul2d result or no return(result store to input0)
+         */
+        template <bool inplace = false, typename feature_t>
+        auto mul2d(const int output_exponent,
+                   Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const Activation<feature_t> *activation,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        {
+            assert(input0.is_same_shape(input1));
+
+            DL_LOG_NN_LATENCY_INIT();
+            Tensor<feature_t> output;
+
+            if constexpr (!inplace)
+            {
+                DL_LOG_NN_LATENCY_START();
+                output.set_exponent(output_exponent).set_shape(input0.shape).malloc_element();
+                DL_LOG_NN_LATENCY_END("apply");
+
+                DL_LOG_NN_LATENCY_START();
+                mul2d(output, input0, input1, activation, assign_core);
+                DL_LOG_NN_LATENCY_END("mul2d");
+
+                return output;
+            }
+            else
+            {
+                DL_LOG_NN_LATENCY_START();
+                mul2d(input0, input0, input1, activation, assign_core, output_exponent);
+                DL_LOG_NN_LATENCY_END("mul2d");
+            }
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_prelu.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_prelu.hpp
new file mode 100644
index 00000000000..e83e8975604
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_prelu.hpp
@@ -0,0 +1,82 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief prelu(input).
+         * 
+         * @param output                as an output
+         * @param input                 as an input
+         * @param activation_element    quantized alpha elements along channel axis
+         * @param activation_exponent   exponent of quantized alpha elements
+         * @param assign_core not effective yet
+         */
+        void prelu(Tensor<int16_t> &output,
+                   Tensor<int16_t> &input,
+                   const int16_t *activation_element,
+                   const int activation_exponent,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief prelu(input).
+         * 
+         * @param output                as an output
+         * @param input                 as an input
+         * @param activation_element    quantized alpha elements along channel axis
+         * @param activation_exponent   exponent of quantized alpha elements
+         * @param assign_core not effective yet
+         */
+        void prelu(Tensor<int8_t> &output,
+                   Tensor<int8_t> &input,
+                   const int8_t *activation_element,
+                   const int activation_exponent,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief prelu(input)
+         * 
+         * @tparam inplace: whether directly store the output to input
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param input                 as an input
+         * @param activation_element    quantized alpha elements along channel axis
+         * @param activation_exponent   exponent of quantized alpha elements
+         * @param assign_core           not effective yet
+         * @return prelu result or no return(result store to input)
+         */
+        template <bool inplace = false, typename feature_t>
+        auto prelu(Tensor<feature_t> &input,
+                   const feature_t *activation_element,
+                   const int activation_exponent,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        {
+            DL_LOG_NN_LATENCY_INIT();
+            Tensor<feature_t> output;
+            if constexpr (!inplace)
+            {
+                DL_LOG_NN_LATENCY_START();
+                output.set_exponent(input.exponent).set_shape(input.shape).malloc_element();
+                DL_LOG_NN_LATENCY_END("apply");
+
+                DL_LOG_NN_LATENCY_START();
+                prelu(output, input, activation_element, activation_exponent, assign_core);
+                DL_LOG_NN_LATENCY_END("prelu");
+
+                return output;
+            }
+            else
+            {
+                DL_LOG_NN_LATENCY_START();
+                prelu(input, input, activation_element, activation_exponent, assign_core);
+                DL_LOG_NN_LATENCY_END("prelu");
+            }
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_relu.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_relu.hpp
new file mode 100644
index 00000000000..308492dfe4b
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_relu.hpp
@@ -0,0 +1,70 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief relu(input).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param assign_core not effective yet
+         */
+        void relu(Tensor<int16_t> &output,
+                  Tensor<int16_t> &input,
+                  const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief relu(input).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param assign_core not effective yet
+         */
+        void relu(Tensor<int8_t> &output,
+                  Tensor<int8_t> &input,
+                  const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief relu(input)
+         * 
+         * @tparam inplace: whether directly store the output to input
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param input           as an input
+         * @param assign_core     not effective yet
+         * @return relu result or no return(result store to input)
+         */
+        template <bool inplace = false, typename feature_t>
+        auto relu(Tensor<feature_t> &input, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        {
+            DL_LOG_NN_LATENCY_INIT();
+            Tensor<feature_t> output;
+
+            if constexpr (!inplace)
+            {
+                DL_LOG_NN_LATENCY_START();
+                output.set_exponent(input.exponent).set_shape(input.shape).malloc_element();
+                DL_LOG_NN_LATENCY_END("apply");
+
+                DL_LOG_NN_LATENCY_START();
+                relu(output, input, assign_core);
+                DL_LOG_NN_LATENCY_END("relu");
+
+                return output;
+            }
+            else
+            {
+                DL_LOG_NN_LATENCY_START();
+                relu(input, input, assign_core);
+                DL_LOG_NN_LATENCY_END("relu");
+            }
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_sub2d.hpp b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_sub2d.hpp
new file mode 100644
index 00000000000..5bbd4940b10
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/nn/dl_nn_sub2d.hpp
@@ -0,0 +1,90 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief activation(sub2d(input0, input1)).
+         * 
+         * @param output      as an output
+         * @param input0      as one input
+         * @param input1      as another input
+         * @param activation  activation of sub2d, if you don't specify anything, no activation is applied
+         * @param assign_core not effective yet
+         * @param output_exponent   exponent of output, only and must specify if inplace operation happens 
+         */
+        void sub2d(Tensor<int16_t> &output,
+                   Tensor<int16_t> &input0,
+                   Tensor<int16_t> &input1,
+                   const Activation<int16_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
+
+        /**
+         * @brief activation(sub2d(input0, input1)).
+         * 
+         * @param output      as an output
+         * @param input0      as one input
+         * @param input1      as another input
+         * @param activation  activation of sub2d, if you don't specify anything, no activation is applied
+         * @param assign_core not effective yet
+         * @param output_exponent   exponent of output, only and must specify if inplace operation happens
+         */
+        void sub2d(Tensor<int8_t> &output,
+                   Tensor<int8_t> &input0,
+                   Tensor<int8_t> &input1,
+                   const Activation<int8_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
+
+        /**
+         * @brief activation(sub2d(input0, input1)).
+         * 
+         * @tparam inplace: whether directly store the output to input0
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param output_exponent exponent of output
+         * @param input0          as one input
+         * @param input1          as another input
+         * @param activation      activation of sub2d, if you don't specify anything, no activation is applied
+         * @param assign_core     not effective yet
+         * @return sub2d result or no return(result store to input0)
+         */
+        template <bool inplace = false, typename feature_t>
+        auto sub2d(const int output_exponent,
+                   Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const Activation<feature_t> *activation,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        {
+            assert(input0.is_same_shape(input1));
+
+            DL_LOG_NN_LATENCY_INIT();
+            Tensor<feature_t> output;
+            if constexpr (!inplace)
+            {
+                DL_LOG_NN_LATENCY_START();
+                output.set_exponent(output_exponent).set_shape(input0.shape).malloc_element();
+                DL_LOG_NN_LATENCY_END("apply");
+
+                DL_LOG_NN_LATENCY_START();
+                sub2d(output, input0, input1, activation, assign_core);
+                DL_LOG_NN_LATENCY_END("sub2d");
+
+                return output;
+            }
+            else
+            {
+                DL_LOG_NN_LATENCY_START();
+                sub2d(input0, input0, input1, activation, assign_core, output_exponent);
+                DL_LOG_NN_LATENCY_END("sub2d");
+            }
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/tool/dl_tool.hpp b/tools/sdk/esp32c3/include/esp-face/include/tool/dl_tool.hpp
new file mode 100644
index 00000000000..e5490e073d1
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/tool/dl_tool.hpp
@@ -0,0 +1,427 @@
+#pragma once
+
+#include <vector>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "esp_system.h"
+#include "esp_timer.h"
+#include "freertos/FreeRTOS.h"
+
+#include "dl_define.hpp"
+
+extern "C"
+{
+#if CONFIG_TIE728_BOOST
+    void dl_tie728_memset_8b(void *ptr, const int value, const int n);
+    void dl_tie728_memset_16b(void *ptr, const int value, const int n);
+    void dl_tie728_memset_32b(void *ptr, const int value, const int n);
+#endif
+}
+
+namespace dl
+{
+    namespace tool
+    {
+        /**
+         * @brief Set memory zero.
+         * 
+         * @param ptr pointer of memory
+         * @param n   byte number
+         */
+        void set_zero(void *ptr, const int n);
+
+        /**
+         * @brief Set array value.
+         * 
+         * @tparam T supports all data type, sizeof(T) equals to 1, 2 and 4 will boost by instruction 
+         * @param ptr   pointer of array
+         * @param value value to set
+         * @param len   length of array
+         */
+        template <typename T>
+        void set_value(T *ptr, const T value, const int len)
+        {
+#if CONFIG_TIE728_BOOST
+            int *temp = (int *)&value;
+            if (sizeof(T) == 1)
+                dl_tie728_memset_8b(ptr, *temp, len);
+            else if (sizeof(T) == 2)
+                dl_tie728_memset_16b(ptr, *temp, len);
+            else if (sizeof(T) == 4)
+                dl_tie728_memset_32b(ptr, *temp, len);
+            else
+#endif
+                for (size_t i = 0; i < len; i++)
+                    ptr[i] = value;
+        }
+
+        /**
+         * @brief Copy memory.
+         * 
+         * @param dst pointer of destination
+         * @param src pointer of source
+         * @param n   byte number
+         */
+        void copy_memory(void *dst, void *src, const int n);
+
+        /**
+         * @brief Apply memory without initialized. Can use free_aligned() to free the memory.
+         * 
+         * @param number number of elements
+         * @param size   size of element
+         * @param align  number of byte aligned, e.g., 16 means 16-byte aligned
+         * @return pointer of allocated memory. NULL for failed
+         */
+        inline void *malloc_aligned(int number, int size, int align = 4)
+        {
+            assert((align > 0) && (((align & (align-1)) == 0)));
+            int total_size = number * size;
+
+            void *res = heap_caps_aligned_alloc(align, total_size, MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL);
+#if DL_SPIRAM_SUPPORT
+            if (NULL == res)
+                res = heap_caps_aligned_alloc(align, total_size, MALLOC_CAP_SPIRAM);
+#endif
+            if (NULL == res)
+            {
+                printf("Fail to malloc %d bytes from DRAM(%d bytyes) and PSRAM(%d bytes), PSRAM is %s.\n",
+                       total_size,
+                       heap_caps_get_free_size(MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL),
+                       heap_caps_get_free_size(MALLOC_CAP_SPIRAM),
+                       DL_SPIRAM_SUPPORT ? "on" : "off");
+                return NULL;
+            }
+
+            return (void *)res;
+        }
+
+        /**
+         * @brief Apply memory with zero-initialized. Can use free_aligned() to free the memory.
+         * 
+         * @param number number of elements
+         * @param size   size of element
+         * @param align  number of byte aligned, e.g., 16 means 16-byte aligned
+         * @return pointer of allocated memory. NULL for failed
+         */
+        inline void *calloc_aligned(int number, int size, int align = 4)
+        {
+
+            void *aligned = malloc_aligned(number, size, align);
+            set_zero(aligned, number * size);
+
+            return (void *)aligned;
+        }
+
+        /**
+         * @brief Free the calloc_aligned() and malloc_aligned() memory
+         * 
+         * @param address pointer of memory to free
+         */
+        inline void free_aligned(void *address)
+        {
+            if (NULL == address)
+                return;
+
+            heap_caps_free(address);
+        }
+
+        /**
+         * @brief Apply memory without initialized in preference order: internal aligned, internal, external aligned
+         * 
+         * @param number number of elements
+         * @param size   size of element
+         * @param align  number of byte aligned, e.g., 16 means 16-byte aligned
+         * @return pointer of allocated memory. NULL for failed
+         */
+        inline void *malloc_aligned_prefer(int number, int size, int align = 4)
+        {
+            assert((align > 0) && (((align & (align-1)) == 0)));
+            int total_size = number * size;
+            void *res = heap_caps_aligned_alloc(align, total_size, MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL);
+            if (NULL == res){
+                res = heap_caps_malloc(total_size, MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL);
+            }
+#if DL_SPIRAM_SUPPORT
+            if (NULL == res){
+                res = heap_caps_aligned_alloc(align, total_size, MALLOC_CAP_SPIRAM);
+            }
+#endif
+            if (NULL == res)
+            {
+                printf("Fail to malloc %d bytes from DRAM(%d bytyes) and PSRAM(%d bytes), PSRAM is %s.\n",
+                       total_size,
+                       heap_caps_get_free_size(MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL),
+                       heap_caps_get_free_size(MALLOC_CAP_SPIRAM),
+                       DL_SPIRAM_SUPPORT ? "on" : "off");
+                return NULL;
+            }
+
+            return res;
+        }
+
+        /**
+         * @brief Apply memory with zero-initialized in preference order: internal aligned, internal, external aligned
+         * 
+         * @param number number of elements
+         * @param size   size of element
+         * @param align  number of byte aligned, e.g., 16 means 16-byte aligned
+         * @return pointer of allocated memory. NULL for failed
+         */
+        inline void *calloc_aligned_prefer(int number, int size, int align = 4)
+        {
+            void *res = malloc_aligned_prefer(number, size, align);
+            set_zero(res, number * size);
+
+            return (void *)res;
+        }
+
+        /**
+         * @brief Free the calloc_aligned_prefer() and malloc_aligned_prefer() memory
+         * 
+         * @param address pointer of memory to free
+         */
+        inline void free_aligned_prefer(void *address)
+        {
+            if (NULL == address)
+                return;
+
+            heap_caps_free(address);
+        }
+
+        /**
+         * @brief Truncate the input into int8_t range.
+         * 
+         * @tparam T supports all integer types
+         * @param output as an output
+         * @param input  as an input
+         */
+        template <typename T>
+        void truncate(int8_t &output, T input)
+        {
+            if (input >= DL_Q8_MAX)
+                output = DL_Q8_MAX;
+            else if (input <= DL_Q8_MIN)
+                output = DL_Q8_MIN;
+            else
+                output = input;
+        }
+
+        /**
+         * @brief Truncate the input into int16_t range.
+         * 
+         * @tparam T supports all integer types
+         * @param output as an output
+         * @param input  as an input
+         */
+        template <typename T>
+        void truncate(int16_t &output, T input)
+        {
+            if (input >= DL_Q16_MAX)
+                output = DL_Q16_MAX;
+            else if (input <= DL_Q16_MIN)
+                output = DL_Q16_MIN;
+            else
+                output = input;
+        }
+
+        /**
+         * @brief Calculate the exponent of quantizing 1/n into max_value range.
+         * 
+         * @param n          1/n: value to be quantized
+         * @param max_value  the max_range
+         */
+        inline int calculate_exponent(int n, int max_value)
+        {
+            int exp = 0;
+            int tmp = 1 / n;
+            while (tmp < max_value)
+            {
+                exp += 1;
+                tmp = (1 << exp) / n;
+            }
+            exp -= 1;
+
+            return exp;
+        }
+
+        /**
+         * @brief Print vector in format "[x1, x2, ...]\n".
+         * 
+         * @param array to print
+         */
+        inline void print_vector(std::vector<int> &array, const char *message = NULL)
+        {
+            if (message)
+                printf("%s: ", message);
+
+            printf("[");
+            for (int i = 0; i < array.size(); i++)
+            {
+                printf(", %d" + (i ? 0 : 2), array[i]);
+            }
+            printf("]\n");
+        }
+
+        /**
+         * @brief Get the cycle object
+         * 
+         * @return cycle count
+         */
+        inline uint32_t get_cycle()
+        {
+            uint32_t ccount;
+            __asm__ __volatile__("rsr %0, ccount"
+                                 : "=a"(ccount)
+                                 :
+                                 : "memory");
+            return ccount;
+        }
+
+        class Latency
+        {
+        private:
+            const uint32_t size; /*<! size of queue */
+            uint32_t *queue;     /*<! queue for storing history period */
+            uint32_t period;     /*<! current period */
+            uint32_t sum;        /*<! sum of period */
+            uint32_t count;      /*<! the number of added period */
+            uint32_t next;       /*<! point to next element in queue */
+            uint32_t timestamp;  /*<! record the start >*/
+
+        public:
+            /**
+             * @brief Construct a new Latency object.
+             * 
+             * @param size 
+             */
+            Latency(const uint32_t size = 1) : size(size),
+                                               period(0),
+                                               sum(0),
+                                               count(0),
+                                               next(0)
+            {
+                this->queue = (this->size > 1) ? (uint32_t *)calloc(this->size, sizeof(uint32_t)) : NULL;
+            }
+
+            /**
+             * @brief Destroy the Latency object.
+             * 
+             */
+            ~Latency()
+            {
+                if (this->queue)
+                    free(this->queue);
+            }
+
+            /**
+             * @brief Record the start timestamp.
+             * 
+             */
+            void start()
+            {
+#if DL_LOG_LATENCY_UNIT
+                this->timestamp = get_cycle();
+#else
+                this->timestamp = esp_timer_get_time();
+#endif
+            }
+
+            /**
+             * @brief Record the period.
+             * 
+             */
+            void end()
+            {
+#if DL_LOG_LATENCY_UNIT
+                this->period = get_cycle() - this->timestamp;
+#else
+                this->period = esp_timer_get_time() - this->timestamp;
+#endif
+                if (this->queue)
+                {
+                    this->sum -= this->queue[this->next];
+                    this->queue[this->next] = this->period;
+                    this->sum += this->queue[this->next];
+                    this->next++;
+                    this->next = this->next % this->size;
+                    if (this->count < this->size)
+                    {
+                        this->count++;
+                    }
+                }
+            }
+
+            /**
+             * @brief Return the period.
+             * 
+             * @return this->timestamp_end - this->timestamp
+             */
+            uint32_t get_period()
+            {
+                return this->period;
+            }
+
+            /**
+             * @brief Get the average period.
+             * 
+             * @return average latency 
+             */
+            uint32_t get_average_period()
+            {
+                return this->queue ? (this->sum / this->count) : this->period;
+            }
+
+            /**
+             * @brief Clear the period
+             * 
+             */
+            void clear_period()
+            {
+                this->period = 0;
+            }
+
+            /**
+             * @brief Print in format "latency: {this->period} {unit}\n".
+             */
+            void print()
+            {
+#if DL_LOG_LATENCY_UNIT
+                printf("latency: %15u cycle\n", this->get_average_period());
+#else
+                printf("latency: %15u us\n", this->get_average_period());
+#endif
+            }
+
+            /**
+             * @brief Print in format "{message}: {this->period} {unit}\n".
+             * 
+             * @param message message of print
+             */
+            void print(const char *message)
+            {
+#if DL_LOG_LATENCY_UNIT
+                printf("%s: %15u cycle\n", message, this->get_average_period());
+#else
+                printf("%s: %15u us\n", message, this->get_average_period());
+#endif
+            }
+
+            /**
+             * @brief Print in format "{prefix}::{key}: {this->period} {unit}\n".
+             * 
+             * @param prefix prefix of print
+             * @param key    key of print
+             */
+            void print(const char *prefix, const char *key)
+            {
+#if DL_LOG_LATENCY_UNIT
+                printf("%s::%s: %u cycle\n", prefix, key, this->get_average_period());
+#else
+                printf("%s::%s: %u us\n", prefix, key, this->get_average_period());
+#endif
+            }
+        };
+    } // namespace tool
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/tool/dl_tool_cache.hpp b/tools/sdk/esp32c3/include/esp-face/include/tool/dl_tool_cache.hpp
new file mode 100644
index 00000000000..74c81717a58
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/tool/dl_tool_cache.hpp
@@ -0,0 +1,74 @@
+#pragma once
+
+#include <stdint.h>
+
+#if CONFIG_IDF_TARGET_ESP32S3
+#include "esp32s3/rom/cache.h"
+#include "soc/extmem_reg.h"
+#endif
+
+namespace dl
+{
+    namespace tool
+    {
+        namespace cache
+        {
+            /**
+             * @brief Initialize preload.
+             * 
+             * @param preload One of 1 or 0,
+             *                - 1: turn on the preload
+             *                - 0: turn off the preload
+             * @return 
+             *         - 1: Initialize successfully
+             *         - 0: Initialize successfully, autoload has been turned off
+             *         - -1: Initialize failed, the chip does not support preload
+             */
+            int8_t preload_init(uint8_t preload = 1);
+
+            /**
+             * @brief Preload memory.
+             * 
+             * @param addr the start address of data to be preloaded
+             * @param size the size of the data in byte to be preloaded
+             */
+            void preload_func(uint32_t addr, uint32_t size);
+
+            /**
+             * @brief Initialize autoload. 
+             * 
+             * @param autoload  One of 1 or 0,
+             *                  - 1: turn on the autoload
+             *                  - 0: turn off the autoload
+             * @param trigger   One of 0 or 1 or 2,
+             *                  - 0: miss, TODO:@yuanjiong
+             *                  - 1: hit, TODO:@yuanjiong
+             *                  - 2: both,TODO:@yuanjiong
+             * @param line_size the number of cache lines to be autoloaded
+             * @return status,
+             *         - 1: Initialize sucessfully
+             *         - 0: Initialize suceesfully, preload has been turned off
+             *         - -1: Initialize failed, the chip does not support autoload
+             */
+            int8_t autoload_init(uint8_t autoload = 1, uint8_t trigger = 2, uint8_t line_size = 0);
+
+            /**
+             * @brief Autoload memory.           
+             * 
+             * @param addr1 the start address of data1 to be autoloaded
+             * @param size1 the size of the data1 in byte to be preloaded
+             * @param addr2 the start address of data2 to be autoloaded
+             * @param size2 the size of the data2 in byte to be preloaded
+             */
+            void autoload_func(uint32_t addr1, uint32_t size1, uint32_t addr2, uint32_t size2);
+
+            /**
+             * @brief Autoload memory.
+             * 
+             * @param addr1 the start address of data1 to be autoloaded
+             * @param size1 the size of the data1 in byte to be preloaded
+             */
+            void autoload_func(uint32_t addr1, uint32_t size1);
+        }
+    } // namespace tool
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/typedef/dl_constant.hpp b/tools/sdk/esp32c3/include/esp-face/include/typedef/dl_constant.hpp
new file mode 100644
index 00000000000..07b2dd24ee1
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/typedef/dl_constant.hpp
@@ -0,0 +1,129 @@
+#pragma once
+
+#include "dl_define.hpp"
+#include <vector>
+
+namespace dl
+{
+    /**
+     * @brief Base class of Filter, Bias, Activation.
+     * 
+     * @tparam T supports int16_t and int8_t,
+     *         - int16_t: stands for operation in int16_t quantize,
+     *         - int8_t: stands for operation in int8_t quantize.
+     */
+    template <typename T>
+    class Constant
+    {
+    public:
+        const T *element;             /*<! point to element. >*/
+        const int exponent;           /*<! exponent of element. >*/
+        const std::vector<int> shape; /*<! shape of element. >*/
+
+        /**
+         * @brief Construct a new Constant object.
+         * 
+         * @param element  point to element.
+         * @param exponent exponent of element.
+         * @param shape    shape of Constant.
+         */
+        Constant(const T *element, const int exponent, const std::vector<int> shape);
+    };
+
+    /**
+     * @brief Filter.
+     * NOTE: The shape format of filter is fixed, but the element sequence depands on optimization method.
+     *       - 1D: reserved
+     *       - 2D: shape format is [filter_height, filter_width, input_channel, output_channel]. dilation format is [height, width]
+     *  
+     * @tparam T supports int16_t and int8_t,
+     *         - int16_t: stands for operation in int16_t quantize,
+     *         - int8_t: stands for operation in int8_t quantize.
+     */
+    template <typename T>
+    class Filter : public Constant<T>
+    {
+    public:
+        const std::vector<int> dilation;      /*<! - 1D: reserved >*/
+                                              /*<! - 2D: [dilation_in_height, dilation_in_width] >*/
+        std::vector<int> shape_with_dilation; /*<! - 1D: reserved >*/
+                                              /*<! - 2D: [filter_height_with_dilation, filter_width_with_dilation, input_channel, output_channel] >*/
+        std::vector<int> channel_exponent;    /*<! exponent for per-channel >*/
+
+        /**
+         * @brief Construct a new Filter object.
+         * 
+         * @param element  point to element
+         * @param exponent exponent of element
+         * @param shape    shape of Filter,
+         *                 - 1D: reserved
+         *                 - 2D: for convolution is [filter_height, filter_width, input_channel, output_channel],
+         *                       for depthwise convolution is [filter_height, filter_width, input_channel, 1]
+         * @param dilation dilation of Filter
+         *                 - 1D: reserved
+         *                 - 2D: [dilation_in_height, dilation_in_width]
+         */
+        Filter(const T *element, const int exponent, const std::vector<int> shape, const std::vector<int> dilation = {1, 1});
+
+        /**
+         * @brief Construct a new Filter object.
+         * 
+         * @param element          point to element
+         * @param channel_exponent exponent for per-channel
+         * @param shape            shape of element
+         * @param dilation         dilation of Filter
+         *                         - 1D: reserved
+         *                         - 2D: [dilation_in_height, dilation_in_width]
+         */
+        Filter(const T *element, const std::vector<int> channel_exponent, const std::vector<int> shape, const std::vector<int> dilation = {1, 1});
+
+        /**
+         * @brief Print the n-th filter.
+         * 
+         * @param n       index of output_channel
+         * @param message to print
+         */
+        void print2d_n(const int n, const char *message) const;
+    };
+
+    /**
+     * @brief Bias.
+     * 
+     * @tparam T supports int16_t and int8_t
+     *         - int16_t: stands for operation in int16_t quantize
+     *         - int8_t: stands for operation in int8_t quantize
+     */
+    template <typename T>
+    class Bias : public Constant<T>
+    {
+    public:
+        using Constant<T>::Constant;
+        std::vector<int> channel_exponent;    /*<! exponent for per-channel >*/
+
+        Bias(const T *element, const std::vector<int> channel_exponent, const std::vector<int> shape);
+    };
+
+    /**
+     * @brief Activation.
+     * 
+     * @tparam T supports int16_t and int8_t
+     *         - int16_t: stands for operation in int16_t quantize
+     *         - int8_t: stands for operation in int8_t quantize
+     */
+    template <typename T>
+    class Activation : public Constant<T>
+    {
+    public:
+        const activation_type_t type; /*<! One of Linear or ReLU or LeakyReLU or PReLU */
+
+        /**
+         * @brief Construct a new Activation object.
+         * 
+         * @param type      One of Linear or ReLU or LeakyReLU or PReLU
+         * @param element   point to element of activation
+         * @param exponent  exponent of element
+         * @param shape     shape of element
+         */
+        Activation(const activation_type_t type, const T *element = NULL, const int exponent = 0, const std::vector<int> shape = {0});
+    };
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp-face/include/typedef/dl_variable.hpp b/tools/sdk/esp32c3/include/esp-face/include/typedef/dl_variable.hpp
new file mode 100644
index 00000000000..18cbe9707e9
--- /dev/null
+++ b/tools/sdk/esp32c3/include/esp-face/include/typedef/dl_variable.hpp
@@ -0,0 +1,495 @@
+#pragma once
+
+#include <stdio.h>
+#include <vector>
+#include <assert.h>
+#include <iostream>
+
+#include "dl_tool.hpp"
+
+namespace dl
+{
+    /**
+     * @brief Tensor
+     * 
+     * @tparam T support uint8_t, int8_t, int16_t and float.
+     */
+    template <typename T>
+    class Tensor
+    {
+    private:
+        int size;                     /*<! size of element including padding */
+        bool auto_free;               /*<! free element when object destroy */
+        std::vector<int> axis_offset; /*<! element offset of each axis */
+
+    public:
+        T *element;             /*<! point to element */
+        int exponent;           /*<! exponent of element */
+        std::vector<int> shape; /*<! shape of Tensor */
+
+        /**
+         * @brief Construct a new Tensor object
+         * 
+         */
+        Tensor() : auto_free(true), element(NULL), exponent(0) { this->set_shape({0}); }
+
+        /**
+         * @brief Construct a new Tensor object by copying from input.
+         * 
+         * @param input an input Tensor
+         * @param deep  one of true or false
+         *              - true: apply a new memory, copy value from input.element to this new memory
+         *              - false: take over input.element to this->element
+         */
+        Tensor(Tensor<T> &input, bool deep) : size(input.size),
+                                              auto_free(input.auto_free),
+                                              exponent(input.exponent)
+        {
+            this->set_shape(input.shape);
+            if (deep && (input.element != NULL))
+            {
+                int size_real = input.get_size();
+                T *new_element = (T *)tool::calloc_aligned_prefer(size_real, sizeof(T), 16);
+                tool::copy_memory(new_element, input.element, size_real * sizeof(T));
+                this->element = new_element;
+            }
+            else
+            {
+                this->element = input.element;
+                this->auto_free = false;
+            }
+        }
+
+        /**
+         * @brief Destroy the Tensor object
+         * 
+         */
+        ~Tensor()
+        {
+            if (this->auto_free)
+                this->free_element();
+        }
+
+        /**
+         * @brief 
+         * 
+         * @param input an input Tensor
+         * @param deep one of true or false
+         *              - true: apply a new memory, copy value from input.element to this new memory
+         *              - false: take over input.element to this->element
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &copy_element(Tensor<T> &input, bool deep)
+        {
+            assert(this->get_size() == input.get_size());
+            assert(input.element != NULL);
+
+            this->malloc_element();
+            if (deep)
+            {
+                tool::copy_memory(this->element, input.element, this->get_size() * sizeof(T));
+            }
+            else
+            {
+                this->element = input.element;
+                this->auto_free = false;
+            }
+            return *this;
+        }
+
+        /**
+         * @brief Set the auto free object.
+         * 
+         * @param auto_free one of true or false
+         *                  - true: free element when object destroyed
+         *                  - false: do not
+         * @return self
+         */
+        Tensor<T> &set_auto_free(const bool auto_free)
+        {
+            this->auto_free = auto_free;
+            return *this;
+        }
+
+        /**
+         * @brief Set the element.
+         * 
+         * @param element point to element memory
+         * @return self
+         */
+        Tensor<T> &set_element(T *element, const bool auto_free = false)
+        {
+            assert(this->element == NULL);
+            this->element = element;
+            this->auto_free = auto_free;
+
+            return *this;
+        }
+
+        /**
+         * @brief Set the exponent.
+         * 
+         * @param exponent exponent of element
+         * @return self
+         */
+        Tensor<T> &set_exponent(const int exponent)
+        {
+            this->exponent = exponent;
+
+            return *this;
+        }
+
+        /**
+         * @brief Set the shape of Tensor.
+         * 
+         * @param shape the target shape 
+         *        
+         * @return self
+         */
+        Tensor<T> &set_shape(const std::vector<int> shape);
+
+        /**
+         * @brief print the shape of the Tensor
+         * 
+         */
+        void print_shape()
+        {
+            if (this->shape.size())
+            {
+                printf("shape = (");
+                for (int i = 0; i < this->shape.size() - 1; i++)
+                {
+                    printf("%d, ", this->shape[i]);
+                }
+                printf("%d)\n", this->shape.back());
+            }
+            else
+            {
+                printf("shape = ()\n");
+            }
+        }
+
+        /**
+         * @brief flatten the Tensor
+         * 
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &flatten();
+
+        /**
+         * @brief Change a new shape to the Tensor without changing its data.
+         * 
+         * @param shape  the target shape
+         * @return Tensor<T>&  self
+         */
+        Tensor<T> &reshape(std::vector<int> shape);
+
+        /**
+         * @brief Remove dims with length==1 from Tensor
+         * 
+         * @param axis the dim to to be remove. make sure the length of the dim is equal to 1.
+         *              if axis == INT32_MAX, all the dims with length==1 will be removed.
+         * @return Tensor<T>& self 
+         */
+        Tensor<T> &squeeze(int axis = INT32_MAX);
+
+        /**
+         * @brief Insert a new dim that will appear at the axis position in the expanded Tensor shape.
+         * 
+         * @param axis the dim to be inserted
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &expand_dims(int axis);
+
+        /**
+         * @brief Insert a new dim that will appear at the axis position in the expanded Tensor shape.
+         * 
+         * @param axis  the dim to be inserted
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &expand_dims(std::vector<int> axis);
+
+        /**
+         * @brief Reverse or permute the axes of the Tensor
+         * 
+         * @param perm the new arangement of the dims. if perm == {}, the dims arangement will be reversed. 
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &transpose(std::vector<int> perm = {});
+
+        /**
+         * @brief Reverse or permute the axes of the input Tensor
+         * 
+         * @param input the input Tensor
+         * @param perm the new arangement of the dims. if perm == {}, the dims arangement will be reversed. 
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &transpose(Tensor<T> &input, std::vector<int> perm = {});
+
+        /**
+         * @brief Get the element pointer.
+         * 
+         * @return pointer to memory
+         */
+        T *get_element_ptr()
+        {
+            return this->element;
+        }
+
+        /**
+         * @brief Get the element value.
+         * 
+         * @param index   the index of each dim. 
+         * @return T element value
+         */
+        T get_element_value(const std::vector<int> index)
+        {
+            return this->element[this->get_element_index(index)];
+        }
+
+        /**
+         * @brief Get the element value.
+         * 
+         * @param index  the index of the element.
+         * @return T  element value
+         */
+        T get_element_value(int index)
+        {
+            return this->element[index];
+        }
+
+        /**
+         * @brief Get the size of Tensor.
+         * 
+         * @return  the size of Tensor.
+         */
+        int get_size()
+        {
+            return this->size;
+        }
+
+        /**
+         * @brief Get the axis offset
+         * 
+         * @return std::vector<int> the axis offset
+         */
+        std::vector<int> get_axis_offset()
+        {
+            return this->axis_offset;
+        }
+
+        /**
+         * @brief Apply memory with zero-initialized only if this->element is NULL.
+         * 
+         * @param auto_free one of true or false
+         *                  - true: free element when object destroyed
+         *                  - false: do not
+         * @return 
+         *         - true: on success
+         *         - false: if applying failed
+         */
+        bool calloc_element(const bool auto_free = true)
+        {
+            if (this->element != NULL)
+                return false;
+
+            this->element = (T *)dl::tool::calloc_aligned_prefer(this->get_size(), sizeof(T), 16);
+            this->auto_free = auto_free;
+
+            return true;
+        }
+
+        /**
+         * @brief Apply memory without initialized only if this->element is NULL.
+         * 
+         * @param auto_free one of true or false
+         *                  - true: free element when object destroyed
+         *                  - false: do not
+         * @return 
+         *         - true: on success
+         *         - false: if applying failed
+         */
+        bool malloc_element(const bool auto_free = true)
+        {
+            if (this->element != NULL)
+                return false;
+
+            this->element = (T *)tool::malloc_aligned_prefer(this->get_size(), sizeof(T), 16);
+            this->auto_free = auto_free;
+
+            return true;
+        }
+
+        /**
+         * @brief free element only if this->element != NULL
+         * set this->element to NULL, after free
+         * @brief Free element if this->element is not NULL.
+         */
+        void free_element()
+        {
+            if (this->auto_free && this->element)
+            {
+                tool::free_aligned_prefer(this->element);
+                this->element = NULL;
+            }
+        }
+
+        /**
+         * @brief print the element of the tensor
+         * 
+         * @param axis_index_range  the element range of each dims to be print. if axis_index_range == {}, all the element will be print. 
+         * @param message  to print  
+         */
+        void print(std::vector<int> axis_index_range = {}, const char *message = "");
+
+        /**
+         * @brief  print all the element of the Tensor.
+         * 
+         * @param message to print  
+         * @param with_padding one of true or false,
+         *                     - true: the padding element will also be ed
+         *                     - false: the padding element will not be ed
+         */
+        void print_all(const char *message = "")
+        {
+            std::cout << "\n"
+                      << message << " | ";
+            this->print_shape();
+
+            for (int i = 0; i < this->get_size(); i++)
+            {
+                std::cout << this->element[i] << " ";
+            }
+            std::cout << "\n";
+            return;
+        }
+
+        /**
+         * @brief Get the index of each dims
+         * 
+         * @param element_index the index of the element
+         * @return std::vector<int> the index of each dims
+         */
+        std::vector<int> get_axis_index(int element_index);
+
+        /**
+         * @brief Get the index of element 
+         * 
+         * @param axis_index the index of each dims
+         * @return int the index of element 
+         */
+        int get_element_index(const std::vector<int> axis_index);
+
+        /**
+         * @brief Check the element value with input ground-truth.
+         * 
+         * @param gt_element ground-truth value of element
+         * @param bias permissible error
+         * @param info one of true or false
+         *             - true:  shape and result
+         *             - false: do not
+         * @param failed_number maximum number of wrong element that will be printed
+         * 
+         * @return 
+         *         - true: in permissible error
+         *         - false: not 
+         */
+        bool check_element(T *gt_element, int bias = 2, bool info = true, int failed_number = 0)
+        {
+            int count = 0;
+            if (info)
+                this->print_shape();
+            int size = this->get_size();
+            for (int i = 0; i < size; i++)
+            {
+                if (DL_ABS(this->element[i] - gt_element[i]) > bias)
+                {
+                    std::vector<int> index = get_axis_index(i);
+                    std::cout << "element[";
+                    for (int j = 0; j < index.size() - 1; j++)
+                    {
+                        std::cout << index[j] << ", ";
+                    }
+                    std::cout << index.back() << "]: ";
+                    std::cout << +this->element[i] << " v.s. " << +gt_element[i] << "\n";
+                    count++;
+                    if (count > failed_number)
+                        return false;
+                }
+            }
+            if (count)
+                return false;
+
+            if (info)
+                printf("PASS\n");
+
+            return true;
+        }
+
+        /**
+         * @brief Check the shape is the same as the shape of input.
+         * 
+         * @param input an input tensor 
+         * @return 
+         *         - true: same shape 
+         *         - false: not 
+         */
+        bool is_same_shape(Tensor<T> &input)
+        {
+            if (input.shape.size() != this->shape.size())
+            {
+                return false;
+            }
+            for (int i = 0; i < this->shape.size(); i++)
+            {
+                if (input.shape[i] != this->shape[i])
+                {
+                    return false;
+                }
+            }
+            return true;
+        }
+
+        Tensor<T> &operator=(const Tensor<T> &input)
+        {
+            this->auto_free = input.auto_free;
+            this->exponent = input.exponent;
+            int size_real_tmp = this->size;
+            int size_input_real = input.size;
+            this->set_shape(input.shape);
+            if (input.element)
+            {
+                if (this->element)
+                {
+                    if (size_real_tmp != size_input_real)
+                    {
+                        tool::free_aligned_prefer(this->element);
+                        T *new_element = (T *)tool::malloc_aligned_prefer(size_input_real, sizeof(T), 16);
+                        tool::copy_memory(new_element, input.element, size_input_real * sizeof(T));
+                        this->element = new_element;
+                    }
+                    else
+                    {
+                        tool::copy_memory(this->element, input.element, size_input_real * sizeof(T));
+                    }
+                }
+                else
+                {
+                    T *new_element = (T *)tool::malloc_aligned_prefer(size_input_real, sizeof(T), 16);
+                    tool::copy_memory(new_element, input.element, size_input_real * sizeof(T));
+                    this->element = new_element;
+                }
+                return *this;
+            }
+            else
+            {
+                if (this->element)
+                {
+                    tool::free_aligned_prefer(this->element);
+                    this->element = NULL;
+                }
+                return *this;
+            }
+        }
+    };
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32c3/include/esp_https_server/include/esp_https_server.h b/tools/sdk/esp32c3/include/esp_https_server/include/esp_https_server.h
index 00565a963ad..efe726e9e70 100644
--- a/tools/sdk/esp32c3/include/esp_https_server/include/esp_https_server.h
+++ b/tools/sdk/esp32c3/include/esp_https_server/include/esp_https_server.h
@@ -10,6 +10,7 @@
 #include <stdbool.h>
 #include "esp_err.h"
 #include "esp_http_server.h"
+#include "esp_tls.h"
 
 #ifdef __cplusplus
 extern "C" {
@@ -20,6 +21,22 @@ typedef enum {
     HTTPD_SSL_TRANSPORT_INSECURE     // SSL disabled
 } httpd_ssl_transport_mode_t;
 
+/**
+ * @brief Callback data struct, contains the ESP-TLS connection handle
+ */
+typedef struct esp_https_server_user_cb_arg {
+    const esp_tls_t *tls;
+} esp_https_server_user_cb_arg_t;
+
+/**
+ * @brief Callback function prototype
+ * Can be used to get connection or client information (SSL context)
+ * E.g. Client certificate, Socket FD, Connection state, etc.
+ *
+ * @param user_cb Callback data struct
+ */
+typedef void esp_https_server_user_cb(esp_https_server_user_cb_arg_t *user_cb);
+
 /**
  * HTTPS server config struct
  *
@@ -66,6 +83,9 @@ struct httpd_ssl_config {
 
     /** Enable tls session tickets */
     bool session_tickets;
+
+    /** User callback for esp_https_server */
+    esp_https_server_user_cb *user_cb;
 };
 
 typedef struct httpd_ssl_config httpd_ssl_config_t;
@@ -113,6 +133,7 @@ typedef struct httpd_ssl_config httpd_ssl_config_t;
     .port_secure = 443,                           \
     .port_insecure = 80,                          \
     .session_tickets = false,                     \
+    .user_cb = NULL,                              \
 }
 
 /**
diff --git a/tools/sdk/esp32c3/include/esp_hw_support/include/esp_sleep.h b/tools/sdk/esp32c3/include/esp_hw_support/include/esp_sleep.h
index cfdfbc04186..75d2f9726c3 100644
--- a/tools/sdk/esp32c3/include/esp_hw_support/include/esp_sleep.h
+++ b/tools/sdk/esp32c3/include/esp_hw_support/include/esp_sleep.h
@@ -44,6 +44,7 @@ typedef enum {
 #if SOC_PM_SUPPORT_CPU_PD
     ESP_PD_DOMAIN_CPU,             //!< CPU core
 #endif
+    ESP_PD_DOMAIN_RTC8M,           //!< Internal 8M oscillator
     ESP_PD_DOMAIN_VDDSDIO,         //!< VDD_SDIO
     ESP_PD_DOMAIN_MAX              //!< Number of domains
 } esp_sleep_pd_domain_t;
diff --git a/tools/sdk/esp32c3/include/esp_lcd/include/esp_lcd_panel_io.h b/tools/sdk/esp32c3/include/esp_lcd/include/esp_lcd_panel_io.h
index eebcabf42b4..330f4d9a165 100644
--- a/tools/sdk/esp32c3/include/esp_lcd/include/esp_lcd_panel_io.h
+++ b/tools/sdk/esp32c3/include/esp_lcd/include/esp_lcd_panel_io.h
@@ -65,6 +65,22 @@ esp_err_t esp_lcd_panel_io_tx_color(esp_lcd_panel_io_handle_t io, int lcd_cmd, c
  */
 esp_err_t esp_lcd_panel_io_del(esp_lcd_panel_io_handle_t io);
 
+/**
+ * @brief Type of LCD panel IO event data
+ */
+typedef struct {
+} esp_lcd_panel_io_event_data_t;
+
+/**
+ * @brief Declare the prototype of the function that will be invoked when panel IO finishes transferring color data
+ *
+ * @param[in] panel_io LCD panel IO handle, which is created by factory API like `esp_lcd_new_panel_io_spi()`
+ * @param[in] edata Panel IO event data, fed by driver
+ * @param[in] user_ctx User data, passed from `esp_lcd_panel_io_xxx_config_t`
+ * @return Whether a high priority task has been waken up by this function
+ */
+typedef bool (*esp_lcd_panel_io_color_trans_done_cb_t)(esp_lcd_panel_io_handle_t panel_io, esp_lcd_panel_io_event_data_t *edata, void *user_ctx);
+
 /**
  * @brief Panel IO configuration structure, for SPI interface
  */
@@ -74,8 +90,8 @@ typedef struct {
     int spi_mode;    /*!< Traditional SPI mode (0~3) */
     unsigned int pclk_hz;    /*!< Frequency of pixel clock */
     size_t trans_queue_depth; /*!< Size of internal transaction queue */
-    bool (*on_color_trans_done)(esp_lcd_panel_io_handle_t panel_io, void *user_data, void *event_data); /*!< Callback, invoked when color data transfer has finished */
-    void *user_data;    /*!< User private data, passed directly to on_trans_frame_done's user_data */
+    esp_lcd_panel_io_color_trans_done_cb_t on_color_trans_done; /*!< Callback invoked when color data transfer has finished */
+    void *user_ctx;    /*!< User private data, passed directly to on_color_trans_done's user_ctx */
     int lcd_cmd_bits;   /*!< Bit-width of LCD command */
     int lcd_param_bits; /*!< Bit-width of LCD parameter */
     struct {
@@ -100,8 +116,8 @@ esp_err_t esp_lcd_new_panel_io_spi(esp_lcd_spi_bus_handle_t bus, const esp_lcd_p
 
 typedef struct {
     uint32_t dev_addr; /*!< I2C device address */
-    bool (*on_color_trans_done)(esp_lcd_panel_io_handle_t panel_io, void *user_data, void *event_data); /*!< Callback, invoked when color data transfer has finished */
-    void *user_data; /*!< User private data, passed directly to on_trans_frame_done's user_data */
+    esp_lcd_panel_io_color_trans_done_cb_t on_color_trans_done; /*!< Callback invoked when color data transfer has finished */
+    void *user_ctx; /*!< User private data, passed directly to on_color_trans_done's user_ctx */
     size_t control_phase_bytes; /*!< I2C LCD panel will encode control information (e.g. D/C seclection) into control phase, in several bytes */
     unsigned int dc_bit_offset; /*!< Offset of the D/C selection bit in control phase */
     int lcd_cmd_bits;           /*!< Bit-width of LCD command */
@@ -168,8 +184,8 @@ typedef struct {
     int cs_gpio_num;         /*!< GPIO used for CS line, set to -1 will declaim exclusively use of I80 bus */
     unsigned int pclk_hz;    /*!< Frequency of pixel clock */
     size_t trans_queue_depth; /*!< Transaction queue size, larger queue, higher throughput */
-    bool (*on_color_trans_done)(esp_lcd_panel_io_handle_t panel_io, void *user_data, void *event_data); /*!< Callback, invoked when color data was tranferred done */
-    void *user_data;    /*!< User private data, passed directly to on_trans_done's user_data */
+    esp_lcd_panel_io_color_trans_done_cb_t on_color_trans_done; /*!< Callback invoked when color data was tranferred done */
+    void *user_ctx;    /*!< User private data, passed directly to on_color_trans_done's user_ctx */
     int lcd_cmd_bits;   /*!< Bit-width of LCD command */
     int lcd_param_bits; /*!< Bit-width of LCD parameter */
     struct {
diff --git a/tools/sdk/esp32c3/include/esp_lcd/include/esp_lcd_panel_rgb.h b/tools/sdk/esp32c3/include/esp_lcd/include/esp_lcd_panel_rgb.h
index 2ddd2b6b9a6..1368bb787f6 100644
--- a/tools/sdk/esp32c3/include/esp_lcd/include/esp_lcd_panel_rgb.h
+++ b/tools/sdk/esp32c3/include/esp_lcd/include/esp_lcd_panel_rgb.h
@@ -18,6 +18,37 @@ extern "C" {
 #if SOC_LCD_RGB_SUPPORTED
 /**
  * @brief LCD RGB timing structure
+ *
+ *                                                 Total Width
+ *                             <--------------------------------------------------->
+ *                       Hsync width HBP             Active Width                HFP
+ *                             <---><--><--------------------------------------><--->
+ *                         ____    ____|_______________________________________|____|
+ *                             |___|   |                                       |    |
+ *                                     |                                       |    |
+ *                         __|         |                                       |    |
+ *            /|\    /|\  |            |                                       |    |
+ *             | VSYNC|   |            |                                       |    |
+ *             |Width\|/  |__          |                                       |    |
+ *             |     /|\     |         |                                       |    |
+ *             |  VBP |      |         |                                       |    |
+ *             |     \|/_____|_________|_______________________________________|    |
+ *             |     /|\     |         | / / / / / / / / / / / / / / / / / / / |    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *    Total    |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *    Heigh    |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |Active|      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |Heigh |      |         |/ / / / / / Active Display Area / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |     \|/_____|_________|_______________________________________|    |
+ *             |     /|\     |                                                      |
+ *             |  VFP |      |                                                      |
+ *            \|/    \|/_____|______________________________________________________|
+ *
  */
 typedef struct {
     unsigned int pclk_hz;           /*!< Frequency of pixel clock */
@@ -38,6 +69,22 @@ typedef struct {
     } flags;
 } esp_lcd_rgb_timing_t;
 
+/**
+ * @brief Type of RGB LCD panel event data
+ */
+typedef struct {
+} esp_lcd_rgb_panel_event_data_t;
+
+/**
+ * @brief Declare the prototype of the function that will be invoked when panel IO finishes transferring color data
+ *
+ * @param[in] panel LCD panel handle, returned from `esp_lcd_new_rgb_panel`
+ * @param[in] edata Panel event data, fed by driver
+ * @param[in] user_ctx User data, passed from `esp_lcd_rgb_panel_config_t`
+ * @return Whether a high priority task has been waken up by this function
+ */
+typedef bool (*esp_lcd_rgb_panel_frame_trans_done_cb_t)(esp_lcd_panel_handle_t panel, esp_lcd_rgb_panel_event_data_t *edata, void *user_ctx);
+
 /**
  * @brief LCD RGB panel configuration structure
  */
@@ -51,8 +98,8 @@ typedef struct {
     int pclk_gpio_num;            /*!< GPIO used for PCLK signal */
     int data_gpio_nums[SOC_LCD_RGB_DATA_WIDTH]; /*!< GPIOs used for data lines */
     int disp_gpio_num; /*!< GPIO used for display control signal, set to -1 if it's not used */
-    bool (*on_frame_trans_done)(esp_lcd_panel_handle_t panel, void *user_data); /*!< Callback, invoked when one frame buffer has transferred done */
-    void *user_data; /*!< User data which would be passed to on_frame_trans_done's user_data */
+    esp_lcd_rgb_panel_frame_trans_done_cb_t on_frame_trans_done; /*!< Callback invoked when one frame buffer has transferred done */
+    void *user_ctx; /*!< User data which would be passed to on_frame_trans_done's user_ctx */
     struct {
         unsigned int disp_active_low: 1; /*!< If this flag is enabled, a low level of display control signal can turn the screen on; vice versa */
         unsigned int relax_on_idle: 1;   /*!< If this flag is enabled, the host won't refresh the LCD if nothing changed in host's frame buffer (this is usefull for LCD with built-in GRAM) */
diff --git a/tools/sdk/esp32c3/include/esp_littlefs/include/esp_littlefs.h b/tools/sdk/esp32c3/include/esp_littlefs/include/esp_littlefs.h
index dbb1028790d..7b5abe9248f 100644
--- a/tools/sdk/esp32c3/include/esp_littlefs/include/esp_littlefs.h
+++ b/tools/sdk/esp32c3/include/esp_littlefs/include/esp_littlefs.h
@@ -14,7 +14,7 @@
 #include <sys/stat.h>
 #include <sys/time.h>
 #include <sys/poll.h>
-#include <dirent.h>
+#include <sys/dirent.h>
 #include <string.h>
 #include "sdkconfig.h"
 
diff --git a/tools/sdk/esp32c3/include/esp_wifi/include/esp_coexist.h b/tools/sdk/esp32c3/include/esp_wifi/include/esp_coexist.h
index 1ff624d9b55..14b7c9aa506 100644
--- a/tools/sdk/esp32c3/include/esp_wifi/include/esp_coexist.h
+++ b/tools/sdk/esp32c3/include/esp_wifi/include/esp_coexist.h
@@ -1,16 +1,8 @@
-// Copyright 2015-2018 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #ifndef __ESP_COEXIST_H__
 #define __ESP_COEXIST_H__
@@ -32,6 +24,13 @@ typedef enum {
     ESP_COEX_PREFER_NUM,            /*!< Prefer value numbers */
 } esp_coex_prefer_t;
 
+typedef enum {
+    EXTERN_COEX_WIRE_1 = 0,
+    EXTERN_COEX_WIRE_2,
+    EXTERN_COEX_WIRE_3,
+    EXTERN_COEX_WIRE_NUM,
+} external_coex_wire_t;
+
 /**
  * @brief coex status type
  */
@@ -41,6 +40,36 @@ typedef enum {
     ESP_COEX_ST_TYPE_BT,
 } esp_coex_status_type_t;
 
+/**
+ * @brief external coex gpio pti
+ */
+typedef struct {
+    int32_t in_pin0;
+    int32_t in_pin1;
+    int32_t out_pin0;
+} esp_external_coex_gpio_set_t;
+
+/**
+ * @brief external coex pti level
+ */
+typedef enum {
+    EXTERN_COEX_PTI_MID = 0,
+    EXTERN_COEX_PTI_HIGH,
+    EXTERN_COEX_PTI_NUM,
+} esp_coex_pti_level_t;
+
+/**
+ * @brief external coex pti
+ */
+typedef struct {
+    uint32_t in_pti1;
+    uint32_t in_pti2;
+    uint32_t in_pti3;
+    uint32_t out_pti1;
+    uint32_t out_pti2;
+    uint32_t out_pti3;
+} esp_external_coex_pti_set_t;
+
 #define ESP_COEX_BLE_ST_MESH_CONFIG        0x08
 #define ESP_COEX_BLE_ST_MESH_TRAFFIC       0x10
 #define ESP_COEX_BLE_ST_MESH_STANDBY       0x20
@@ -84,6 +113,18 @@ esp_err_t esp_coex_status_bit_set(esp_coex_status_type_t type, uint32_t status);
  */
 esp_err_t esp_coex_status_bit_clear(esp_coex_status_type_t type, uint32_t status);
 
+#if CONFIG_EXTERNAL_COEX_ENABLE
+/**
+ * @brief Setup gpio pin and corresponding pti level, start external coex.
+ * @param wire_type : to select the whole external coex gpio number.
+ * @param gpio_pin : gpio pin number to choose.
+ * @return : ESP_OK - success, other - failed
+ */
+esp_err_t esp_enable_extern_coex_gpio_pin(external_coex_wire_t wire_type,
+                    esp_external_coex_gpio_set_t gpio_pin);
+
+esp_err_t esp_disable_extern_coex_gpio_pin();
+#endif
 
 #ifdef __cplusplus
 }
diff --git a/tools/sdk/esp32c3/include/esp_wifi/include/esp_coexist_internal.h b/tools/sdk/esp32c3/include/esp_wifi/include/esp_coexist_internal.h
index 3501f0da6fb..7ba06d4c690 100644
--- a/tools/sdk/esp32c3/include/esp_wifi/include/esp_coexist_internal.h
+++ b/tools/sdk/esp32c3/include/esp_wifi/include/esp_coexist_internal.h
@@ -1,21 +1,14 @@
-// Copyright 2018-2018 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2018-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #ifndef __ESP_COEXIST_INTERNAL_H__
 #define __ESP_COEXIST_INTERNAL_H__
 
 #include <stdbool.h>
+#include "esp_coexist.h"
 #include "esp_coexist_adapter.h"
 
 #ifdef __cplusplus
@@ -210,6 +203,29 @@ int coex_schm_curr_phase_idx_get(void);
  */
 esp_err_t esp_coex_adapter_register(coex_adapter_funcs_t *funcs);
 
+#if CONFIG_EXTERNAL_COEX_ENABLE
+/**
+  * @brief     Set external coexistence pti level and enable it.
+  *
+  * @param     level1    external coex low pti
+  * @param     level2    external coex mid pti
+  * @param     level3    external coex high pti
+  *
+  * @return
+  *    - ESP_OK: succeed
+  */
+esp_err_t esp_coex_external_set(esp_coex_pti_level_t level1,
+         esp_coex_pti_level_t level2, esp_coex_pti_level_t level3);
+
+/**
+  * @brief     Disable external coexist
+  *
+  * @return
+  *    - ESP_OK: succeed
+  */
+void esp_coex_external_stop(void);
+#endif    /*External Coex*/
+
 /**
   * @brief     Check the MD5 values of the coexistence adapter header files in IDF and WiFi library
   *
diff --git a/tools/sdk/esp32c3/include/esp_wifi/include/esp_wifi_types.h b/tools/sdk/esp32c3/include/esp_wifi/include/esp_wifi_types.h
index 503e8d7bb05..cee23f3fafc 100644
--- a/tools/sdk/esp32c3/include/esp_wifi/include/esp_wifi_types.h
+++ b/tools/sdk/esp32c3/include/esp_wifi/include/esp_wifi_types.h
@@ -1,16 +1,8 @@
-// Copyright 2015-2016 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 
 #ifndef __ESP_WIFI_TYPES_H__
@@ -80,6 +72,7 @@ typedef enum {
     WIFI_REASON_ASSOC_NOT_AUTHED         = 9,
     WIFI_REASON_DISASSOC_PWRCAP_BAD      = 10,
     WIFI_REASON_DISASSOC_SUPCHAN_BAD     = 11,
+    WIFI_REASON_BSS_TRANSITION_DISASSOC  = 12,
     WIFI_REASON_IE_INVALID               = 13,
     WIFI_REASON_MIC_FAILURE              = 14,
     WIFI_REASON_4WAY_HANDSHAKE_TIMEOUT   = 15,
@@ -250,7 +243,8 @@ typedef struct {
     wifi_pmf_config_t pmf_cfg;    /**< Configuration for Protected Management Frame. Will be advertized in RSN Capabilities in RSN IE. */
     uint32_t rm_enabled:1;        /**< Whether Radio Measurements are enabled for the connection */
     uint32_t btm_enabled:1;       /**< Whether BSS Transition Management is enabled for the connection */
-    uint32_t reserved:30;         /**< Reserved for future feature set */
+    uint32_t mbo_enabled:1;       /**< Whether MBO is enabled for the connection */
+    uint32_t reserved:29;         /**< Reserved for future feature set */
 } wifi_sta_config_t;
 
 /** @brief Configuration data for ESP32 AP or STA.
diff --git a/tools/sdk/esp32c3/include/freertos/include/esp_additions/freertos/FreeRTOSConfig.h b/tools/sdk/esp32c3/include/freertos/include/esp_additions/freertos/FreeRTOSConfig.h
index a01a56e9fd4..675af141ebc 100644
--- a/tools/sdk/esp32c3/include/freertos/include/esp_additions/freertos/FreeRTOSConfig.h
+++ b/tools/sdk/esp32c3/include/freertos/include/esp_additions/freertos/FreeRTOSConfig.h
@@ -90,7 +90,6 @@
 #define portNUM_PROCESSORS                              1
 #endif
 
-#define configASSERT_2                                  0
 #define portUSING_MPU_WRAPPERS                          0
 #define configUSE_MUTEX                                 1
 
@@ -206,7 +205,6 @@
 #define configGENERATE_RUN_TIME_STATS                   1       /* Used by vTaskGetRunTimeStats() */
 #endif
 
-#define configUSE_TRACE_FACILITY_2                      0
 #define configBENCHMARK                                 0
 #define configUSE_16_BIT_TICKS                          0
 #define configIDLE_SHOULD_YIELD                         0
@@ -306,4 +304,9 @@ extern void vPortCleanUpTCB ( void *pxTCB );
 
 #define configTASK_NOTIFICATION_ARRAY_ENTRIES           1
 
+// backward compatibility for 4.4
+#define xTaskRemoveFromUnorderedEventList vTaskRemoveFromUnorderedEventList
+
+#define configNUM_CORES                                 portNUM_PROCESSORS
+
 #endif /* FREERTOS_CONFIG_H */
diff --git a/tools/sdk/esp32c3/include/freertos/include/freertos/event_groups.h b/tools/sdk/esp32c3/include/freertos/include/freertos/event_groups.h
index 84505ddaaa0..9792296e566 100644
--- a/tools/sdk/esp32c3/include/freertos/include/freertos/event_groups.h
+++ b/tools/sdk/esp32c3/include/freertos/include/freertos/event_groups.h
@@ -64,7 +64,7 @@
  * used to create a synchronisation point between multiple tasks (a
  * 'rendezvous').
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup EventGroup EventGroup
  * @endcond
  */
@@ -78,7 +78,7 @@
  * xEventGroupCreate() returns an EventGroupHandle_t variable that can then
  * be used as a parameter to other event group functions.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup EventGroupHandle_t EventGroupHandle_t
  * @endcond
  * \ingroup EventGroup
@@ -94,7 +94,7 @@ typedef struct EventGroupDef_t * EventGroupHandle_t;
  * number of bits it holds is set by configUSE_16_BIT_TICKS (16 bits if set to 1,
  * 32 bits if set to 0.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup EventBits_t EventBits_t
  * @endcond
  * \ingroup EventGroup
@@ -102,7 +102,7 @@ typedef struct EventGroupDef_t * EventGroupHandle_t;
 typedef TickType_t               EventBits_t;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  * EventGroupHandle_t xEventGroupCreate( void );
@@ -152,7 +152,7 @@ typedef TickType_t               EventBits_t;
  *      // The event group was created.
  *  }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupCreate xEventGroupCreate
  * @endcond
  * \ingroup EventGroup
@@ -162,7 +162,7 @@ typedef TickType_t               EventBits_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  * EventGroupHandle_t xEventGroupCreateStatic( EventGroupHandle_t * pxEventGroupBuffer );
@@ -217,7 +217,7 @@ typedef TickType_t               EventBits_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupWaitBits(    EventGroupHandle_t xEventGroup,
@@ -307,7 +307,7 @@ typedef TickType_t               EventBits_t;
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupWaitBits xEventGroupWaitBits
  * @endcond
  * \ingroup EventGroup
@@ -319,7 +319,7 @@ EventBits_t xEventGroupWaitBits( EventGroupHandle_t xEventGroup,
                                  TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup, const EventBits_t uxBitsToClear );
@@ -372,7 +372,7 @@ EventBits_t xEventGroupWaitBits( EventGroupHandle_t xEventGroup,
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupClearBits xEventGroupClearBits
  * @endcond
  * \ingroup EventGroup
@@ -381,7 +381,7 @@ EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup,
                                   const EventBits_t uxBitsToClear ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  BaseType_t xEventGroupClearBitsFromISR( EventGroupHandle_t xEventGroup, const EventBits_t uxBitsToSet );
@@ -432,7 +432,7 @@ EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup,
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupClearBitsFromISR xEventGroupClearBitsFromISR
  * @endcond
  * \ingroup EventGroup
@@ -446,7 +446,7 @@ EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup,
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup, const EventBits_t uxBitsToSet );
@@ -516,7 +516,7 @@ EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup,
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupSetBits xEventGroupSetBits
  * @endcond
  * \ingroup EventGroup
@@ -525,7 +525,7 @@ EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup,
                                 const EventBits_t uxBitsToSet ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  BaseType_t xEventGroupSetBitsFromISR( EventGroupHandle_t xEventGroup, const EventBits_t uxBitsToSet, BaseType_t *pxHigherPriorityTaskWoken );
@@ -595,7 +595,7 @@ EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup,
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupSetBitsFromISR xEventGroupSetBitsFromISR
  * @endcond
  * \ingroup EventGroup
@@ -610,7 +610,7 @@ EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup,
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupSync(    EventGroupHandle_t xEventGroup,
@@ -732,7 +732,7 @@ EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup,
  * }
  *
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupSync xEventGroupSync
  * @endcond
  * \ingroup EventGroup
@@ -744,7 +744,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupGetBits( EventGroupHandle_t xEventGroup );
@@ -758,7 +758,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
  *
  * @return The event group bits at the time xEventGroupGetBits() was called.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupGetBits xEventGroupGetBits
  * @endcond
  * \ingroup EventGroup
@@ -766,7 +766,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
 #define xEventGroupGetBits( xEventGroup )    xEventGroupClearBits( xEventGroup, 0 )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupGetBitsFromISR( EventGroupHandle_t xEventGroup );
@@ -779,7 +779,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
  *
  * @return The event group bits at the time xEventGroupGetBitsFromISR() was called.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupGetBitsFromISR xEventGroupGetBitsFromISR
  * @endcond
  * \ingroup EventGroup
@@ -787,7 +787,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
 EventBits_t xEventGroupGetBitsFromISR( EventGroupHandle_t xEventGroup ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  void xEventGroupDelete( EventGroupHandle_t xEventGroup );
@@ -802,7 +802,7 @@ EventBits_t xEventGroupGetBitsFromISR( EventGroupHandle_t xEventGroup ) PRIVILEG
  */
 void vEventGroupDelete( EventGroupHandle_t xEventGroup ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 
 /* For internal use only. */
 void vEventGroupSetBitsCallback( void * pvEventGroup,
diff --git a/tools/sdk/esp32c3/include/freertos/include/freertos/message_buffer.h b/tools/sdk/esp32c3/include/freertos/include/freertos/message_buffer.h
index e57c589fbac..af5c3290b7c 100644
--- a/tools/sdk/esp32c3/include/freertos/include/freertos/message_buffer.h
+++ b/tools/sdk/esp32c3/include/freertos/include/freertos/message_buffer.h
@@ -85,7 +85,7 @@ typedef void * MessageBufferHandle_t;
 /*-----------------------------------------------------------*/
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -139,7 +139,7 @@ typedef void * MessageBufferHandle_t;
  *  }
  *
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferCreate xMessageBufferCreate
  * @endcond
  * \ingroup MessageBufferManagement
@@ -148,7 +148,7 @@ typedef void * MessageBufferHandle_t;
     ( MessageBufferHandle_t ) xStreamBufferGenericCreate( xBufferSizeBytes, ( size_t ) 0, pdTRUE )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -210,7 +210,7 @@ typedef void * MessageBufferHandle_t;
  * }
  *
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferCreateStatic xMessageBufferCreateStatic
  * @endcond
  * \ingroup MessageBufferManagement
@@ -219,7 +219,7 @@ typedef void * MessageBufferHandle_t;
     ( MessageBufferHandle_t ) xStreamBufferGenericCreateStatic( xBufferSizeBytes, 0, pdTRUE, pucMessageBufferStorageArea, pxStaticMessageBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -314,7 +314,7 @@ typedef void * MessageBufferHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferSend xMessageBufferSend
  * @endcond
  * \ingroup MessageBufferManagement
@@ -323,7 +323,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferSend( ( StreamBufferHandle_t ) xMessageBuffer, pvTxData, xDataLengthBytes, xTicksToWait )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -423,7 +423,7 @@ typedef void * MessageBufferHandle_t;
  *  portYIELD_FROM_ISR( xHigherPriorityTaskWoken );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferSendFromISR xMessageBufferSendFromISR
  * @endcond
  * \ingroup MessageBufferManagement
@@ -432,7 +432,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferSendFromISR( ( StreamBufferHandle_t ) xMessageBuffer, pvTxData, xDataLengthBytes, pxHigherPriorityTaskWoken )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -516,7 +516,7 @@ typedef void * MessageBufferHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferReceive xMessageBufferReceive
  * @endcond
  * \ingroup MessageBufferManagement
@@ -526,7 +526,7 @@ typedef void * MessageBufferHandle_t;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -622,7 +622,7 @@ typedef void * MessageBufferHandle_t;
  *  portYIELD_FROM_ISR( xHigherPriorityTaskWoken );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferReceiveFromISR xMessageBufferReceiveFromISR
  * @endcond
  * \ingroup MessageBufferManagement
@@ -631,7 +631,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferReceiveFromISR( ( StreamBufferHandle_t ) xMessageBuffer, pvRxData, xBufferLengthBytes, pxHigherPriorityTaskWoken )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -654,7 +654,7 @@ typedef void * MessageBufferHandle_t;
     vStreamBufferDelete( ( StreamBufferHandle_t ) xMessageBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * BaseType_t xMessageBufferIsFull( MessageBufferHandle_t xMessageBuffer ) );
@@ -674,7 +674,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferIsFull( ( StreamBufferHandle_t ) xMessageBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * BaseType_t xMessageBufferIsEmpty( MessageBufferHandle_t xMessageBuffer ) );
@@ -693,7 +693,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferIsEmpty( ( StreamBufferHandle_t ) xMessageBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * BaseType_t xMessageBufferReset( MessageBufferHandle_t xMessageBuffer );
@@ -712,7 +712,7 @@ typedef void * MessageBufferHandle_t;
  * the message queue to wait for space to become available, or to wait for a
  * a message to be available, then pdFAIL is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferReset xMessageBufferReset
  * @endcond
  * \ingroup MessageBufferManagement
@@ -722,7 +722,7 @@ typedef void * MessageBufferHandle_t;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * size_t xMessageBufferSpaceAvailable( MessageBufferHandle_t xMessageBuffer ) );
@@ -740,7 +740,7 @@ typedef void * MessageBufferHandle_t;
  * architecture, so if xMessageBufferSpacesAvailable() returns 10, then the size
  * of the largest message that can be written to the message buffer is 6 bytes.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferSpaceAvailable xMessageBufferSpaceAvailable
  * @endcond
  * \ingroup MessageBufferManagement
@@ -751,7 +751,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferSpacesAvailable( ( StreamBufferHandle_t ) xMessageBuffer ) /* Corrects typo in original macro name. */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * size_t xMessageBufferNextLengthBytes( MessageBufferHandle_t xMessageBuffer ) );
@@ -767,7 +767,7 @@ typedef void * MessageBufferHandle_t;
  * @return The length (in bytes) of the next message in the message buffer, or 0
  * if the message buffer is empty.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferNextLengthBytes xMessageBufferNextLengthBytes
  * @endcond
  * \ingroup MessageBufferManagement
@@ -776,7 +776,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferNextMessageLengthBytes( ( StreamBufferHandle_t ) xMessageBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -811,7 +811,7 @@ typedef void * MessageBufferHandle_t;
  * @return If a task was removed from the Blocked state then pdTRUE is returned.
  * Otherwise pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferSendCompletedFromISR xMessageBufferSendCompletedFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -820,7 +820,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferSendCompletedFromISR( ( StreamBufferHandle_t ) xMessageBuffer, pxHigherPriorityTaskWoken )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -856,7 +856,7 @@ typedef void * MessageBufferHandle_t;
  * @return If a task was removed from the Blocked state then pdTRUE is returned.
  * Otherwise pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferReceiveCompletedFromISR xMessageBufferReceiveCompletedFromISR
  * @endcond
  * \ingroup StreamBufferManagement
diff --git a/tools/sdk/esp32c3/include/freertos/include/freertos/queue.h b/tools/sdk/esp32c3/include/freertos/include/freertos/queue.h
index 81cccc05df3..05ca7de4546 100644
--- a/tools/sdk/esp32c3/include/freertos/include/freertos/queue.h
+++ b/tools/sdk/esp32c3/include/freertos/include/freertos/queue.h
@@ -62,7 +62,7 @@ typedef struct QueueDefinition   * QueueSetHandle_t;
  */
 typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 
 /* For internal use only. */
 #define queueSEND_TO_BACK                     ( ( BaseType_t ) 0 )
@@ -80,7 +80,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 /** @endcond */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * QueueHandle_t xQueueCreate(
@@ -146,7 +146,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueCreate xQueueCreate
  * @endcond
  * \ingroup QueueManagement
@@ -156,7 +156,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * QueueHandle_t xQueueCreateStatic(
@@ -235,7 +235,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueCreateStatic xQueueCreateStatic
  * @endcond
  * \ingroup QueueManagement
@@ -245,7 +245,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendToToFront(
@@ -321,7 +321,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSend xQueueSend
  * @endcond
  * \ingroup QueueManagement
@@ -330,7 +330,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
     xQueueGenericSend( ( xQueue ), ( pvItemToQueue ), ( xTicksToWait ), queueSEND_TO_FRONT )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendToBack(
@@ -408,7 +408,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSend xQueueSend
  * @endcond
  * \ingroup QueueManagement
@@ -417,7 +417,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
     xQueueGenericSend( ( xQueue ), ( pvItemToQueue ), ( xTicksToWait ), queueSEND_TO_BACK )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSend(
@@ -497,7 +497,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSend xQueueSend
  * @endcond
  * \ingroup QueueManagement
@@ -506,7 +506,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
     xQueueGenericSend( ( xQueue ), ( pvItemToQueue ), ( xTicksToWait ), queueSEND_TO_BACK )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueOverwrite(
@@ -585,7 +585,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ...
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueOverwrite xQueueOverwrite
  * @endcond
  * \ingroup QueueManagement
@@ -595,7 +595,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueGenericSend(
@@ -678,7 +678,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSend xQueueSend
  * @endcond
  * \ingroup QueueManagement
@@ -689,7 +689,7 @@ BaseType_t xQueueGenericSend( QueueHandle_t xQueue,
                               const BaseType_t xCopyPosition ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueuePeek(
@@ -780,7 +780,7 @@ BaseType_t xQueueGenericSend( QueueHandle_t xQueue,
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueuePeek xQueuePeek
  * @endcond
  * \ingroup QueueManagement
@@ -790,7 +790,7 @@ BaseType_t xQueuePeek( QueueHandle_t xQueue,
                        TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueuePeekFromISR(
@@ -820,7 +820,7 @@ BaseType_t xQueuePeek( QueueHandle_t xQueue,
  * @return pdTRUE if an item was successfully received from the queue,
  * otherwise pdFALSE.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueuePeekFromISR xQueuePeekFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -829,7 +829,7 @@ BaseType_t xQueuePeekFromISR( QueueHandle_t xQueue,
                               void * const pvBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueReceive(
@@ -917,7 +917,7 @@ BaseType_t xQueuePeekFromISR( QueueHandle_t xQueue,
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueReceive xQueueReceive
  * @endcond
  * \ingroup QueueManagement
@@ -927,7 +927,7 @@ BaseType_t xQueueReceive( QueueHandle_t xQueue,
                           TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * UBaseType_t uxQueueMessagesWaiting( const QueueHandle_t xQueue );
@@ -940,7 +940,7 @@ BaseType_t xQueueReceive( QueueHandle_t xQueue,
  *
  * @return The number of messages available in the queue.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup uxQueueMessagesWaiting uxQueueMessagesWaiting
  * @endcond
  * \ingroup QueueManagement
@@ -948,7 +948,7 @@ BaseType_t xQueueReceive( QueueHandle_t xQueue,
 UBaseType_t uxQueueMessagesWaiting( const QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * UBaseType_t uxQueueSpacesAvailable( const QueueHandle_t xQueue );
@@ -963,7 +963,7 @@ UBaseType_t uxQueueMessagesWaiting( const QueueHandle_t xQueue ) PRIVILEGED_FUNC
  *
  * @return The number of spaces available in the queue.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup uxQueueMessagesWaiting uxQueueMessagesWaiting
  * @endcond
  * \ingroup QueueManagement
@@ -971,7 +971,7 @@ UBaseType_t uxQueueMessagesWaiting( const QueueHandle_t xQueue ) PRIVILEGED_FUNC
 UBaseType_t uxQueueSpacesAvailable( const QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * void vQueueDelete( QueueHandle_t xQueue );
@@ -983,7 +983,7 @@ UBaseType_t uxQueueSpacesAvailable( const QueueHandle_t xQueue ) PRIVILEGED_FUNC
  *
  * @param xQueue A handle to the queue to be deleted.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vQueueDelete vQueueDelete
  * @endcond
  * \ingroup QueueManagement
@@ -991,7 +991,7 @@ UBaseType_t uxQueueSpacesAvailable( const QueueHandle_t xQueue ) PRIVILEGED_FUNC
 void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendToFrontFromISR(
@@ -1057,7 +1057,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSendFromISR xQueueSendFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1067,7 +1067,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendToBackFromISR(
@@ -1133,7 +1133,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSendFromISR xQueueSendFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1142,7 +1142,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
     xQueueGenericSendFromISR( ( xQueue ), ( pvItemToQueue ), ( pxHigherPriorityTaskWoken ), queueSEND_TO_BACK )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueOverwriteFromISR(
@@ -1225,7 +1225,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueOverwriteFromISR xQueueOverwriteFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1234,7 +1234,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
     xQueueGenericSendFromISR( ( xQueue ), ( pvItemToQueue ), ( pxHigherPriorityTaskWoken ), queueOVERWRITE )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendFromISR(
@@ -1304,7 +1304,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSendFromISR xQueueSendFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1312,10 +1312,10 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 #define xQueueSendFromISR( xQueue, pvItemToQueue, pxHigherPriorityTaskWoken ) \
     xQueueGenericSendFromISR( ( xQueue ), ( pvItemToQueue ), ( pxHigherPriorityTaskWoken ), queueSEND_TO_BACK )
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /**@{*/
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueGenericSendFromISR(
@@ -1402,7 +1402,7 @@ BaseType_t xQueueGiveFromISR( QueueHandle_t xQueue,
 /** @endcond */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueReceiveFromISR(
@@ -1487,7 +1487,7 @@ BaseType_t xQueueGiveFromISR( QueueHandle_t xQueue,
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueReceiveFromISR xQueueReceiveFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1504,7 +1504,7 @@ BaseType_t xQueueIsQueueEmptyFromISR( const QueueHandle_t xQueue ) PRIVILEGED_FU
 BaseType_t xQueueIsQueueFullFromISR( const QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 UBaseType_t uxQueueMessagesWaitingFromISR( const QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /*
  * The functions defined above are for passing data to and from tasks.  The
  * functions below are the equivalents for passing data to and from
@@ -1778,7 +1778,7 @@ QueueSetMemberHandle_t xQueueSelectFromSet( QueueSetHandle_t xQueueSet,
  */
 QueueSetMemberHandle_t xQueueSelectFromSetFromISR( QueueSetHandle_t xQueueSet ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 
 /* Not public API functions. */
 void vQueueWaitForMessageRestricted( QueueHandle_t xQueue,
diff --git a/tools/sdk/esp32c3/include/freertos/include/freertos/semphr.h b/tools/sdk/esp32c3/include/freertos/include/freertos/semphr.h
index 7e99c0b396c..2041641b91d 100644
--- a/tools/sdk/esp32c3/include/freertos/include/freertos/semphr.h
+++ b/tools/sdk/esp32c3/include/freertos/include/freertos/semphr.h
@@ -39,7 +39,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define semSEMAPHORE_QUEUE_ITEM_LENGTH      ( ( uint8_t ) 0U )
 #define semGIVE_BLOCK_TIME                  ( ( TickType_t ) 0U )
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /**
  * semphr. h
  * @code{c}
@@ -88,7 +88,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vSemaphoreCreateBinary vSemaphoreCreateBinary
  * @endcond
  * \ingroup Semaphores
@@ -106,7 +106,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 /** @endcond */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateBinary( void );
@@ -163,7 +163,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateBinary xSemaphoreCreateBinary
  * @endcond
  * \ingroup Semaphores
@@ -173,7 +173,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateBinaryStatic( StaticSemaphore_t *pxSemaphoreBuffer );
@@ -229,7 +229,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  // Rest of task code goes here.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateBinaryStatic xSemaphoreCreateBinaryStatic
  * @endcond
  * \ingroup Semaphores
@@ -239,7 +239,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * xSemaphoreTake(
@@ -304,7 +304,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreTake xSemaphoreTake
  * @endcond
  * \ingroup Semaphores
@@ -312,7 +312,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreTake( xSemaphore, xBlockTime )    xQueueSemaphoreTake( ( xSemaphore ), ( xBlockTime ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * xSemaphoreTakeRecursive(
@@ -403,7 +403,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreTakeRecursive xSemaphoreTakeRecursive
  * @endcond
  * \ingroup Semaphores
@@ -465,7 +465,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreGive xSemaphoreGive
  * @endcond
  * \ingroup Semaphores
@@ -473,7 +473,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreGive( xSemaphore )    xQueueGenericSend( ( QueueHandle_t ) ( xSemaphore ), NULL, semGIVE_BLOCK_TIME, queueSEND_TO_BACK )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * xSemaphoreGiveRecursive( SemaphoreHandle_t xMutex );
@@ -555,7 +555,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreGiveRecursive xSemaphoreGiveRecursive
  * @endcond
  * \ingroup Semaphores
@@ -641,7 +641,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreGiveFromISR xSemaphoreGiveFromISR
  * @endcond
  * \ingroup Semaphores
@@ -649,7 +649,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreGiveFromISR( xSemaphore, pxHigherPriorityTaskWoken )    xQueueGiveFromISR( ( QueueHandle_t ) ( xSemaphore ), ( pxHigherPriorityTaskWoken ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * xSemaphoreTakeFromISR(
@@ -686,7 +686,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreTakeFromISR( xSemaphore, pxHigherPriorityTaskWoken )    xQueueReceiveFromISR( ( QueueHandle_t ) ( xSemaphore ), NULL, ( pxHigherPriorityTaskWoken ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateMutex( void );
@@ -741,7 +741,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateMutex xSemaphoreCreateMutex
  * @endcond
  * \ingroup Semaphores
@@ -751,7 +751,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateMutexStatic( StaticSemaphore_t *pxMutexBuffer );
@@ -808,7 +808,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  // so there is no need to check it.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateMutexStatic xSemaphoreCreateMutexStatic
  * @endcond
  * \ingroup Semaphores
@@ -951,7 +951,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateCounting( UBaseType_t uxMaxCount, UBaseType_t uxInitialCount );
@@ -1027,7 +1027,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateCounting xSemaphoreCreateCounting
  * @endcond
  * \ingroup Semaphores
@@ -1037,7 +1037,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateCountingStatic( UBaseType_t uxMaxCount, UBaseType_t uxInitialCount, StaticSemaphore_t *pxSemaphoreBuffer );
@@ -1118,7 +1118,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  // is no need to check its value.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateCountingStatic xSemaphoreCreateCountingStatic
  * @endcond
  * \ingroup Semaphores
@@ -1128,7 +1128,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * void vSemaphoreDelete( SemaphoreHandle_t xSemaphore );
@@ -1140,7 +1140,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *
  * @param xSemaphore A handle to the semaphore to be deleted.
  *
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * \defgroup vSemaphoreDelete vSemaphoreDelete
  * @endcond
  * \ingroup Semaphores
@@ -1148,7 +1148,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define vSemaphoreDelete( xSemaphore )                   vQueueDelete( ( QueueHandle_t ) ( xSemaphore ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr.h
  * @code{c}
  * TaskHandle_t xSemaphoreGetMutexHolder( SemaphoreHandle_t xMutex );
@@ -1167,7 +1167,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreGetMutexHolder( xSemaphore )           xQueueGetMutexHolder( ( xSemaphore ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr.h
  * @code{c}
  * TaskHandle_t xSemaphoreGetMutexHolderFromISR( SemaphoreHandle_t xMutex );
@@ -1182,7 +1182,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreGetMutexHolderFromISR( xSemaphore )    xQueueGetMutexHolderFromISR( ( xSemaphore ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr.h
  * @code{c}
  * UBaseType_t uxSemaphoreGetCount( SemaphoreHandle_t xSemaphore );
diff --git a/tools/sdk/esp32c3/include/freertos/include/freertos/stream_buffer.h b/tools/sdk/esp32c3/include/freertos/include/freertos/stream_buffer.h
index 9e58cff120c..a20dcf0375e 100644
--- a/tools/sdk/esp32c3/include/freertos/include/freertos/stream_buffer.h
+++ b/tools/sdk/esp32c3/include/freertos/include/freertos/stream_buffer.h
@@ -71,7 +71,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -134,7 +134,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferCreate xStreamBufferCreate
  * @endcond
  * \ingroup StreamBufferManagement
@@ -142,7 +142,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
 #define xStreamBufferCreate( xBufferSizeBytes, xTriggerLevelBytes )    xStreamBufferGenericCreate( xBufferSizeBytes, xTriggerLevelBytes, pdFALSE )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -220,7 +220,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
  * }
  *
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferCreateStatic xStreamBufferCreateStatic
  * @endcond
  * \ingroup StreamBufferManagement
@@ -229,7 +229,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
     xStreamBufferGenericCreateStatic( xBufferSizeBytes, xTriggerLevelBytes, pdFALSE, pucStreamBufferStorageArea, pxStaticStreamBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -319,7 +319,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSend xStreamBufferSend
  * @endcond
  * \ingroup StreamBufferManagement
@@ -330,7 +330,7 @@ size_t xStreamBufferSend( StreamBufferHandle_t xStreamBuffer,
                           TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -424,7 +424,7 @@ size_t xStreamBufferSend( StreamBufferHandle_t xStreamBuffer,
  *  taskYIELD_FROM_ISR( xHigherPriorityTaskWoken );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSendFromISR xStreamBufferSendFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -435,7 +435,7 @@ size_t xStreamBufferSendFromISR( StreamBufferHandle_t xStreamBuffer,
                                  BaseType_t * const pxHigherPriorityTaskWoken ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -517,7 +517,7 @@ size_t xStreamBufferSendFromISR( StreamBufferHandle_t xStreamBuffer,
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferReceive xStreamBufferReceive
  * @endcond
  * \ingroup StreamBufferManagement
@@ -528,7 +528,7 @@ size_t xStreamBufferReceive( StreamBufferHandle_t xStreamBuffer,
                              TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -607,7 +607,7 @@ size_t xStreamBufferReceive( StreamBufferHandle_t xStreamBuffer,
  *  taskYIELD_FROM_ISR( xHigherPriorityTaskWoken );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferReceiveFromISR xStreamBufferReceiveFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -618,7 +618,7 @@ size_t xStreamBufferReceiveFromISR( StreamBufferHandle_t xStreamBuffer,
                                     BaseType_t * const pxHigherPriorityTaskWoken ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -636,7 +636,7 @@ size_t xStreamBufferReceiveFromISR( StreamBufferHandle_t xStreamBuffer,
  *
  * @param xStreamBuffer The handle of the stream buffer to be deleted.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vStreamBufferDelete vStreamBufferDelete
  * @endcond
  * \ingroup StreamBufferManagement
@@ -644,7 +644,7 @@ size_t xStreamBufferReceiveFromISR( StreamBufferHandle_t xStreamBuffer,
 void vStreamBufferDelete( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -660,7 +660,7 @@ void vStreamBufferDelete( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTI
  * @return If the stream buffer is full then pdTRUE is returned.  Otherwise
  * pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferIsFull xStreamBufferIsFull
  * @endcond
  * \ingroup StreamBufferManagement
@@ -668,7 +668,7 @@ void vStreamBufferDelete( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTI
 BaseType_t xStreamBufferIsFull( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -684,7 +684,7 @@ BaseType_t xStreamBufferIsFull( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_
  * @return If the stream buffer is empty then pdTRUE is returned.  Otherwise
  * pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferIsEmpty xStreamBufferIsEmpty
  * @endcond
  * \ingroup StreamBufferManagement
@@ -692,7 +692,7 @@ BaseType_t xStreamBufferIsFull( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_
 BaseType_t xStreamBufferIsEmpty( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -711,7 +711,7 @@ BaseType_t xStreamBufferIsEmpty( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED
  * a task blocked waiting to send to or read from the stream buffer then the
  * stream buffer is not reset and pdFAIL is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferReset xStreamBufferReset
  * @endcond
  * \ingroup StreamBufferManagement
@@ -719,7 +719,7 @@ BaseType_t xStreamBufferIsEmpty( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED
 BaseType_t xStreamBufferReset( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -736,7 +736,7 @@ BaseType_t xStreamBufferReset( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_F
  * @return The number of bytes that can be written to the stream buffer before
  * the stream buffer would be full.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSpacesAvailable xStreamBufferSpacesAvailable
  * @endcond
  * \ingroup StreamBufferManagement
@@ -744,7 +744,7 @@ BaseType_t xStreamBufferReset( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_F
 size_t xStreamBufferSpacesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -761,7 +761,7 @@ size_t xStreamBufferSpacesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVIL
  * @return The number of bytes that can be read from the stream buffer before
  * the stream buffer would be empty.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferBytesAvailable xStreamBufferBytesAvailable
  * @endcond
  * \ingroup StreamBufferManagement
@@ -769,7 +769,7 @@ size_t xStreamBufferSpacesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVIL
 size_t xStreamBufferBytesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -802,7 +802,7 @@ size_t xStreamBufferBytesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVILE
  * then the trigger level will be updated and pdTRUE is returned.  Otherwise
  * pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSetTriggerLevel xStreamBufferSetTriggerLevel
  * @endcond
  * \ingroup StreamBufferManagement
@@ -811,7 +811,7 @@ BaseType_t xStreamBufferSetTriggerLevel( StreamBufferHandle_t xStreamBuffer,
                                          size_t xTriggerLevel ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -846,7 +846,7 @@ BaseType_t xStreamBufferSetTriggerLevel( StreamBufferHandle_t xStreamBuffer,
  * @return If a task was removed from the Blocked state then pdTRUE is returned.
  * Otherwise pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSendCompletedFromISR xStreamBufferSendCompletedFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -855,7 +855,7 @@ BaseType_t xStreamBufferSendCompletedFromISR( StreamBufferHandle_t xStreamBuffer
                                               BaseType_t * pxHigherPriorityTaskWoken ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -891,7 +891,7 @@ BaseType_t xStreamBufferSendCompletedFromISR( StreamBufferHandle_t xStreamBuffer
  * @return If a task was removed from the Blocked state then pdTRUE is returned.
  * Otherwise pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferReceiveCompletedFromISR xStreamBufferReceiveCompletedFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -899,7 +899,7 @@ BaseType_t xStreamBufferSendCompletedFromISR( StreamBufferHandle_t xStreamBuffer
 BaseType_t xStreamBufferReceiveCompletedFromISR( StreamBufferHandle_t xStreamBuffer,
                                                  BaseType_t * pxHigherPriorityTaskWoken ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /* Functions below here are not part of the public API. */
 StreamBufferHandle_t xStreamBufferGenericCreate( size_t xBufferSizeBytes,
                                                  size_t xTriggerLevelBytes,
diff --git a/tools/sdk/esp32c3/include/freertos/include/freertos/task.h b/tools/sdk/esp32c3/include/freertos/include/freertos/task.h
index 9135b76f014..125a924d061 100644
--- a/tools/sdk/esp32c3/include/freertos/include/freertos/task.h
+++ b/tools/sdk/esp32c3/include/freertos/include/freertos/task.h
@@ -76,7 +76,7 @@
  * returns (via a pointer parameter) an TaskHandle_t variable that can then
  * be used as a parameter to vTaskDelete to delete the task.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup TaskHandle_t TaskHandle_t
  * @endcond
  * \ingroup Tasks
@@ -114,7 +114,7 @@ typedef enum
     eSetValueWithoutOverwrite     /* Set the task's notification value if the previous value has been read by the task. */
 } eNotifyAction;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /**
  * Used internally only.
  */
@@ -189,11 +189,13 @@ typedef enum
 #define tskIDLE_PRIORITY    ( ( UBaseType_t ) 0U )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro for forcing a context switch.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskYIELD taskYIELD
  * @endcond
  * \ingroup SchedulerControl
@@ -201,7 +203,9 @@ typedef enum
 #define taskYIELD()                        portYIELD()
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro to mark the start of a critical code region.  Preemptive context
  * switches cannot occur when in a critical region.
@@ -209,7 +213,7 @@ typedef enum
  * @note This may alter the stack (depending on the portable implementation)
  * so must be used with care!
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskENTER_CRITICAL taskENTER_CRITICAL
  * @endcond
  * \ingroup SchedulerControl
@@ -228,7 +232,9 @@ typedef enum
 #endif //  ESP_PLATFORM
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro to mark the end of a critical code region.  Preemptive context
  * switches cannot occur when in a critical region.
@@ -236,7 +242,7 @@ typedef enum
  * @note This may alter the stack (depending on the portable implementation)
  * so must be used with care!
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskEXIT_CRITICAL taskEXIT_CRITICAL
  * @endcond
  * \ingroup SchedulerControl
@@ -255,11 +261,13 @@ typedef enum
 #define taskEXIT_CRITICAL_ISR( )        portEXIT_CRITICAL_ISR( )
 #endif // ESP_PLATFORM
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro to disable all maskable interrupts.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskDISABLE_INTERRUPTS taskDISABLE_INTERRUPTS
  * @endcond
  * \ingroup SchedulerControl
@@ -267,11 +275,13 @@ typedef enum
 #define taskDISABLE_INTERRUPTS()           portDISABLE_INTERRUPTS()
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro to enable microcontroller interrupts.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskENABLE_INTERRUPTS taskENABLE_INTERRUPTS
  * @endcond
  * \ingroup SchedulerControl
@@ -422,7 +432,7 @@ typedef enum
  *   }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCreate xTaskCreate
  * @endcond
  * \ingroup Tasks
@@ -430,14 +440,14 @@ typedef enum
 #if ( configSUPPORT_DYNAMIC_ALLOCATION == 1 )
 
     static inline IRAM_ATTR BaseType_t xTaskCreate(
-            TaskFunction_t pvTaskCode,
-            const char * const pcName,
-            const uint32_t usStackDepth,
-            void * const pvParameters,
-            UBaseType_t uxPriority,
-            TaskHandle_t * const pvCreatedTask)
+                            TaskFunction_t pvTaskCode,
+                            const char * const pcName,     /*lint !e971 Unqualified char types are allowed for strings and single characters only. */
+                            const uint32_t usStackDepth,
+                            void * const pvParameters,
+                            UBaseType_t uxPriority,
+                            TaskHandle_t * const pxCreatedTask) PRIVILEGED_FUNCTION
     {
-        return xTaskCreatePinnedToCore( pvTaskCode, pcName, usStackDepth, pvParameters, uxPriority, pvCreatedTask, tskNO_AFFINITY );
+        return xTaskCreatePinnedToCore( pvTaskCode, pcName, usStackDepth, pvParameters, uxPriority, pxCreatedTask, tskNO_AFFINITY );
     }
 
 #endif
@@ -599,20 +609,20 @@ typedef enum
 
 #if( configSUPPORT_STATIC_ALLOCATION == 1 )
     static inline IRAM_ATTR TaskHandle_t xTaskCreateStatic(
-            TaskFunction_t pvTaskCode,
-            const char * const pcName,
-            const uint32_t ulStackDepth,
-            void * const pvParameters,
-            UBaseType_t uxPriority,
-            StackType_t * const pxStackBuffer,
-            StaticTask_t * const pxTaskBuffer)
+                                    TaskFunction_t pvTaskCode,
+                                    const char * const pcName,     /*lint !e971 Unqualified char types are allowed for strings and single characters only. */
+                                    const uint32_t ulStackDepth,
+                                    void * const pvParameters,
+                                    UBaseType_t uxPriority,
+                                    StackType_t * const puxStackBuffer,
+                                    StaticTask_t * const pxTaskBuffer) PRIVILEGED_FUNCTION
     {
-        return xTaskCreateStaticPinnedToCore( pvTaskCode, pcName, ulStackDepth, pvParameters, uxPriority, pxStackBuffer, pxTaskBuffer, tskNO_AFFINITY );
+        return xTaskCreateStaticPinnedToCore( pvTaskCode, pcName, ulStackDepth, pvParameters, uxPriority, puxStackBuffer, pxTaskBuffer, tskNO_AFFINITY );
     }
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskCreateRestricted( TaskParameters_t *pxTaskDefinition, TaskHandle_t *pxCreatedTask );
@@ -683,18 +693,18 @@ typedef enum
  *  for( ;; );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCreateRestricted xTaskCreateRestricted
  * @endcond
  * \ingroup Tasks
  */
 #if ( portUSING_MPU_WRAPPERS == 1 )
     BaseType_t xTaskCreateRestricted( const TaskParameters_t * const pxTaskDefinition,
-                                      TaskHandle_t * pxCreatedTask );
+                                      TaskHandle_t * pxCreatedTask ) PRIVILEGED_FUNCTION;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskCreateRestrictedStatic( TaskParameters_t *pxTaskDefinition, TaskHandle_t *pxCreatedTask );
@@ -777,7 +787,7 @@ typedef enum
  *  for( ;; );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCreateRestrictedStatic xTaskCreateRestrictedStatic
  * @endcond
  * \ingroup Tasks
@@ -788,7 +798,7 @@ typedef enum
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskAllocateMPURegions( TaskHandle_t xTask, const MemoryRegion_t * const pxRegions );
@@ -833,7 +843,7 @@ typedef enum
  *  // defined or shared regions have been declared elsewhere).
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCreateRestricted xTaskCreateRestricted
  * @endcond
  * \ingroup Tasks
@@ -842,7 +852,7 @@ void vTaskAllocateMPURegions( TaskHandle_t xTask,
                               const MemoryRegion_t * const pxRegions ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskDelete( TaskHandle_t xTask );
@@ -881,7 +891,7 @@ void vTaskAllocateMPURegions( TaskHandle_t xTask,
  *   vTaskDelete( xHandle );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskDelete vTaskDelete
  * @endcond
  * \ingroup Tasks
@@ -893,10 +903,12 @@ void vTaskDelete( TaskHandle_t xTaskToDelete ) PRIVILEGED_FUNCTION;
  *----------------------------------------------------------*/
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskDelay( const TickType_t xTicksToDelay );
  * @endcode
+ * @endcond
  *
  * Delay a task for a given number of ticks.  The actual time that the
  * task remains blocked depends on the tick rate.  The constant
@@ -938,7 +950,7 @@ void vTaskDelete( TaskHandle_t xTaskToDelete ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskDelay vTaskDelay
  * @endcond
  * \ingroup TaskCtrl
@@ -946,10 +958,12 @@ void vTaskDelete( TaskHandle_t xTaskToDelete ) PRIVILEGED_FUNCTION;
 void vTaskDelay( const TickType_t xTicksToDelay ) PRIVILEGED_FUNCTION;
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskDelayUntil( TickType_t *pxPreviousWakeTime, const TickType_t xTimeIncrement );
  * @endcode
+ * @endcond
  *
  * INCLUDE_xTaskDelayUntil must be defined as 1 for this function to be available.
  * See the configuration section for more information.
@@ -1007,7 +1021,7 @@ void vTaskDelay( const TickType_t xTicksToDelay ) PRIVILEGED_FUNCTION;
  *     }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskDelayUntil xTaskDelayUntil
  * @endcond
  * \ingroup TaskCtrl
@@ -1026,7 +1040,7 @@ BaseType_t xTaskDelayUntil( TickType_t * const pxPreviousWakeTime,
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskAbortDelay( TaskHandle_t xTask );
@@ -1054,7 +1068,7 @@ BaseType_t xTaskDelayUntil( TickType_t * const pxPreviousWakeTime,
  * @return If the task referenced by xTask was not in the Blocked state then
  * pdFAIL is returned.  Otherwise pdPASS is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskAbortDelay xTaskAbortDelay
  * @endcond
  * \ingroup TaskCtrl
@@ -1062,7 +1076,7 @@ BaseType_t xTaskDelayUntil( TickType_t * const pxPreviousWakeTime,
 BaseType_t xTaskAbortDelay( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * UBaseType_t uxTaskPriorityGet( const TaskHandle_t xTask );
@@ -1107,7 +1121,7 @@ BaseType_t xTaskAbortDelay( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
  *   }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup uxTaskPriorityGet uxTaskPriorityGet
  * @endcond
  * \ingroup TaskCtrl
@@ -1115,7 +1129,7 @@ BaseType_t xTaskAbortDelay( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 UBaseType_t uxTaskPriorityGet( const TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * UBaseType_t uxTaskPriorityGetFromISR( const TaskHandle_t xTask );
@@ -1127,7 +1141,7 @@ UBaseType_t uxTaskPriorityGet( const TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 UBaseType_t uxTaskPriorityGetFromISR( const TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * eTaskState eTaskGetState( TaskHandle_t xTask );
@@ -1149,7 +1163,7 @@ UBaseType_t uxTaskPriorityGetFromISR( const TaskHandle_t xTask ) PRIVILEGED_FUNC
 eTaskState eTaskGetState( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskGetInfo( TaskHandle_t xTask, TaskStatus_t *pxTaskStatus, BaseType_t xGetFreeStackSpace, eTaskState eState );
@@ -1203,7 +1217,7 @@ eTaskState eTaskGetState( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
  *                eInvalid ); // Include the task state in xTaskDetails.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskGetInfo vTaskGetInfo
  * @endcond
  * \ingroup TaskCtrl
@@ -1214,7 +1228,7 @@ void vTaskGetInfo( TaskHandle_t xTask,
                    eTaskState eState ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskPrioritySet( TaskHandle_t xTask, UBaseType_t uxNewPriority );
@@ -1254,7 +1268,7 @@ void vTaskGetInfo( TaskHandle_t xTask,
  *   vTaskPrioritySet( NULL, tskIDLE_PRIORITY + 1 );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskPrioritySet vTaskPrioritySet
  * @endcond
  * \ingroup TaskCtrl
@@ -1263,7 +1277,7 @@ void vTaskPrioritySet( TaskHandle_t xTask,
                        UBaseType_t uxNewPriority ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskSuspend( TaskHandle_t xTaskToSuspend );
@@ -1312,7 +1326,7 @@ void vTaskPrioritySet( TaskHandle_t xTask,
  *   // with our handle as the parameter.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskSuspend vTaskSuspend
  * @endcond
  * \ingroup TaskCtrl
@@ -1320,7 +1334,7 @@ void vTaskPrioritySet( TaskHandle_t xTask,
 void vTaskSuspend( TaskHandle_t xTaskToSuspend ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskResume( TaskHandle_t xTaskToResume );
@@ -1367,7 +1381,7 @@ void vTaskSuspend( TaskHandle_t xTaskToSuspend ) PRIVILEGED_FUNCTION;
  *   // time in accordance with its priority within the system.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskResume vTaskResume
  * @endcond
  * \ingroup TaskCtrl
@@ -1375,7 +1389,7 @@ void vTaskSuspend( TaskHandle_t xTaskToSuspend ) PRIVILEGED_FUNCTION;
 void vTaskResume( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void xTaskResumeFromISR( TaskHandle_t xTaskToResume );
@@ -1402,7 +1416,7 @@ void vTaskResume( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
  * otherwise pdFALSE. This is used by the ISR to determine if a context switch
  * may be required following the ISR.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskResumeFromISR vTaskResumeFromISR
  * @endcond
  * \ingroup TaskCtrl
@@ -1412,9 +1426,9 @@ BaseType_t xTaskResumeFromISR( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
 /*-----------------------------------------------------------
  * SCHEDULER CONTROL
  *----------------------------------------------------------*/
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskStartScheduler( void );
@@ -1445,7 +1459,7 @@ BaseType_t xTaskResumeFromISR( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskStartScheduler vTaskStartScheduler
  * @endcond
  * \ingroup SchedulerControl
@@ -1453,7 +1467,7 @@ BaseType_t xTaskResumeFromISR( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
 void vTaskStartScheduler( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskEndScheduler( void );
@@ -1507,7 +1521,7 @@ void vTaskStartScheduler( void ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskEndScheduler vTaskEndScheduler
  * @endcond
  * \ingroup SchedulerControl
@@ -1517,7 +1531,7 @@ void vTaskEndScheduler( void ) PRIVILEGED_FUNCTION;
 /** @endcond */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskSuspendAll( void );
@@ -1566,7 +1580,7 @@ void vTaskEndScheduler( void ) PRIVILEGED_FUNCTION;
  *   }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskSuspendAll vTaskSuspendAll
  * @endcond
  * \ingroup SchedulerControl
@@ -1574,7 +1588,7 @@ void vTaskEndScheduler( void ) PRIVILEGED_FUNCTION;
 void vTaskSuspendAll( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskResumeAll( void );
@@ -1626,7 +1640,7 @@ void vTaskSuspendAll( void ) PRIVILEGED_FUNCTION;
  *   }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskResumeAll xTaskResumeAll
  * @endcond
  * \ingroup SchedulerControl
@@ -1638,7 +1652,7 @@ BaseType_t xTaskResumeAll( void ) PRIVILEGED_FUNCTION;
  *----------------------------------------------------------*/
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * TickType_t xTaskGetTickCount( void );
@@ -1647,7 +1661,7 @@ BaseType_t xTaskResumeAll( void ) PRIVILEGED_FUNCTION;
  *
  * @return The count of ticks since vTaskStartScheduler was called.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskGetTickCount xTaskGetTickCount
  * @endcond
  * \ingroup TaskUtils
@@ -1655,7 +1669,7 @@ BaseType_t xTaskResumeAll( void ) PRIVILEGED_FUNCTION;
 TickType_t xTaskGetTickCount( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * TickType_t xTaskGetTickCountFromISR( void );
@@ -1669,7 +1683,7 @@ TickType_t xTaskGetTickCount( void ) PRIVILEGED_FUNCTION;
  * microcontroller being used or interrupt nesting is either not supported or
  * not being used.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskGetTickCountFromISR xTaskGetTickCountFromISR
  * @endcond
  * \ingroup TaskUtils
@@ -1677,7 +1691,7 @@ TickType_t xTaskGetTickCount( void ) PRIVILEGED_FUNCTION;
 TickType_t xTaskGetTickCountFromISR( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * uint16_t uxTaskGetNumberOfTasks( void );
@@ -1689,7 +1703,7 @@ TickType_t xTaskGetTickCountFromISR( void ) PRIVILEGED_FUNCTION;
  * has been deleted but not yet freed by the idle task will also be
  * included in the count.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup uxTaskGetNumberOfTasks uxTaskGetNumberOfTasks
  * @endcond
  * \ingroup TaskUtils
@@ -1697,7 +1711,7 @@ TickType_t xTaskGetTickCountFromISR( void ) PRIVILEGED_FUNCTION;
 UBaseType_t uxTaskGetNumberOfTasks( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * char *pcTaskGetName( TaskHandle_t xTaskToQuery );
@@ -1708,7 +1722,7 @@ UBaseType_t uxTaskGetNumberOfTasks( void ) PRIVILEGED_FUNCTION;
  * xTaskToQuery.  A task can query its own name by either passing in its own
  * handle, or by setting xTaskToQuery to NULL.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup pcTaskGetName pcTaskGetName
  * @endcond
  * \ingroup TaskUtils
@@ -1716,7 +1730,7 @@ UBaseType_t uxTaskGetNumberOfTasks( void ) PRIVILEGED_FUNCTION;
 char * pcTaskGetName( TaskHandle_t xTaskToQuery ) PRIVILEGED_FUNCTION;     /*lint !e971 Unqualified char types are allowed for strings and single characters only. */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * TaskHandle_t xTaskGetHandle( const char *pcNameToQuery );
@@ -1730,7 +1744,7 @@ char * pcTaskGetName( TaskHandle_t xTaskToQuery ) PRIVILEGED_FUNCTION;     /*lin
  * NULL is returned if no matching name is found.  INCLUDE_xTaskGetHandle
  * must be set to 1 in FreeRTOSConfig.h for pcTaskGetHandle() to be available.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup pcTaskGetHandle pcTaskGetHandle
  * @endcond
  * \ingroup TaskUtils
@@ -1813,7 +1827,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 #ifdef configUSE_APPLICATION_TASK_TAG
     #if configUSE_APPLICATION_TASK_TAG == 1
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * void vTaskSetApplicationTaskTag( TaskHandle_t xTask, TaskHookFunction_t pxHookFunction );
@@ -1830,7 +1844,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
                                          TaskHookFunction_t pxHookFunction ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * void xTaskGetApplicationTaskTag( TaskHandle_t xTask );
@@ -1844,7 +1858,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
         TaskHookFunction_t xTaskGetApplicationTaskTag( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * void xTaskGetApplicationTaskTagFromISR( TaskHandle_t xTask );
@@ -1932,7 +1946,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 #if ( configCHECK_FOR_STACK_OVERFLOW > 0 )
 
      /**
-      * @cond
+      * @cond !DOC_EXCLUDE_HEADER_SECTION
       * task.h
       * @code{c}
       * void vApplicationStackOverflowHook( TaskHandle_t xTask char *pcTaskName);
@@ -1952,7 +1966,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 
 #if  (  configUSE_TICK_HOOK > 0 )
     /**
-     * @cond
+     * @cond !DOC_EXCLUDE_HEADER_SECTION
      *  task.h
      * @code{c}
      * void vApplicationTickHook( void );
@@ -1967,7 +1981,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 
 #if ( configSUPPORT_STATIC_ALLOCATION == 1 )
     /**
-     * @cond
+     * @cond !DOC_EXCLUDE_HEADER_SECTION
      * task.h
      * @code{c}
      * void vApplicationGetIdleTaskMemory( StaticTask_t ** ppxIdleTaskTCBBuffer, StackType_t ** ppxIdleTaskStackBuffer, uint32_t *pulIdleTaskStackSize )
@@ -1986,7 +2000,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * BaseType_t xTaskCallApplicationTaskHook( TaskHandle_t xTask, void *pvParameter );
@@ -2155,7 +2169,7 @@ UBaseType_t uxTaskGetSystemState( TaskStatus_t * const pxTaskStatusArray,
  * enough to contain the generated report.  Approximately 40 bytes per
  * task should be sufficient.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskList vTaskList
  * @endcond
  * \ingroup TaskUtils
@@ -2210,7 +2224,7 @@ void vTaskList( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lint !e971 Unq
  * contain the generated report.  Approximately 40 bytes per task should
  * be sufficient.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskGetRunTimeStats vTaskGetRunTimeStats
  * @endcond
  * \ingroup TaskUtils
@@ -2218,7 +2232,7 @@ void vTaskList( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lint !e971 Unq
 void vTaskGetRunTimeStats( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lint !e971 Unqualified char types are allowed for strings and single characters only. */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code
  * uint32_t ulTaskGetIdleRunTimeCounter( void );
@@ -2246,7 +2260,7 @@ void vTaskGetRunTimeStats( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lin
  * frequency configured using the portCONFIGURE_TIMER_FOR_RUN_TIME_STATS() and
  * portGET_RUN_TIME_COUNTER_VALUE() macros.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup ulTaskGetIdleRunTimeCounter ulTaskGetIdleRunTimeCounter
  * @endcond
  * \ingroup TaskUtils
@@ -2254,11 +2268,13 @@ void vTaskGetRunTimeStats( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lin
 uint32_t ulTaskGetIdleRunTimeCounter( void ) PRIVILEGED_FUNCTION;
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyIndexed( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify, uint32_t ulValue, eNotifyAction eAction );
  * BaseType_t xTaskNotify( TaskHandle_t xTaskToNotify, uint32_t ulValue, eNotifyAction eAction );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2359,7 +2375,9 @@ uint32_t ulTaskGetIdleRunTimeCounter( void ) PRIVILEGED_FUNCTION;
  * @return Dependent on the value of eAction.  See the description of the
  * eAction parameter.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyIndexed xTaskNotifyIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
@@ -2373,11 +2391,13 @@ BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
     xTaskGenericNotify( ( xTaskToNotify ), ( uxIndexToNotify ), ( ulValue ), ( eAction ), NULL )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyAndQueryIndexed( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify, uint32_t ulValue, eNotifyAction eAction, uint32_t *pulPreviousNotifyValue );
  * BaseType_t xTaskNotifyAndQuery( TaskHandle_t xTaskToNotify, uint32_t ulValue, eNotifyAction eAction, uint32_t *pulPreviousNotifyValue );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2393,7 +2413,9 @@ BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
  * than when the function returns) in the additional pulPreviousNotifyValue
  * parameter.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyAndQueryIndexed xTaskNotifyAndQueryIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 #define xTaskNotifyAndQuery( xTaskToNotify, ulValue, eAction, pulPreviousNotifyValue ) \
@@ -2402,11 +2424,13 @@ BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
     xTaskGenericNotify( ( xTaskToNotify ), ( uxIndexToNotify ), ( ulValue ), ( eAction ), ( pulPreviousNotifyValue ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyIndexedFromISR( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify, uint32_t ulValue, eNotifyAction eAction, BaseType_t *pxHigherPriorityTaskWoken );
  * BaseType_t xTaskNotifyFromISR( TaskHandle_t xTaskToNotify, uint32_t ulValue, eNotifyAction eAction, BaseType_t *pxHigherPriorityTaskWoken );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2511,7 +2535,9 @@ BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
  * @return Dependent on the value of eAction.  See the description of the
  * eAction parameter.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyIndexedFromISR xTaskNotifyIndexedFromISR
+ * @endcond
  * \ingroup TaskNotifications
  */
 BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
@@ -2526,11 +2552,13 @@ BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
     xTaskGenericNotifyFromISR( ( xTaskToNotify ), ( uxIndexToNotify ), ( ulValue ), ( eAction ), NULL, ( pxHigherPriorityTaskWoken ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyAndQueryIndexedFromISR( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify, uint32_t ulValue, eNotifyAction eAction, uint32_t *pulPreviousNotificationValue, BaseType_t *pxHigherPriorityTaskWoken );
  * BaseType_t xTaskNotifyAndQueryFromISR( TaskHandle_t xTaskToNotify, uint32_t ulValue, eNotifyAction eAction, uint32_t *pulPreviousNotificationValue, BaseType_t *pxHigherPriorityTaskWoken );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2546,7 +2574,9 @@ BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
  * function is called rather than at the time the function returns) in the
  * additional pulPreviousNotifyValue parameter.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyAndQueryIndexedFromISR xTaskNotifyAndQueryIndexedFromISR
+ * @endcond
  * \ingroup TaskNotifications
  */
 #define xTaskNotifyAndQueryIndexedFromISR( xTaskToNotify, uxIndexToNotify, ulValue, eAction, pulPreviousNotificationValue, pxHigherPriorityTaskWoken ) \
@@ -2555,12 +2585,14 @@ BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
     xTaskGenericNotifyFromISR( ( xTaskToNotify ), ( tskDEFAULT_INDEX_TO_NOTIFY ), ( ulValue ), ( eAction ), ( pulPreviousNotificationValue ), ( pxHigherPriorityTaskWoken ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyWaitIndexed( UBaseType_t uxIndexToWaitOn, uint32_t ulBitsToClearOnEntry, uint32_t ulBitsToClearOnExit, uint32_t *pulNotificationValue, TickType_t xTicksToWait );
  *
  * BaseType_t xTaskNotifyWait( uint32_t ulBitsToClearOnEntry, uint32_t ulBitsToClearOnExit, uint32_t *pulNotificationValue, TickType_t xTicksToWait );
  * @endcode
+ * @endcond
  *
  * Waits for a direct to task notification to be pending at a given index within
  * an array of direct to task notifications.
@@ -2655,7 +2687,9 @@ BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
  * already pending when xTaskNotifyWait was called) then pdPASS is
  * returned.  Otherwise pdFAIL is returned.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyWaitIndexed xTaskNotifyWaitIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
@@ -2669,11 +2703,13 @@ BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
     xTaskGenericNotifyWait( ( uxIndexToWaitOn ), ( ulBitsToClearOnEntry ), ( ulBitsToClearOnExit ), ( pulNotificationValue ), ( xTicksToWait ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyGiveIndexed( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify );
  * BaseType_t xTaskNotifyGive( TaskHandle_t xTaskToNotify );
  * @endcode
+ * @endcond
  *
  * Sends a direct to task notification to a particular index in the target
  * task's notification array in a manner similar to giving a counting semaphore.
@@ -2737,7 +2773,9 @@ BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
  * @return xTaskNotifyGive() is a macro that calls xTaskNotify() with the
  * eAction parameter set to eIncrement - so pdPASS is always returned.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyGiveIndexed xTaskNotifyGiveIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 #define xTaskNotifyGive( xTaskToNotify ) \
@@ -2746,11 +2784,13 @@ BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
     xTaskGenericNotify( ( xTaskToNotify ), ( uxIndexToNotify ), ( 0 ), eIncrement, NULL )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskNotifyGiveIndexedFromISR( TaskHandle_t xTaskHandle, UBaseType_t uxIndexToNotify, BaseType_t *pxHigherPriorityTaskWoken );
  * void vTaskNotifyGiveFromISR( TaskHandle_t xTaskHandle, BaseType_t *pxHigherPriorityTaskWoken );
  * @endcode
+ * @endcond
  *
  * A version of xTaskNotifyGiveIndexed() that can be called from an interrupt
  * service routine (ISR).
@@ -2821,7 +2861,9 @@ BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
  * requested from an ISR is dependent on the port - see the documentation page
  * for the port in use.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskNotifyGiveIndexedFromISR vTaskNotifyGiveIndexedFromISR
+ * @endcond
  * \ingroup TaskNotifications
  */
 void vTaskGenericNotifyGiveFromISR( TaskHandle_t xTaskToNotify,
@@ -2833,12 +2875,14 @@ void vTaskGenericNotifyGiveFromISR( TaskHandle_t xTaskToNotify,
     vTaskGenericNotifyGiveFromISR( ( xTaskToNotify ), ( uxIndexToNotify ), ( pxHigherPriorityTaskWoken ) );
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * uint32_t ulTaskNotifyTakeIndexed( UBaseType_t uxIndexToWaitOn, BaseType_t xClearCountOnExit, TickType_t xTicksToWait );
  *
  * uint32_t ulTaskNotifyTake( BaseType_t xClearCountOnExit, TickType_t xTicksToWait );
  * @endcode
+ * @endcond
  *
  * Waits for a direct to task notification on a particular index in the calling
  * task's notification array in a manner similar to taking a counting semaphore.
@@ -2927,7 +2971,9 @@ void vTaskGenericNotifyGiveFromISR( TaskHandle_t xTaskToNotify,
  * @return The task's notification count before it is either cleared to zero or
  * decremented (see the xClearCountOnExit parameter).
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup ulTaskNotifyTakeIndexed ulTaskNotifyTakeIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 uint32_t ulTaskGenericNotifyTake( UBaseType_t uxIndexToWaitOn,
@@ -2939,12 +2985,14 @@ uint32_t ulTaskGenericNotifyTake( UBaseType_t uxIndexToWaitOn,
     ulTaskGenericNotifyTake( ( uxIndexToWaitOn ), ( xClearCountOnExit ), ( xTicksToWait ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyStateClearIndexed( TaskHandle_t xTask, UBaseType_t uxIndexToCLear );
  *
  * BaseType_t xTaskNotifyStateClear( TaskHandle_t xTask );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2992,7 +3040,9 @@ uint32_t ulTaskGenericNotifyTake( UBaseType_t uxIndexToWaitOn,
  * @return pdTRUE if the task's notification state was set to
  * eNotWaitingNotification, otherwise pdFALSE.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyStateClearIndexed xTaskNotifyStateClearIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 BaseType_t xTaskGenericNotifyStateClear( TaskHandle_t xTask,
@@ -3003,12 +3053,14 @@ BaseType_t xTaskGenericNotifyStateClear( TaskHandle_t xTask,
     xTaskGenericNotifyStateClear( ( xTask ), ( uxIndexToClear ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * uint32_t ulTaskNotifyValueClearIndexed( TaskHandle_t xTask, UBaseType_t uxIndexToClear, uint32_t ulBitsToClear );
  *
  * uint32_t ulTaskNotifyValueClear( TaskHandle_t xTask, uint32_t ulBitsToClear );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -3057,7 +3109,9 @@ BaseType_t xTaskGenericNotifyStateClear( TaskHandle_t xTask,
  *
  * @return The value of the target task's notification value before the bits
  * specified by ulBitsToClear were cleared.
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup ulTaskNotifyValueClear ulTaskNotifyValueClear
+ * @endcond
  * \ingroup TaskNotifications
  */
 uint32_t ulTaskGenericNotifyValueClear( TaskHandle_t xTask,
@@ -3069,7 +3123,7 @@ uint32_t ulTaskGenericNotifyValueClear( TaskHandle_t xTask,
     ulTaskGenericNotifyValueClear( ( xTask ), ( uxIndexToClear ), ( ulBitsToClear ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * void vTaskSetTimeOutState( TimeOut_t * const pxTimeOut );
@@ -3082,14 +3136,14 @@ uint32_t ulTaskGenericNotifyValueClear( TaskHandle_t xTask,
  * is to be captured.  The captured time includes the tick count and the number
  * of times the tick count has overflowed since the system first booted.
  * \defgroup vTaskSetTimeOutState vTaskSetTimeOutState
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \ingroup TaskCtrl
  * @endcond
  */
 void vTaskSetTimeOutState( TimeOut_t * const pxTimeOut ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code
  * BaseType_t xTaskCheckForTimeOut( TimeOut_t * const pxTimeOut, TickType_t * const pxTicksToWait );
@@ -3170,7 +3224,7 @@ void vTaskSetTimeOutState( TimeOut_t * const pxTimeOut ) PRIVILEGED_FUNCTION;
  *      return uxReceived;
  *  }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCheckForTimeOut xTaskCheckForTimeOut
  * @endcond
  * \ingroup TaskCtrl
@@ -3179,7 +3233,7 @@ BaseType_t xTaskCheckForTimeOut( TimeOut_t * const pxTimeOut,
                                  TickType_t * const pxTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * BaseType_t xTaskCatchUpTicks( TickType_t xTicksToCatchUp );
@@ -3204,7 +3258,7 @@ BaseType_t xTaskCheckForTimeOut( TimeOut_t * const pxTimeOut,
  * blocked state and a context switch being performed.  Otherwise pdFALSE.
  *
  * \defgroup xTaskCatchUpTicks xTaskCatchUpTicks
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \ingroup TaskCtrl
  * @endcond
  */
@@ -3214,7 +3268,7 @@ BaseType_t xTaskCatchUpTicks( TickType_t xTicksToCatchUp ) PRIVILEGED_FUNCTION;
 /*-----------------------------------------------------------
  * SCHEDULER INTERNALS AVAILABLE FOR PORTING PURPOSES
  *----------------------------------------------------------*/
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /*
  * Return the handle of the task running on a certain CPU. Because of
  * the nature of SMP processing, there is no guarantee that this
@@ -3335,8 +3389,8 @@ void vTaskPlaceOnEventListRestricted( List_t * const pxEventList,
  * making the call, otherwise pdFALSE.
  */
 BaseType_t xTaskRemoveFromEventList( const List_t * const pxEventList ) PRIVILEGED_FUNCTION;
-BaseType_t xTaskRemoveFromUnorderedEventList( ListItem_t * pxEventListItem,
-                                              const TickType_t xItemValue ) PRIVILEGED_FUNCTION;
+void vTaskRemoveFromUnorderedEventList( ListItem_t * pxEventListItem,
+                                        const TickType_t xItemValue ) PRIVILEGED_FUNCTION;
 
 /*
  * THIS FUNCTION MUST NOT BE USED FROM APPLICATION CODE.  IT IS ONLY
@@ -3399,11 +3453,6 @@ void vTaskPriorityDisinheritAfterTimeout( TaskHandle_t const pxMutexHolder,
  */
 UBaseType_t uxTaskGetTaskNumber( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
-/*
- * Get the current core affinity of a task
- */
-BaseType_t xTaskGetAffinity( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
-
 /*
  * Set the uxTaskNumber of the task referenced by the xTask parameter to
  * uxHandle.
diff --git a/tools/sdk/esp32c3/include/freertos/include/freertos/timers.h b/tools/sdk/esp32c3/include/freertos/include/freertos/timers.h
index a8bc4f38c78..af6dcb23501 100644
--- a/tools/sdk/esp32c3/include/freertos/include/freertos/timers.h
+++ b/tools/sdk/esp32c3/include/freertos/include/freertos/timers.h
@@ -450,7 +450,7 @@ void vTimerSetTimerID( TimerHandle_t xTimer,
 BaseType_t xTimerIsTimerActive( TimerHandle_t xTimer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * TaskHandle_t xTimerGetTimerDaemonTaskHandle( void );
  * @endcond
  *
@@ -1315,7 +1315,7 @@ TickType_t xTimerGetPeriod( TimerHandle_t xTimer ) PRIVILEGED_FUNCTION;
  */
 TickType_t xTimerGetExpiryTime( TimerHandle_t xTimer ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 
 /*
  * Functions beyond this part are not part of the public API and are intended
@@ -1339,7 +1339,7 @@ BaseType_t xTimerGenericCommand( TimerHandle_t xTimer,
 #if ( configSUPPORT_STATIC_ALLOCATION == 1 )
 
     /**
-     * @cond
+     * @cond !DOC_EXCLUDE_HEADER_SECTION
      * task.h
      * @code{c}
      * void vApplicationGetTimerTaskMemory( StaticTask_t ** ppxTimerTaskTCBBuffer, StackType_t ** ppxTimerTaskStackBuffer, uint32_t *pulTimerTaskStackSize )
diff --git a/tools/sdk/esp32c3/include/freertos/port/riscv/include/freertos/portmacro.h b/tools/sdk/esp32c3/include/freertos/port/riscv/include/freertos/portmacro.h
index cb58b422658..8c09ba7df2f 100644
--- a/tools/sdk/esp32c3/include/freertos/port/riscv/include/freertos/portmacro.h
+++ b/tools/sdk/esp32c3/include/freertos/port/riscv/include/freertos/portmacro.h
@@ -24,178 +24,419 @@
  *
  * 1 tab == 4 spaces!
  */
+
 #ifndef PORTMACRO_H
 #define PORTMACRO_H
 
-#ifdef __cplusplus
-extern "C" {
-#endif
-
 #ifndef __ASSEMBLER__
 
+#include "sdkconfig.h"
 #include <stdint.h>
 #include <stdlib.h>
 #include <stdbool.h>
 #include <stdio.h>
-#include <limits.h>
-#include "esp_timer.h"              /* required for FreeRTOS run time stats */
-
-#include "sdkconfig.h"
+#include "soc/spinlock.h"
+#include "soc/interrupt_core0_reg.h"
+#include "soc/cpu.h"
 #include "esp_attr.h"
+#include "esp_rom_sys.h"
+#include "esp_timer.h"              /* required for FreeRTOS run time stats */
 #include "esp_heap_caps.h"
+#include "esp_system.h"             /* required by esp_get_...() functions in portable.h. [refactor-todo] Update portable.h */
+#include "esp_newlib.h"
+#include "portbenchmark.h"
+
+/* [refactor-todo] These includes are not directly used in this file. They are kept into to prevent a breaking change. Remove these. */
+#include <limits.h>
 #ifdef CONFIG_LEGACY_INCLUDE_COMMON_HEADERS
 #include "soc/soc_memory_layout.h"
 #endif
-#include "soc/spinlock.h"
-#include "soc/interrupt_core0_reg.h"
-#include "esp_rom_sys.h"
-#include "soc/cpu.h"
-#include "esp_system.h"
-#include "esp_newlib.h"
 
-/*-----------------------------------------------------------
- * Port specific definitions.
- *
- * The settings in this file configure FreeRTOS correctly for the
- * given hardware and compiler.
- *
- * These settings should not be altered.
- *-----------------------------------------------------------
- */
+#ifdef __cplusplus
+extern "C" {
+#endif
 
-/* Type definitions. */
-#define portCHAR		uint8_t
-#define portFLOAT		float
-#define portDOUBLE		double
-#define portLONG		int32_t
-#define portSHORT		int16_t
-#define portSTACK_TYPE	uint8_t
-#define portBASE_TYPE	int
-// interrupt module will mask interrupt with priority less than threshold
-#define RVHAL_EXCM_LEVEL    4
 
-typedef portSTACK_TYPE			StackType_t;
-typedef portBASE_TYPE			BaseType_t;
-typedef unsigned portBASE_TYPE	UBaseType_t;
 
-#if( configUSE_16_BIT_TICKS == 1 )
-	typedef uint16_t TickType_t;
-	#define portMAX_DELAY ( TickType_t ) 0xffff
+/* --------------------------------------------------- Port Types ------------------------------------------------------
+ * - Port specific types.
+ * - The settings in this file configure FreeRTOS correctly for the given hardware and compiler.
+ * - These settings should not be altered.
+ * - The port types must come first as they are used further down in this file
+ * ------------------------------------------------------------------------------------------------------------------ */
+
+#define portCHAR                    uint8_t
+#define portFLOAT                   float
+#define portDOUBLE                  double
+#define portLONG                    int32_t
+#define portSHORT                   int16_t
+#define portSTACK_TYPE              uint8_t
+#define portBASE_TYPE               int
+
+typedef portSTACK_TYPE              StackType_t;
+typedef portBASE_TYPE               BaseType_t;
+typedef unsigned portBASE_TYPE      UBaseType_t;
+
+#if (configUSE_16_BIT_TICKS == 1)
+typedef uint16_t TickType_t;
+#define portMAX_DELAY (TickType_t)  0xffff
 #else
-	typedef uint32_t TickType_t;
-	#define portMAX_DELAY ( TickType_t ) 0xffffffffUL
+typedef uint32_t TickType_t;
+#define portMAX_DELAY (TickType_t)  0xffffffffUL
 #endif
-/*------------------------------------------------------*/
 
-/* Architecture specifics. */
-#define portSTACK_GROWTH			( -1 )
-#define portTICK_PERIOD_MS			( ( TickType_t ) (1000 / configTICK_RATE_HZ) )
-#define portBYTE_ALIGNMENT			16
-/*-----------------------------------------------------------*/
-#include "portbenchmark.h"
+/* Task function macros as described on the FreeRTOS.org WEB site. */
+#define portTASK_FUNCTION_PROTO(vFunction, pvParameters) void vFunction(void *pvParameters)
+#define portTASK_FUNCTION(vFunction, pvParameters) void vFunction(void *pvParameters)
 
-static inline BaseType_t IRAM_ATTR xPortGetCoreID(void) {
-    return cpu_hal_get_core_id();
-}
+// interrupt module will mask interrupt with priority less than threshold
+#define RVHAL_EXCM_LEVEL            4
 
 
-static inline bool IRAM_ATTR xPortCanYield(void)
-{
-    uint32_t threshold = REG_READ(INTERRUPT_CORE0_CPU_INT_THRESH_REG);
-    /* when enter critical code, freertos will mask threshold to RVHAL_EXCM_LEVEL
-     * and exit critical code, will recover threshold value (1). so threshold <= 1
-     * means not in critical code
-     */
-    return (threshold <= 1);
-}
+/* ----------------------------------------------- Port Configurations -------------------------------------------------
+ * - Configurations values supplied by each port
+ * - Required by FreeRTOS
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-static inline void uxPortCompareSetExtram(volatile uint32_t *addr, uint32_t compare, uint32_t *set)
-{
-#if defined(CONFIG_SPIRAM)
-    compare_and_set_extram(addr, compare, set);
-#endif
-}
+#define portCRITICAL_NESTING_IN_TCB     0
+#define portSTACK_GROWTH                (-1)
+#define portTICK_PERIOD_MS              ((TickType_t) (1000 / configTICK_RATE_HZ))
+#define portBYTE_ALIGNMENT              16
+#define portNOP() __asm volatile        (" nop ")
 
-static inline void __attribute__((always_inline)) uxPortCompareSet(volatile uint32_t *addr, uint32_t compare, uint32_t *set) {
-    compare_and_set_native(addr, compare, set);
-}
 
-#define portCRITICAL_NESTING_IN_TCB 0
 
-/*
- * Send an interrupt to another core in order to make the task running
- * on it yield for a higher-priority task.
- */
-void vPortYieldOtherCore( BaseType_t coreid);
+/* ---------------------------------------------- Forward Declarations -------------------------------------------------
+ * - Forward declarations of all the port functions and macros need to implement the FreeRTOS porting interface
+ * - These must come before definition/declaration of the FreeRTOS porting interface
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-/*
- Callback to set a watchpoint on the end of the stack. Called every context switch to change the stack
- watchpoint around.
- */
-void vPortSetStackWatchpoint( void* pxStackStart );
+// --------------------- Interrupts ------------------------
 
-/*
- * Returns true if the current core is in ISR context; low prio ISR, med prio ISR or timer tick ISR. High prio ISRs
- * aren't detected here, but they normally cannot call C code, so that should not be an issue anyway.
+/**
+ * @brief Checks if the current core is in an ISR context
+ *
+ * - ISR context consist of Low/Mid priority ISR, or time tick ISR
+ * - High priority ISRs aren't detected here, but they normally cannot call C code, so that should not be an issue anyway.
+ *
+ * @note [refactor-todo] Check if this should be inlined
+ * @return
+ *  - pdTRUE if in ISR
+ *  - pdFALSE otherwise
  */
 BaseType_t xPortInIsrContext(void);
 
-/*
- * This function will be called in High prio ISRs. Returns true if the current core was in ISR context
- * before calling into high prio ISR context.
+/**
+ * @brief Check if in ISR context from High priority ISRs
+ *
+ * - Called from High priority ISR
+ * - Checks if the previous context (before high priority interrupt) was in ISR context (meaning low/med priority)
+ *
+ * @note [refactor-todo] Check if this should be inlined
+ * @return
+ *  - pdTRUE if in previous in ISR context
+ *  - pdFALSE otherwise
  */
 BaseType_t xPortInterruptedFromISRContext(void);
 
-/* "mux" data structure (spinlock) */
+/**
+ * @brief Disable interrupts in a nested manner
+ *
+ * - Cleaner solution allows nested interrupts disabling and restoring via local registers or stack.
+ * - They can be called from interrupts too.
+ * - WARNING Only applies to current CPU.
+ *
+ * @note [refactor-todo] Define this as portSET_INTERRUPT_MASK_FROM_ISR() instead
+ * @return unsigned Previous interrupt state
+ */
+static inline unsigned portENTER_CRITICAL_NESTED(void);
+
+/* ---------------------- Spinlocks ------------------------
+ - Spinlocks added to match API with SMP FreeRTOS. Single core RISC-V does not need spin locks
+ - Because single core does not have a primitive spinlock data type, we have to implement one here
+ * @note [refactor-todo] Refactor critical section API so that this is no longer required
+ * ------------------------------------------------------ */
+
+/**
+ * @brief Spinlock object
+ * Owner:
+ *  - Set to 0 if uninitialized
+ *  - Set to portMUX_FREE_VAL when free
+ *  - Set to CORE_ID_REGVAL_PRO or CORE_ID_REGVAL_AP when locked
+ *  - Any other value indicates corruption
+ * Count:
+ *  - 0 if unlocked
+ *  - Recursive count if locked
+ *
+ * @note Not a true spinlock as single core RISC-V does not have atomic compare and set instruction
+ * @note Keep portMUX_INITIALIZER_UNLOCKED in sync with this struct
+ */
 typedef struct {
-	/* owner field values:
-	 * 0				- Uninitialized (invalid)
-	 * portMUX_FREE_VAL - Mux is free, can be locked by either CPU
-	 * CORE_ID_REGVAL_PRO / CORE_ID_REGVAL_APP - Mux is locked to the particular core
-	 *
-	 *
-	 * Any value other than portMUX_FREE_VAL, CORE_ID_REGVAL_PRO, CORE_ID_REGVAL_APP indicates corruption
-	 */
-	uint32_t owner;
-	/* count field:
-	 * If mux is unlocked, count should be zero.
-	 * If mux is locked, count is non-zero & represents the number of recursive locks on the mux.
-	 */
-	uint32_t count;
+    uint32_t owner;
+    uint32_t count;
 #ifdef CONFIG_FREERTOS_PORTMUX_DEBUG
-	const char *lastLockedFn;
-	int lastLockedLine;
+    const char *lastLockedFn;
+    int lastLockedLine;
 #endif
 } portMUX_TYPE;
+/**< Spinlock initializer */
+#ifndef CONFIG_FREERTOS_PORTMUX_DEBUG
+#define portMUX_INITIALIZER_UNLOCKED {                  \
+            .owner = portMUX_FREE_VAL,                      \
+            .count = 0,                                     \
+        }
+#else
+#define portMUX_INITIALIZER_UNLOCKED {                  \
+            .owner = portMUX_FREE_VAL,                      \
+            .count = 0,                                     \
+            .lastLockedFn = "(never locked)",               \
+            .lastLockedLine = -1                            \
+        }
+#endif /* CONFIG_FREERTOS_PORTMUX_DEBUG */
+#define portMUX_FREE_VAL        SPINLOCK_FREE           /**< Spinlock is free. [refactor-todo] check if this is still required */
+#define portMUX_NO_TIMEOUT      SPINLOCK_WAIT_FOREVER   /**< When passed for 'timeout_cycles', spin forever if necessary. [refactor-todo] check if this is still required */
+#define portMUX_TRY_LOCK        SPINLOCK_NO_WAIT        /**< Try to acquire the spinlock a single time only. [refactor-todo] check if this is still required */
+
+/**
+ * @brief Initialize a spinlock
+ *
+ * - Initializes a spinlock that is used by FreeRTOS SMP critical sections
+ *
+ * @note [refactor-todo] We can make this inline or consider making it a macro
+ * @param[in] mux Spinlock
+ */
+void vPortCPUInitializeMutex(portMUX_TYPE *mux);
 
-#define portMUX_FREE_VAL		SPINLOCK_FREE
+/**
+ * @brief Acquire a spinlock
+ *
+ * @note [refactor-todo] check if we still need this
+ * @note [refactor-todo] Check if this should be inlined
+ * @param[in] mux Spinlock
+ */
+void vPortCPUAcquireMutex(portMUX_TYPE *mux);
 
-/* Special constants for vPortCPUAcquireMutexTimeout() */
-#define portMUX_NO_TIMEOUT      SPINLOCK_WAIT_FOREVER  /* When passed for 'timeout_cycles', spin forever if necessary */
-#define portMUX_TRY_LOCK        SPINLOCK_NO_WAIT     /* Try to acquire the spinlock a single time only */
+/**
+ * @brief Acquire a spinlock but with a specified timeout
+ *
+ * @note [refactor-todo] Check if we still need this
+ * @note [refactor-todo] Check if this should be inlined
+ * @note [refactor-todo] Check if this function should be renamed (due to bool return type)
+ * @param[in] mux Spinlock
+ * @param[in] timeout Timeout in number of CPU cycles
+ * @return true Spinlock acquired
+ * @return false Timed out
+ */
+bool vPortCPUAcquireMutexTimeout(portMUX_TYPE *mux, int timeout_cycles);
 
-// Keep this in sync with the portMUX_TYPE struct definition please.
-#ifndef CONFIG_FREERTOS_PORTMUX_DEBUG
-#define portMUX_INITIALIZER_UNLOCKED {					\
-		.owner = portMUX_FREE_VAL,						\
-		.count = 0,										\
-	}
-#else
-#define portMUX_INITIALIZER_UNLOCKED {					\
-		.owner = portMUX_FREE_VAL,						\
-		.count = 0,										\
-		.lastLockedFn = "(never locked)",				\
-		.lastLockedLine = -1							\
-	}
-#endif
+/**
+ * @brief Release a spinlock
+ *
+ * @note [refactor-todo] check if we still need this
+ * @note [refactor-todo] Check if this should be inlined
+ * @param[in] mux Spinlock
+ */
+void vPortCPUReleaseMutex(portMUX_TYPE *mux);
 
-/* Scheduler utilities. */
-extern void vPortYield( void );
-extern void vPortYieldFromISR( void );
+/**
+ * @brief Wrapper for atomic compare-and-set instruction
+ *
+ * @note Isn't a real atomic CAS.
+ * @note [refactor-todo] check if we still need this
+ * @note [refactor-todo] Check if this function should be renamed (due to void return type)
+ *
+ * @param[inout] addr Pointer to target address
+ * @param[in] compare Compare value
+ * @param[inout] set Pointer to set value
+ */
+static inline void __attribute__((always_inline)) uxPortCompareSet(volatile uint32_t *addr, uint32_t compare, uint32_t *set);
 
-#define portYIELD()			    vPortYield()
-#define portYIELD_FROM_ISR()    vPortYieldFromISR()
+/**
+ * @brief Wrapper for atomic compare-and-set instruction in external RAM
+ *
+ * @note Isn't a real atomic CAS.
+ * @note [refactor-todo] check if we still need this
+ * @note [refactor-todo] Check if this function should be renamed (due to void return type)
+ *
+ * @param[inout] addr Pointer to target address
+ * @param[in] compare Compare value
+ * @param[inout] set Pointer to set value
+ */
+static inline void uxPortCompareSetExtram(volatile uint32_t *addr, uint32_t compare, uint32_t *set);
+
+// ------------------ Critical Sections --------------------
+
+/**
+ * @brief Enter a critical section
+ *
+ * - Simply disable interrupts
+ * - Can be nested
+ */
+void vPortEnterCritical(void);
+
+/**
+ * @brief Exit a critical section
+ *
+ * - Reenables interrupts
+ * - Can be nested
+ */
+void vPortExitCritical(void);
+
+// ---------------------- Yielding -------------------------
 
+/**
+ * @brief Set interrupt mask and return current interrupt enable register
+ *
+ * @note [refactor-todo] Check if this function should be renamed (due to int return type)
+ * @return int Current interrupt enable register before set
+ */
+int vPortSetInterruptMask(void);
+
+/**
+ * @brief Clear current interrupt mask and set given mask
+ *
+ * @param mask Interrupt mask
+ */
+void vPortClearInterruptMask(int mask);
+
+/**
+ * @brief Perform a context switch from a task
+ *
+ * @note [refactor-todo] The rest of ESP-IDF should call taskYield() instead
+ */
+void vPortYield(void);
+
+/**
+ * @brief Perform a context switch from an ISR
+ */
+void vPortYieldFromISR(void);
+
+/**
+ * @brief Yields the other core
+ *
+ * @note Added to be compatible with SMP API
+ * @note [refactor-todo] Put this into private macros as its only called from task.c and is not public API
+ * @param coreid ID of core to yield
+ */
+void vPortYieldOtherCore(BaseType_t coreid);
+
+/**
+ * @brief Checks if the current core can yield
+ *
+ * - A core cannot yield if its in an ISR or in a critical section
+ *
+ * @note [refactor-todo] See if this can be separated from port macro
+ * @note [refactor-todo] Check if this function should be renamed (due to bool return type)
+ * @return true Core can yield
+ * @return false Core cannot yield
+ */
+static inline bool IRAM_ATTR xPortCanYield(void);
+
+// ------------------- Hook Functions ----------------------
+
+extern void esp_vApplicationIdleHook(void);
+extern void esp_vApplicationTickHook(void);
+
+/**
+ * @brief Hook function called on entry to tickless idle
+ *
+ * - Implemented in pm_impl.c
+ *
+ * @param xExpectedIdleTime Expected idle time
+ */
+void vApplicationSleep(TickType_t xExpectedIdleTime);
+
+// ----------------------- System --------------------------
+
+/**
+ * @brief Get the tick rate per second
+ *
+ * @note [refactor-todo] make this inline
+ * @note [refactor-todo] Check if this function should be renamed (due to uint return type)
+ * @return uint32_t Tick rate in Hz
+ */
+uint32_t xPortGetTickRateHz(void);
+
+/**
+ * @brief Set a watchpoint to watch the last 32 bytes of the stack
+ *
+ * Callback to set a watchpoint on the end of the stack. Called every context switch to change the stack watchpoint
+ * around.
+ *
+ * @param pxStackStart Pointer to the start of the stack
+ */
+void vPortSetStackWatchpoint(void *pxStackStart);
+
+/**
+ * @brief Get the current core's ID
+ *
+ * @note Added to be compatible with SMP API
+ * @note [refactor-todo] IDF should call a FreeRTOS like macro instead of port function directly
+ * @return BaseType_t Core ID
+ */
+static inline BaseType_t IRAM_ATTR xPortGetCoreID(void)
+{
+    return (uint32_t) cpu_hal_get_core_id();
+}
+
+
+
+/* ------------------------------------------- FreeRTOS Porting Interface ----------------------------------------------
+ * - Contains all the mappings of the macros required by FreeRTOS
+ * - Most come after forward declare as porting macros map to declared functions
+ * - Maps to forward declared functions
+ * ------------------------------------------------------------------------------------------------------------------ */
+
+// ----------------------- Memory --------------------------
+
+/**
+ * @brief Task memory allocation macros
+ *
+ * @note Because the ROM routines don't necessarily handle a stack in external RAM correctly, we force the stack
+ * memory to always be internal.
+ * @note [refactor-todo] Update portable.h to match v10.4.3 to use new malloc prototypes
+ */
+#define portTcbMemoryCaps               (MALLOC_CAP_INTERNAL|MALLOC_CAP_8BIT)
+#define portStackMemoryCaps             (MALLOC_CAP_INTERNAL|MALLOC_CAP_8BIT)
+#define pvPortMallocTcbMem(size)        pvPortMalloc(size)
+#define pvPortMallocStackMem(size)      pvPortMalloc(size)
+
+// --------------------- Interrupts ------------------------
+
+#define portEXIT_CRITICAL_NESTED(state)     do { portCLEAR_INTERRUPT_MASK_FROM_ISR(state);} while(0);
+#define portDISABLE_INTERRUPTS()            portSET_INTERRUPT_MASK_FROM_ISR()
+#define portENABLE_INTERRUPTS()             portCLEAR_INTERRUPT_MASK_FROM_ISR(1)
+#define portSET_INTERRUPT_MASK_FROM_ISR()                       vPortSetInterruptMask()
+#define portCLEAR_INTERRUPT_MASK_FROM_ISR(uxSavedStatusValue) vPortClearInterruptMask(uxSavedStatusValue)
+
+// ------------------ Critical Sections --------------------
+
+#define portENTER_CRITICAL(mux)         {(void)mux;  vPortEnterCritical();}
+#define portEXIT_CRITICAL(mux)          {(void)mux;  vPortExitCritical();}
+//In single-core RISC-V, we can use the same critical section API
+#define portENTER_CRITICAL_ISR(mux)     portENTER_CRITICAL(mux)
+#define portEXIT_CRITICAL_ISR(mux)      portEXIT_CRITICAL(mux)
+/* [refactor-todo] on RISC-V, both ISR and non-ISR cases result in the same call. We can redefine this macro */
+#define portENTER_CRITICAL_SAFE(mux)    ({  \
+    if (xPortInIsrContext()) {              \
+        portENTER_CRITICAL_ISR(mux);        \
+    } else {                                \
+        portENTER_CRITICAL(mux);            \
+    }                                       \
+})
+#define portEXIT_CRITICAL_SAFE(mux)     ({  \
+    if (xPortInIsrContext()) {              \
+        portEXIT_CRITICAL_ISR(mux);         \
+    } else {                                \
+        portEXIT_CRITICAL(mux);             \
+    }                                       \
+})
+
+// ---------------------- Yielding -------------------------
+
+#define portYIELD() vPortYield()
+#define portYIELD_FROM_ISR() vPortYieldFromISR()
+#define portEND_SWITCHING_ISR(xSwitchRequired) if(xSwitchRequired) vPortYield()
 /* Yielding within an API call (when interrupts are off), means the yield should be delayed
    until interrupts are re-enabled.
    To do this, we use the "cross-core" interrupt as a trigger to yield on this core when interrupts are re-enabled.This
@@ -203,112 +444,104 @@ extern void vPortYieldFromISR( void );
    happening on the same CPU.
 */
 #define portYIELD_WITHIN_API() portYIELD()
-/*-----------------------------------------------------------*/
-
-/* Critical section management. */
-extern int vPortSetInterruptMask(void);
-extern void vPortClearInterruptMask( int );
-
-void vPortCPUInitializeMutex(portMUX_TYPE *mux);
-void vPortCPUAcquireMutex(portMUX_TYPE *mux);
-bool vPortCPUAcquireMutexTimeout(portMUX_TYPE *mux, int timeout_cycles);
-void vPortCPUReleaseMutex(portMUX_TYPE *mux);
 
-extern void vPortEnterCritical( void );
-extern void vPortExitCritical( void );
-
-#define portDISABLE_INTERRUPTS()				portSET_INTERRUPT_MASK_FROM_ISR()
-#define portENABLE_INTERRUPTS()					portCLEAR_INTERRUPT_MASK_FROM_ISR(1)
-
-#define portENTER_CRITICAL(mux)		{(void)mux;  vPortEnterCritical();}
-#define portEXIT_CRITICAL(mux)		{(void)mux;  vPortExitCritical();}
-
-#define portENTER_CRITICAL_ISR(mux)    portENTER_CRITICAL(mux)
-#define portEXIT_CRITICAL_ISR(mux)     portEXIT_CRITICAL(mux)
-
-#define portENTER_CRITICAL_SAFE(mux)  do {                                             \
-                                         if (xPortInIsrContext()) {                    \
-                                             portENTER_CRITICAL_ISR(mux);              \
-                                         } else {                                      \
-                                             portENTER_CRITICAL(mux);                  \
-                                         }                                             \
-                                      } while(0)
-
-#define portEXIT_CRITICAL_SAFE(mux)  do {                                              \
-                                         if (xPortInIsrContext()) {                    \
-                                             portEXIT_CRITICAL_ISR(mux);               \
-                                         } else {                                      \
-                                             portEXIT_CRITICAL(mux);                   \
-                                         }                                             \
-                                      } while(0)
-
-/*------------------------------------------------------------*/
-#define portSET_INTERRUPT_MASK_FROM_ISR()       vPortSetInterruptMask()
-#define portCLEAR_INTERRUPT_MASK_FROM_ISR( uxSavedStatusValue )       vPortClearInterruptMask( uxSavedStatusValue )
-#define portEND_SWITCHING_ISR( xSwitchRequired ) if( xSwitchRequired ) vPortYield()
-
-// Cleaner solution allows nested interrupts disabling and restoring via local registers or stack.
-// They can be called from interrupts too.
-static inline unsigned portENTER_CRITICAL_NESTED(void) {
-    unsigned state = portSET_INTERRUPT_MASK_FROM_ISR();
-	return state;
-}
+// ------------------- Hook Functions ----------------------
 
-#define portEXIT_CRITICAL_NESTED(state)   do { portCLEAR_INTERRUPT_MASK_FROM_ISR( state );} while(0);
-/*-----------------------------------------------------------*/
-
-//Because the ROM routines don't necessarily handle a stack in external RAM correctly, we force
-//the stack memory to always be internal.
-#define portTcbMemoryCaps (MALLOC_CAP_INTERNAL|MALLOC_CAP_8BIT)
-#define portStackMemoryCaps (MALLOC_CAP_INTERNAL|MALLOC_CAP_8BIT)
+#ifndef CONFIG_FREERTOS_LEGACY_HOOKS
+#define vApplicationIdleHook    esp_vApplicationIdleHook
+#define vApplicationTickHook    esp_vApplicationTickHook
+#endif /* !CONFIG_FREERTOS_LEGACY_HOOKS */
+#define portSUPPRESS_TICKS_AND_SLEEP(idleTime) vApplicationSleep(idleTime)
 
-#define pvPortMallocTcbMem(size) pvPortMalloc(size)
-#define pvPortMallocStackMem(size)  pvPortMalloc(size)
+// ------------------- Run Time Stats ----------------------
 
-/* Fine resolution time */
-#define portGET_RUN_TIME_COUNTER_VALUE() 0
 #define portCONFIGURE_TIMER_FOR_RUN_TIME_STATS()
-
+#define portGET_RUN_TIME_COUNTER_VALUE() 0
 #ifdef CONFIG_FREERTOS_RUN_TIME_STATS_USING_ESP_TIMER
 /* Coarse resolution time (us) */
 #define portALT_GET_RUN_TIME_COUNTER_VALUE(x)    do {x = (uint32_t)esp_timer_get_time();} while(0)
 #endif
 
-extern void esp_vApplicationIdleHook( void );
-extern void esp_vApplicationTickHook( void );
 
-#ifndef CONFIG_FREERTOS_LEGACY_HOOKS
-#define vApplicationIdleHook    esp_vApplicationIdleHook
-#define vApplicationTickHook    esp_vApplicationTickHook
-#endif /* !CONFIG_FREERTOS_LEGACY_HOOKS */
 
-/* Task function macros as described on the FreeRTOS.org WEB site. */
-#define portTASK_FUNCTION_PROTO( vFunction, pvParameters ) void vFunction( void *pvParameters )
-#define portTASK_FUNCTION( vFunction, pvParameters ) void vFunction( void *pvParameters )
+/* --------------------------------------------- Inline Implementations ------------------------------------------------
+ * - Implementation of inline functions of the forward declares
+ * - Should come after forward declare and FreeRTOS Porting interface, as implementation may use both.
+ * - For implementation of non-inlined functions, see port.c, port_common.c, or other assembly files
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-void vApplicationSleep( TickType_t xExpectedIdleTime );
-#define portSUPPRESS_TICKS_AND_SLEEP( idleTime ) vApplicationSleep( idleTime )
+// --------------------- Interrupts ------------------------
 
-#define portNOP() __asm volatile 	( " nop " )
+static inline unsigned portENTER_CRITICAL_NESTED(void)
+{
+    unsigned state = portSET_INTERRUPT_MASK_FROM_ISR();
+    return state;
+}
 
-#define portVALID_TCB_MEM(ptr) esp_ptr_byte_accessible(ptr)
-#define portVALID_STACK_MEM(ptr) esp_ptr_byte_accessible(ptr)
+// ---------------------- Spinlocks ------------------------
 
-/* Get tick rate per second */
-uint32_t xPortGetTickRateHz(void);
+static inline void __attribute__((always_inline)) uxPortCompareSet(volatile uint32_t *addr, uint32_t compare, uint32_t *set)
+{
+    compare_and_set_native(addr, compare, set);
+}
 
-// configASSERT_2 if requested
-#if configASSERT_2
-#include <stdio.h>
-void exit(int);
-#define configASSERT( x )   if (!(x)) { porttracePrint(-1); printf("\nAssertion failed in %s:%d\n", __FILE__, __LINE__); exit(-1); }
+static inline void uxPortCompareSetExtram(volatile uint32_t *addr, uint32_t compare, uint32_t *set)
+{
+#if defined(CONFIG_SPIRAM)
+    compare_and_set_extram(addr, compare, set);
 #endif
+}
 
+// ---------------------- Yielding -------------------------
 
-#endif //__ASSEMBLER__
+static inline bool IRAM_ATTR xPortCanYield(void)
+{
+    uint32_t threshold = REG_READ(INTERRUPT_CORE0_CPU_INT_THRESH_REG);
+    /* when enter critical code, FreeRTOS will mask threshold to RVHAL_EXCM_LEVEL
+     * and exit critical code, will recover threshold value (1). so threshold <= 1
+     * means not in critical code
+     */
+    return (threshold <= 1);
+}
+
+
+
+/* ------------------------------------------------------ Misc ---------------------------------------------------------
+ * - Miscellaneous porting macros
+ * - These are not port of the FreeRTOS porting interface, but are used by other FreeRTOS dependent components
+ * ------------------------------------------------------------------------------------------------------------------ */
+
+// -------------------- Heap Related -----------------------
+
+/**
+ * @brief Checks if a given piece of memory can be used to store a task's TCB
+ *
+ * - Defined in port_common.c
+ *
+ * @param ptr Pointer to memory
+ * @return true Memory can be used to store a TCB
+ * @return false Otherwise
+ */
+bool xPortCheckValidTCBMem(const void *ptr);
+
+/**
+ * @brief Checks if a given piece of memory can be used to store a task's stack
+ *
+ * - Defined in port_common.c
+ *
+ * @param ptr Pointer to memory
+ * @return true Memory can be used to store a task stack
+ * @return false Otherwise
+ */
+bool xPortcheckValidStackMem(const void *ptr);
+
+#define portVALID_TCB_MEM(ptr) xPortCheckValidTCBMem(ptr)
+#define portVALID_STACK_MEM(ptr) xPortcheckValidStackMem(ptr)
 
 #ifdef __cplusplus
 }
 #endif
 
+#endif //__ASSEMBLER__
+
 #endif /* PORTMACRO_H */
diff --git a/tools/sdk/esp32c3/include/hal/esp32c3/include/hal/i2s_ll.h b/tools/sdk/esp32c3/include/hal/esp32c3/include/hal/i2s_ll.h
index 670db91616b..0cfb05a266a 100644
--- a/tools/sdk/esp32c3/include/hal/esp32c3/include/hal/i2s_ll.h
+++ b/tools/sdk/esp32c3/include/hal/esp32c3/include/hal/i2s_ll.h
@@ -1,16 +1,8 @@
-// Copyright 2021 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 // The LL layer for I2S register operations
 /*******************************************************************************
@@ -86,6 +78,26 @@ static inline void i2s_ll_rx_enable_clock(i2s_dev_t *hw)
     hw->rx_clkm_conf.rx_clk_active = 1;
 }
 
+/**
+ * @brief Disable I2S tx module clock
+ *
+ * @param hw Peripheral I2S hardware instance address.
+ */
+static inline void i2s_ll_tx_disable_clock(i2s_dev_t *hw)
+{
+    hw->tx_clkm_conf.tx_clk_active = 0;
+}
+
+/**
+ * @brief Disable I2S rx module clock
+ *
+ * @param hw Peripheral I2S hardware instance address.
+ */
+static inline void i2s_ll_rx_disable_clock(i2s_dev_t *hw)
+{
+    hw->rx_clkm_conf.rx_clk_active = 0;
+}
+
 /**
  * @brief I2S mclk use tx module clock
  *
diff --git a/tools/sdk/esp32c3/include/hal/esp32c3/include/hal/rmt_ll.h b/tools/sdk/esp32c3/include/hal/esp32c3/include/hal/rmt_ll.h
index 0f703aebc75..662e63f784e 100644
--- a/tools/sdk/esp32c3/include/hal/esp32c3/include/hal/rmt_ll.h
+++ b/tools/sdk/esp32c3/include/hal/esp32c3/include/hal/rmt_ll.h
@@ -1,16 +1,9 @@
-// Copyright 2020 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2020-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
 #pragma once
 
 #include <stdint.h>
@@ -23,6 +16,9 @@
 extern "C" {
 #endif
 
+
+#define RMT_LL_MAX_LOOP_COUNT           (1023)/*!< Max loop count that hardware is supported */
+
 #define RMT_LL_HW_BASE  (&RMT)
 #define RMT_LL_MEM_BASE (&RMTMEM)
 
diff --git a/tools/sdk/esp32c3/include/hal/include/hal/i2s_hal.h b/tools/sdk/esp32c3/include/hal/include/hal/i2s_hal.h
index a08813db808..037970fa2b3 100644
--- a/tools/sdk/esp32c3/include/hal/include/hal/i2s_hal.h
+++ b/tools/sdk/esp32c3/include/hal/include/hal/i2s_hal.h
@@ -1,16 +1,8 @@
-// Copyright 2020 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2020-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 /*******************************************************************************
  * NOTICE
@@ -176,28 +168,28 @@ void i2s_hal_enable_slave_fd_mode(i2s_hal_context_t *hal);
  *
  * @param hal Context of the HAL layer
  */
-#define i2s_hal_start_tx(hal)               i2s_ll_tx_start((hal)->dev)
+void i2s_hal_start_tx(i2s_hal_context_t *hal);
 
 /**
  * @brief Start I2S rx
  *
  * @param hal Context of the HAL layer
  */
-#define i2s_hal_start_rx(hal)               i2s_ll_rx_start((hal)->dev)
+void i2s_hal_start_rx(i2s_hal_context_t *hal);
 
 /**
  * @brief Stop I2S tx
  *
  * @param hal Context of the HAL layer
  */
-#define i2s_hal_stop_tx(hal)                i2s_ll_tx_stop((hal)->dev)
+void i2s_hal_stop_tx(i2s_hal_context_t *hal);
 
 /**
  * @brief Stop I2S rx
  *
  * @param hal Context of the HAL layer
  */
-#define i2s_hal_stop_rx(hal)                i2s_ll_rx_stop((hal)->dev)
+void i2s_hal_stop_rx(i2s_hal_context_t *hal);
 
 /**
  * @brief Set the received data length to trigger `in_suc_eof` interrupt.
diff --git a/tools/sdk/esp32c3/include/hal/include/hal/lcd_hal.h b/tools/sdk/esp32c3/include/hal/include/hal/lcd_hal.h
index 76e28dda0e4..db255b3d1e4 100644
--- a/tools/sdk/esp32c3/include/hal/include/hal/lcd_hal.h
+++ b/tools/sdk/esp32c3/include/hal/include/hal/lcd_hal.h
@@ -1,22 +1,8 @@
-// Copyright 2021 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
-
-/*******************************************************************************
- * NOTICE
- * The HAL is not public api, don't use in application code.
- * See readme.md in soc/README.md
- ******************************************************************************/
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
diff --git a/tools/sdk/esp32c3/include/hal/include/hal/lcd_types.h b/tools/sdk/esp32c3/include/hal/include/hal/lcd_types.h
index 69dab14801d..01e6d0c2949 100644
--- a/tools/sdk/esp32c3/include/hal/include/hal/lcd_types.h
+++ b/tools/sdk/esp32c3/include/hal/include/hal/lcd_types.h
@@ -1,16 +1,8 @@
-// Copyright 2021 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
@@ -21,15 +13,12 @@ extern "C" {
 /**
  * @brief LCD clock source
  * @note User should select the clock source based on the real requirement:
- * ╔═════════════════════╦══════════════════════════╦════════════════════════════╗
- * ║ LCD clock source    ║ Features                 ║ Power Management           ║
- * ╠═════════════════════╬══════════════════════════╬════════════════════════════╣
- * ║ LCD_CLK_SRC_PLL160M ║ High resolution, fixed   ║ ESP_PM_APB_FREQ_MAX lock   ║
- * ╠═════════════════════╬══════════════════════════╬════════════════════════════╣
- * ║ LCD_CLK_SRC_APLL    ║ Configurable resolution  ║ ESP_PM_NO_LIGHT_SLEEP lock ║
- * ╠═════════════════════╬══════════════════════════╬════════════════════════════╣
- * ║ LCD_CLK_SRC_XTAL    ║ Medium resolution, fixed ║ No PM lock                 ║
- * ╚═════════════════════╩══════════════════════════╩════════════════════════════╝
+ *
+ * | LCD clock source    | Features                 | Power Management           |
+ * |---------------------|--------------------------|----------------------------|
+ * | LCD_CLK_SRC_PLL160M | High resolution, fixed   | ESP_PM_APB_FREQ_MAX lock   |
+ * | LCD_CLK_SRC_APLL    | Configurable resolution  | ESP_PM_NO_LIGHT_SLEEP lock |
+ * | LCD_CLK_SRC_XTAL    | Medium resolution, fixed | No PM lock                 |
  */
 typedef enum {
     LCD_CLK_SRC_PLL160M, /*!< Select PLL160M as the source clock */
diff --git a/tools/sdk/esp32c3/include/hal/include/hal/touch_sensor_types.h b/tools/sdk/esp32c3/include/hal/include/hal/touch_sensor_types.h
index ec027bf8705..9085f5eecd8 100644
--- a/tools/sdk/esp32c3/include/hal/include/hal/touch_sensor_types.h
+++ b/tools/sdk/esp32c3/include/hal/include/hal/touch_sensor_types.h
@@ -1,16 +1,8 @@
-// Copyright 2015-2020 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
@@ -155,12 +147,23 @@ typedef enum {
     TOUCH_PAD_INTR_MASK_INACTIVE = BIT(2),  /*!<Inactive for one of the enabled channels. */
     TOUCH_PAD_INTR_MASK_SCAN_DONE = BIT(3), /*!<Measurement done for all the enabled channels. */
     TOUCH_PAD_INTR_MASK_TIMEOUT = BIT(4),   /*!<Timeout for one of the enabled channels. */
+#if SOC_TOUCH_PROXIMITY_MEAS_DONE_SUPPORTED
+    TOUCH_PAD_INTR_MASK_PROXI_MEAS_DONE = BIT(5),   /*!<For proximity sensor, when the number of measurements reaches the set count of measurements, an interrupt will be generated. */
+    TOUCH_PAD_INTR_MASK_MAX
+#define TOUCH_PAD_INTR_MASK_ALL (TOUCH_PAD_INTR_MASK_TIMEOUT    \
+                                | TOUCH_PAD_INTR_MASK_SCAN_DONE \
+                                | TOUCH_PAD_INTR_MASK_INACTIVE  \
+                                | TOUCH_PAD_INTR_MASK_ACTIVE    \
+                                | TOUCH_PAD_INTR_MASK_DONE      \
+                                | TOUCH_PAD_INTR_MASK_PROXI_MEAS_DONE) /*!<All touch interrupt type enable. */
+#else
     TOUCH_PAD_INTR_MASK_MAX
 #define TOUCH_PAD_INTR_MASK_ALL (TOUCH_PAD_INTR_MASK_TIMEOUT    \
                                 | TOUCH_PAD_INTR_MASK_SCAN_DONE \
                                 | TOUCH_PAD_INTR_MASK_INACTIVE  \
                                 | TOUCH_PAD_INTR_MASK_ACTIVE    \
                                 | TOUCH_PAD_INTR_MASK_DONE) /*!<All touch interrupt type enable. */
+#endif
 } touch_pad_intr_mask_t;
 FLAG_ATTR(touch_pad_intr_mask_t)
 
diff --git a/tools/sdk/esp32c3/include/lwip/port/esp32/include/lwipopts.h b/tools/sdk/esp32c3/include/lwip/port/esp32/include/lwipopts.h
index 4bc2ad13960..b9cfe17aca9 100644
--- a/tools/sdk/esp32c3/include/lwip/port/esp32/include/lwipopts.h
+++ b/tools/sdk/esp32c3/include/lwip/port/esp32/include/lwipopts.h
@@ -266,6 +266,12 @@
 
 #endif
 
+/**
+ * CONFIG_LWIP_DHCP_OPTIONS_LEN: The total length of outgoing DHCP option msg. If you have many options
+ * and options value is too long, you can configure the length according to your requirements
+ */
+#define DHCP_OPTIONS_LEN                CONFIG_LWIP_DHCP_OPTIONS_LEN
+
 /*
    ------------------------------------
    ---------- AUTOIP options ----------
diff --git a/tools/sdk/esp32c3/include/mdns/include/mdns.h b/tools/sdk/esp32c3/include/mdns/include/mdns.h
index 4bb288d4174..35c8d538abc 100644
--- a/tools/sdk/esp32c3/include/mdns/include/mdns.h
+++ b/tools/sdk/esp32c3/include/mdns/include/mdns.h
@@ -187,7 +187,10 @@ esp_err_t mdns_instance_name_set(const char * instance_name);
  * @note The value length of txt items will be automatically decided by strlen
  *
  * @param  instance_name    instance name to set. If NULL,
- *                          global instance name or hostname will be used
+ *                          global instance name or hostname will be used.
+ *                          Note that MDNS_MULTIPLE_INSTANCE config option
+ *                          needs to be enabled for adding multiple instances
+ *                          with the same instance type.
  * @param  service_type     service type (_http, _ftp, etc)
  * @param  proto            service protocol (_tcp, _udp)
  * @param  port             service port
@@ -209,6 +212,9 @@ esp_err_t mdns_service_add(const char * instance_name, const char * service_type
  *
  * @param  instance_name    instance name to set. If NULL,
  *                          global instance name or hostname will be used
+ *                          Note that MDNS_MULTIPLE_INSTANCE config option
+ *                          needs to be enabled for adding multiple instances
+ *                          with the same instance type.
  * @param  service_type     service type (_http, _ftp, etc)
  * @param  proto            service protocol (_tcp, _udp)
  * @param  hostname         service hostname. If NULL, local hostname will be used.
@@ -238,6 +244,22 @@ esp_err_t mdns_service_add_for_host(const char * instance_name, const char * ser
  */
 bool mdns_service_exists(const char * service_type, const char * proto, const char * hostname);
 
+
+/**
+ * @brief  Check whether a service has been added.
+ *
+ * @param  instance         instance name
+ * @param  service_type     service type (_http, _ftp, etc)
+ * @param  proto            service protocol (_tcp, _udp)
+ * @param  hostname         service hostname. If NULL, checks for the local hostname.
+ *
+ * @return
+ *     - true   Correspondding service has been added.
+ *     - false  Service not found.
+ */
+bool mdns_service_exists_with_instance(const char *instance, const char *service_type, const char *proto,
+                                       const char *hostname);
+
 /**
  * @brief  Remove service from mDNS server
  *
diff --git a/tools/sdk/esp32c3/include/mqtt/esp-mqtt/include/mqtt_client.h b/tools/sdk/esp32c3/include/mqtt/esp-mqtt/include/mqtt_client.h
index 348d0729f1d..475581b50db 100644
--- a/tools/sdk/esp32c3/include/mqtt/esp-mqtt/include/mqtt_client.h
+++ b/tools/sdk/esp32c3/include/mqtt/esp-mqtt/include/mqtt_client.h
@@ -206,6 +206,7 @@ typedef struct {
     int network_timeout_ms;                 /*!< Abort network operation if it is not completed after this value, in milliseconds (defaults to 10s) */
     bool disable_keepalive;                 /*!< Set disable_keepalive=true to turn off keep-alive mechanism, false by default (keepalive is active by default). Note: setting the config value `keepalive` to `0` doesn't disable keepalive feature, but uses a default keepalive period */
     const char *path;                       /*!< Path in the URI*/
+    int message_retransmit_timeout;         /*!< timeout for retansmit of failded packet */
 } esp_mqtt_client_config_t;
 
 /**
@@ -375,6 +376,7 @@ esp_err_t esp_mqtt_client_destroy(esp_mqtt_client_handle_t client);
  * @param config    mqtt configuration structure
  *
  * @return ESP_ERR_NO_MEM if failed to allocate
+ *         ESP_ERR_INVALID_ARG if conflicts on transport configuration.
  *         ESP_OK on success
  */
 esp_err_t esp_mqtt_set_config(esp_mqtt_client_handle_t client, const esp_mqtt_client_config_t *config);
diff --git a/tools/sdk/esp32c3/include/openssl/include/openssl/ssl.h b/tools/sdk/esp32c3/include/openssl/include/openssl/ssl.h
index 4a3376c0dbf..9a33e0d33b2 100644
--- a/tools/sdk/esp32c3/include/openssl/include/openssl/ssl.h
+++ b/tools/sdk/esp32c3/include/openssl/include/openssl/ssl.h
@@ -1,16 +1,13 @@
-// Copyright 2015-2016 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+// Allow for this warning suppression only in IDF_CI_BUILD
+#if !defined(ESP_OPENSSL_SUPPRESS_LEGACY_WARNING) || !defined(IDF_CI_BUILD)
+#warning "OpenSSL component will be removed from ESP-IDF in v5.0, please use esp_tls instead"
+#endif
 
 #ifndef _SSL_H_
 #define _SSL_H_
diff --git a/tools/sdk/esp32c3/include/protocomm/include/transports/protocomm_ble.h b/tools/sdk/esp32c3/include/protocomm/include/transports/protocomm_ble.h
index a9e7b6a209e..f899f95b188 100644
--- a/tools/sdk/esp32c3/include/protocomm/include/transports/protocomm_ble.h
+++ b/tools/sdk/esp32c3/include/protocomm/include/transports/protocomm_ble.h
@@ -1,16 +1,8 @@
-// Copyright 2018 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2018-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
@@ -86,6 +78,10 @@ typedef struct protocomm_ble_config {
      * Pointer to the Name-UUID lookup table
      */
     protocomm_ble_name_uuid_t *nu_lookup;
+
+    /* BLE bonding */
+    unsigned ble_bonding:1;
+
 } protocomm_ble_config_t;
 
 /**
diff --git a/tools/sdk/esp32c3/include/soc/esp32c3/include/soc/rtc.h b/tools/sdk/esp32c3/include/soc/esp32c3/include/soc/rtc.h
index 40e6ec81d14..db67d3ac517 100644
--- a/tools/sdk/esp32c3/include/soc/esp32c3/include/soc/rtc.h
+++ b/tools/sdk/esp32c3/include/soc/esp32c3/include/soc/rtc.h
@@ -114,11 +114,14 @@ extern "C" {
 set sleep_init default param
 */
 #define RTC_CNTL_DBG_ATTEN_LIGHTSLEEP_DEFAULT  3
+#define RTC_CNTL_DBG_ATTEN_LIGHTSLEEP_NODROP  0
 #define RTC_CNTL_DBG_ATTEN_DEEPSLEEP_DEFAULT  15
 #define RTC_CNTL_DBG_ATTEN_MONITOR_DEFAULT  0
 #define RTC_CNTL_BIASSLP_MONITOR_DEFAULT  0
+#define RTC_CNTL_BIASSLP_SLEEP_ON  0
 #define RTC_CNTL_BIASSLP_SLEEP_DEFAULT  1
 #define RTC_CNTL_PD_CUR_MONITOR_DEFAULT  0
+#define RTC_CNTL_PD_CUR_SLEEP_ON  0
 #define RTC_CNTL_PD_CUR_SLEEP_DEFAULT  1
 #define RTC_CNTL_DG_VDD_DRV_B_SLP_DEFAULT 254
 
@@ -639,7 +642,8 @@ typedef struct {
     uint32_t rtc_peri_pd_en : 1;        //!< power down RTC peripherals
     uint32_t wifi_pd_en : 1;            //!< power down WiFi
     uint32_t bt_pd_en : 1;              //!< power down BT
-    uint32_t cpu_pd_en : 1;              //!< power down CPU, but not restart when lightsleep.
+    uint32_t cpu_pd_en : 1;             //!< power down CPU, but not restart when lightsleep.
+    uint32_t int_8m_pd_en : 1;          //!< Power down Internal 8M oscillator
     uint32_t dig_peri_pd_en : 1;        //!< power down digital peripherals
     uint32_t deep_slp : 1;              //!< power down digital domain
     uint32_t wdt_flashboot_mod_en : 1;  //!< enable WDT flashboot mode
@@ -648,6 +652,7 @@ typedef struct {
     uint32_t rtc_dbias_wak : 5;         //!< set bias for RTC domain, in active mode
     uint32_t rtc_dbias_slp : 5;         //!< set bias for RTC domain, in sleep mode
     uint32_t vddsdio_pd_en : 1;         //!< power down VDDSDIO regulator
+    uint32_t xtal_fpu : 1;              //!< keep main XTAL powered up in sleep
     uint32_t deep_slp_reject : 1;
     uint32_t light_slp_reject : 1;
 } rtc_sleep_config_t;
@@ -660,6 +665,7 @@ typedef struct {
  *
  * @param RTC_SLEEP_PD_x flags combined using bitwise OR
  */
+#define is_dslp(pd_flags)   ((pd_flags) & RTC_SLEEP_PD_DIG)
 #define RTC_SLEEP_CONFIG_DEFAULT(sleep_flags) { \
     .lslp_mem_inf_fpu = 0, \
     .rtc_mem_inf_follow_cpu = ((sleep_flags) & RTC_SLEEP_PD_RTC_MEM_FOLLOW_CPU) ? 1 : 0, \
@@ -669,14 +675,20 @@ typedef struct {
     .wifi_pd_en = ((sleep_flags) & RTC_SLEEP_PD_WIFI) ? 1 : 0, \
     .bt_pd_en = ((sleep_flags) & RTC_SLEEP_PD_BT) ? 1 : 0, \
     .cpu_pd_en = ((sleep_flags) & RTC_SLEEP_PD_CPU) ? 1 : 0, \
+    .int_8m_pd_en = is_dslp(sleep_flags) ? 1 : ((sleep_flags) & RTC_SLEEP_PD_INT_8M) ? 1 : 0, \
     .dig_peri_pd_en = ((sleep_flags) & RTC_SLEEP_PD_DIG_PERIPH) ? 1 : 0, \
     .deep_slp = ((sleep_flags) & RTC_SLEEP_PD_DIG) ? 1 : 0, \
     .wdt_flashboot_mod_en = 0, \
     .dig_dbias_wak = RTC_CNTL_DBIAS_1V10, \
-    .dig_dbias_slp = RTC_CNTL_DBIAS_SLP, \
+    .dig_dbias_slp = is_dslp(sleep_flags)                   ? RTC_CNTL_DBIAS_SLP  \
+                   : !((sleep_flags) & RTC_SLEEP_PD_INT_8M) ? RTC_CNTL_DBIAS_1V10 \
+                   : RTC_CNTL_DBIAS_SLP, \
     .rtc_dbias_wak = RTC_CNTL_DBIAS_1V10, \
-    .rtc_dbias_slp = RTC_CNTL_DBIAS_SLP, \
+    .rtc_dbias_slp = is_dslp(sleep_flags)                   ? RTC_CNTL_DBIAS_SLP  \
+                   : !((sleep_flags) & RTC_SLEEP_PD_INT_8M) ? RTC_CNTL_DBIAS_1V10 \
+                   : RTC_CNTL_DBIAS_SLP, \
     .vddsdio_pd_en = ((sleep_flags) & RTC_SLEEP_PD_VDDSDIO) ? 1 : 0, \
+    .xtal_fpu = is_dslp(sleep_flags) ? 0 : ((sleep_flags) & RTC_SLEEP_PD_XTAL) ? 0 : 1, \
     .deep_slp_reject = 1, \
     .light_slp_reject = 1 \
 };
@@ -691,6 +703,8 @@ typedef struct {
 #define RTC_SLEEP_PD_BT                 BIT(7)  //!< Power down BT
 #define RTC_SLEEP_PD_CPU                BIT(8)  //!< Power down CPU when in lightsleep, but not restart
 #define RTC_SLEEP_PD_DIG_PERIPH         BIT(9)  //!< Power down DIG peripherals
+#define RTC_SLEEP_PD_INT_8M             BIT(10) //!< Power down Internal 8M oscillator
+#define RTC_SLEEP_PD_XTAL               BIT(11) //!< Power down main XTAL
 
 /**
  * @brief Prepare the chip to enter sleep mode
diff --git a/tools/sdk/esp32c3/include/soc/include/soc/lcd_periph.h b/tools/sdk/esp32c3/include/soc/include/soc/lcd_periph.h
index d7d3ebc68bf..7bb7c530cf2 100644
--- a/tools/sdk/esp32c3/include/soc/include/soc/lcd_periph.h
+++ b/tools/sdk/esp32c3/include/soc/include/soc/lcd_periph.h
@@ -1,16 +1,8 @@
-// Copyright 2021 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
diff --git a/tools/sdk/esp32c3/include/spi_flash/include/esp_partition.h b/tools/sdk/esp32c3/include/spi_flash/include/esp_partition.h
index 36baea29860..550bf6b6b26 100644
--- a/tools/sdk/esp32c3/include/spi_flash/include/esp_partition.h
+++ b/tools/sdk/esp32c3/include/spi_flash/include/esp_partition.h
@@ -181,7 +181,9 @@ esp_partition_iterator_t esp_partition_next(esp_partition_iterator_t iterator);
 /**
  * @brief Release partition iterator
  *
- * @param iterator Iterator obtained using esp_partition_find. Must be non-NULL.
+ * @param iterator Iterator obtained using esp_partition_find.
+ *                 The iterator is allowed to be NULL, so it is not necessary to check its value
+ *                 before calling this function.
  *
  */
 void esp_partition_iterator_release(esp_partition_iterator_t iterator);
diff --git a/tools/sdk/esp32/include/spi_flash/include/spi_flash_private.h b/tools/sdk/esp32c3/include/spi_flash/include/esp_private/spi_flash_os.h
similarity index 88%
rename from tools/sdk/esp32/include/spi_flash/include/spi_flash_private.h
rename to tools/sdk/esp32c3/include/spi_flash/include/esp_private/spi_flash_os.h
index 9d822fb9475..e302df86ea2 100644
--- a/tools/sdk/esp32/include/spi_flash/include/spi_flash_private.h
+++ b/tools/sdk/esp32c3/include/spi_flash/include/esp_private/spi_flash_os.h
@@ -4,7 +4,9 @@
  * SPDX-License-Identifier: Apache-2.0
  */
 
-
+/**
+ * System level MSPI APIs (private)
+ */
 /**
  * Currently the MSPI timing tuning related APIs are designed to be private.
  * Because:
@@ -17,6 +19,10 @@
  */
 #pragma once
 
+#include <stdint.h>
+#include <stdbool.h>
+#include "sdkconfig.h"
+#include "esp_err.h"
 #if CONFIG_IDF_TARGET_ESP32
 #include "esp32/rom/spi_flash.h"
 #elif CONFIG_IDF_TARGET_ESP32S2
@@ -34,9 +40,9 @@ extern "C" {
 #endif
 
 /**
- * @brief Register ROM functions and init flash device registers to make use of octal flash
+ * @brief To setup Flash chip
  */
-esp_err_t esp_opiflash_init(void);
+esp_err_t spi_flash_init_chip_state(void);
 
 /**
  * @brief Make MSPI work under 20Mhz
@@ -88,6 +94,12 @@ void spi_timing_get_flash_timing_param(spi_flash_hal_timing_config_t *out_timing
  */
 bool spi_timine_config_flash_is_tuned(void);
 
+/**
+ * @brief Set Flash chip specifically required MSPI register settings here
+ */
+void spi_flash_set_vendor_required_regs(void);
+
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/tools/sdk/esp32c3/include/wpa_supplicant/esp_supplicant/include/esp_mbo.h b/tools/sdk/esp32c3/include/wpa_supplicant/esp_supplicant/include/esp_mbo.h
new file mode 100644
index 00000000000..4292213943f
--- /dev/null
+++ b/tools/sdk/esp32c3/include/wpa_supplicant/esp_supplicant/include/esp_mbo.h
@@ -0,0 +1,64 @@
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+#ifndef _ESP_MBO_H
+#define _ESP_MBO_H
+
+#include <stdbool.h>
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/**
+  * enum non_pref_chan_reason: Reason for non preference of channel
+  */
+enum non_pref_chan_reason {
+	NON_PREF_CHAN_REASON_UNSPECIFIED = 0,
+	NON_PREF_CHAN_REASON_RSSI = 1,
+	NON_PREF_CHAN_REASON_EXT_INTERFERENCE = 2,
+	NON_PREF_CHAN_REASON_INT_INTERFERENCE = 3,
+};
+
+/**
+  * @brief  Channel structure for non preferred channel
+  *
+  * @param  reason: enum non_pref_chan_reason
+  * @param  oper_class: operating class for the channel
+  * @param  chan: channel number
+  * @param  preference: channel preference
+  */
+struct non_pref_chan {
+	enum non_pref_chan_reason reason;
+	uint8_t oper_class;
+	uint8_t chan;
+	uint8_t preference;
+};
+
+/**
+  * @brief  Array structure for non preferred channel struct
+  *
+  * @param  non_pref_chan_num: channel count
+  * @param  chan: array of non_pref_chan type
+  */
+struct non_pref_chan_s {
+	size_t non_pref_chan_num;
+	struct non_pref_chan chan[];
+};
+
+/**
+  * @brief  Update channel preference for MBO IE
+  *
+  * @param  non_pref_chan: Non preference channel list
+  *
+  * @return
+  *    - 0: success else failure
+  */
+int esp_mbo_update_non_pref_chan(struct non_pref_chan_s *non_pref_chan);
+
+#ifdef __cplusplus
+}
+#endif
+#endif
diff --git a/tools/sdk/esp32c3/include/wpa_supplicant/esp_supplicant/include/esp_wnm.h b/tools/sdk/esp32c3/include/wpa_supplicant/esp_supplicant/include/esp_wnm.h
index a1dcfb655c5..4301385d2cf 100644
--- a/tools/sdk/esp32c3/include/wpa_supplicant/esp_supplicant/include/esp_wnm.h
+++ b/tools/sdk/esp32c3/include/wpa_supplicant/esp_supplicant/include/esp_wnm.h
@@ -1,17 +1,7 @@
-/**
- * Copyright 2020 Espressif Systems (Shanghai) PTE LTD
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *   http://www.apache.org/licenses/LICENSE-2.0
+/*
+ * SPDX-FileCopyrightText: 2020-2021 Espressif Systems (Shanghai) CO LTD
  *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
+ * SPDX-License-Identifier: Apache-2.0
  */
 
 #ifndef _ESP_WNM_H
@@ -29,11 +19,13 @@ enum btm_query_reason {
 	REASON_UNSPECIFIED = 0,
 	REASON_FRAME_LOSS = 1,
 	REASON_DELAY = 2,
-	REASON_QOS_CAPACITY = 3,
-	REASON_FIRST_ASSOC = 4,
-	REASON_LOAD_BALALNCE = 5,
-	REASON_BETTER_AP = 6,
-	REASON_CURRENT_DEAUTH = 7,
+	REASON_BANDWIDTH = 3,
+	REASON_LOAD_BALANCE = 4,
+	REASON_RSSI = 5,
+	REASON_RETRANSMISSIONS = 6,
+	REASON_INTERFERENCE = 7,
+	REASON_GRAY_ZONE = 8,
+	REASON_PREMIUM_AP = 9,
 };
 
 /**
diff --git a/tools/sdk/esp32c3/ld/libcat_face_detect.a b/tools/sdk/esp32c3/ld/libcat_face_detect.a
new file mode 100644
index 00000000000..7a747df92d8
Binary files /dev/null and b/tools/sdk/esp32c3/ld/libcat_face_detect.a differ
diff --git a/tools/sdk/esp32c3/ld/libdl.a b/tools/sdk/esp32c3/ld/libdl.a
new file mode 100644
index 00000000000..266d5b7036e
Binary files /dev/null and b/tools/sdk/esp32c3/ld/libdl.a differ
diff --git a/tools/sdk/esp32c3/ld/libhuman_face_detect.a b/tools/sdk/esp32c3/ld/libhuman_face_detect.a
new file mode 100644
index 00000000000..cd08ef9b491
Binary files /dev/null and b/tools/sdk/esp32c3/ld/libhuman_face_detect.a differ
diff --git a/tools/sdk/esp32c3/ld/sections.ld b/tools/sdk/esp32c3/ld/sections.ld
index 2338ee8b467..692ccf1f130 100644
--- a/tools/sdk/esp32c3/ld/sections.ld
+++ b/tools/sdk/esp32c3/ld/sections.ld
@@ -164,7 +164,7 @@ SECTIONS
     *libesp_system.a:ubsan.*(.literal .literal.* .text .text.*)
     *libfreertos.a:(EXCLUDE_FILE(*libfreertos.a:port.* *libfreertos.a:port_common.*) .literal EXCLUDE_FILE(*libfreertos.a:port.* *libfreertos.a:port_common.*) .literal.* EXCLUDE_FILE(*libfreertos.a:port.* *libfreertos.a:port_common.*) .text EXCLUDE_FILE(*libfreertos.a:port.* *libfreertos.a:port_common.*) .text.*)
     *libfreertos.a:port.*(.text .text.prvTaskExitError .text.pxPortInitialiseStack .text.vApplicationStackOverflowHook .text.vPortCPUAcquireMutex .text.vPortCPUAcquireMutexTimeout .text.vPortCPUInitializeMutex .text.vPortCPUReleaseMutex .text.vPortClearInterruptMask .text.vPortEndScheduler .text.vPortEnterCritical .text.vPortExitCritical .text.vPortSetInterruptMask .text.vPortSetStackWatchpoint .text.vPortYield .text.vPortYieldFromISR .text.vPortYieldOtherCore .text.xPortGetTickRateHz .text.xPortInIsrContext .text.xPortStartScheduler)
-    *libfreertos.a:port_common.*(.text .text.esp_startup_start_app_common)
+    *libfreertos.a:port_common.*(.text .text.esp_startup_start_app_common .text.xPortCheckValidTCBMem .text.xPortcheckValidStackMem)
     *libgcc.a:_divsf3.*(.literal .literal.* .text .text.*)
     *libgcc.a:lib2funcs.*(.literal .literal.* .text .text.*)
     *libgcov.a:(.literal .literal.* .text .text.*)
@@ -311,8 +311,8 @@ SECTIONS
     _bss_start = ABSOLUTE(.);
 
     *(.bss .bss.*)
-    *(.dynbss .dynsbss .gnu.linkonce.b .gnu.linkonce.b.* .gnu.linkonce.sb .gnu.linkonce.sb.* .gnu.linkonce.sb2 .gnu.linkonce.sb2.* .sbss .sbss.* .sbss2 .sbss2.* .scommon .share.mem)
     *(.ext_ram.bss .ext_ram.bss.*)
+    *(.dynbss .dynsbss .gnu.linkonce.b .gnu.linkonce.b.* .gnu.linkonce.sb .gnu.linkonce.sb.* .gnu.linkonce.sb2 .gnu.linkonce.sb2.* .sbss .sbss.* .sbss2 .sbss2.* .scommon .share.mem)
     *(COMMON)
     _bt_bss_start = ABSOLUTE(.);
     *libbt.a:(.bss .bss.* COMMON)
diff --git a/tools/sdk/esp32c3/lib/libapp_trace.a b/tools/sdk/esp32c3/lib/libapp_trace.a
index f41ecffbefc..05a6e036ee3 100644
Binary files a/tools/sdk/esp32c3/lib/libapp_trace.a and b/tools/sdk/esp32c3/lib/libapp_trace.a differ
diff --git a/tools/sdk/esp32c3/lib/libapp_update.a b/tools/sdk/esp32c3/lib/libapp_update.a
index a76a5a0d19e..240e222f54d 100644
Binary files a/tools/sdk/esp32c3/lib/libapp_update.a and b/tools/sdk/esp32c3/lib/libapp_update.a differ
diff --git a/tools/sdk/esp32c3/lib/libbootloader_support.a b/tools/sdk/esp32c3/lib/libbootloader_support.a
index 8dee3407a13..3d1b9717628 100644
Binary files a/tools/sdk/esp32c3/lib/libbootloader_support.a and b/tools/sdk/esp32c3/lib/libbootloader_support.a differ
diff --git a/tools/sdk/esp32c3/lib/libbt.a b/tools/sdk/esp32c3/lib/libbt.a
index 97808822530..c40b123135d 100644
Binary files a/tools/sdk/esp32c3/lib/libbt.a and b/tools/sdk/esp32c3/lib/libbt.a differ
diff --git a/tools/sdk/esp32c3/lib/libcoap.a b/tools/sdk/esp32c3/lib/libcoap.a
index 407310ba1cb..96337f123e7 100644
Binary files a/tools/sdk/esp32c3/lib/libcoap.a and b/tools/sdk/esp32c3/lib/libcoap.a differ
diff --git a/tools/sdk/esp32c3/lib/libcoexist.a b/tools/sdk/esp32c3/lib/libcoexist.a
index 6f03cab0b8e..de5915cbc2d 100644
Binary files a/tools/sdk/esp32c3/lib/libcoexist.a and b/tools/sdk/esp32c3/lib/libcoexist.a differ
diff --git a/tools/sdk/esp32c3/lib/libconsole.a b/tools/sdk/esp32c3/lib/libconsole.a
index 91bd32add2d..6ce07b6431b 100644
Binary files a/tools/sdk/esp32c3/lib/libconsole.a and b/tools/sdk/esp32c3/lib/libconsole.a differ
diff --git a/tools/sdk/esp32c3/lib/libcore.a b/tools/sdk/esp32c3/lib/libcore.a
index 10e7ca5a83b..38e7e437963 100644
Binary files a/tools/sdk/esp32c3/lib/libcore.a and b/tools/sdk/esp32c3/lib/libcore.a differ
diff --git a/tools/sdk/esp32c3/lib/libcxx.a b/tools/sdk/esp32c3/lib/libcxx.a
index 9617dd86b74..235bc776b7a 100644
Binary files a/tools/sdk/esp32c3/lib/libcxx.a and b/tools/sdk/esp32c3/lib/libcxx.a differ
diff --git a/tools/sdk/esp32c3/lib/libdriver.a b/tools/sdk/esp32c3/lib/libdriver.a
index 71809d7ec58..9f9222d7812 100644
Binary files a/tools/sdk/esp32c3/lib/libdriver.a and b/tools/sdk/esp32c3/lib/libdriver.a differ
diff --git a/tools/sdk/esp32c3/lib/libefuse.a b/tools/sdk/esp32c3/lib/libefuse.a
index f4bb53d7dae..99ee734cff0 100644
Binary files a/tools/sdk/esp32c3/lib/libefuse.a and b/tools/sdk/esp32c3/lib/libefuse.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp-tls.a b/tools/sdk/esp32c3/lib/libesp-tls.a
index 4e304a6646f..f493374d2c5 100644
Binary files a/tools/sdk/esp32c3/lib/libesp-tls.a and b/tools/sdk/esp32c3/lib/libesp-tls.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_common.a b/tools/sdk/esp32c3/lib/libesp_common.a
index 577a8884b7f..c789ccd4d02 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_common.a and b/tools/sdk/esp32c3/lib/libesp_common.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_eth.a b/tools/sdk/esp32c3/lib/libesp_eth.a
index 457368b2747..5e6eae54ae7 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_eth.a and b/tools/sdk/esp32c3/lib/libesp_eth.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_event.a b/tools/sdk/esp32c3/lib/libesp_event.a
index eb12ef2bb70..7e12d28723a 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_event.a and b/tools/sdk/esp32c3/lib/libesp_event.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_hid.a b/tools/sdk/esp32c3/lib/libesp_hid.a
index f976d5ab228..ed01f19522e 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_hid.a and b/tools/sdk/esp32c3/lib/libesp_hid.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_http_client.a b/tools/sdk/esp32c3/lib/libesp_http_client.a
index f82ad10086d..355edfdbd75 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_http_client.a and b/tools/sdk/esp32c3/lib/libesp_http_client.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_http_server.a b/tools/sdk/esp32c3/lib/libesp_http_server.a
index 78a0206aab2..862a0c58399 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_http_server.a and b/tools/sdk/esp32c3/lib/libesp_http_server.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_https_ota.a b/tools/sdk/esp32c3/lib/libesp_https_ota.a
index e1702191f75..ad874bcf2a4 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_https_ota.a and b/tools/sdk/esp32c3/lib/libesp_https_ota.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_https_server.a b/tools/sdk/esp32c3/lib/libesp_https_server.a
index 41c852b2f76..18fc0f04916 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_https_server.a and b/tools/sdk/esp32c3/lib/libesp_https_server.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_hw_support.a b/tools/sdk/esp32c3/lib/libesp_hw_support.a
index 441d8231622..8689cdc3f30 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_hw_support.a and b/tools/sdk/esp32c3/lib/libesp_hw_support.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_lcd.a b/tools/sdk/esp32c3/lib/libesp_lcd.a
index d5f9a5d15a7..e627bdf82f6 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_lcd.a and b/tools/sdk/esp32c3/lib/libesp_lcd.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_littlefs.a b/tools/sdk/esp32c3/lib/libesp_littlefs.a
index 10ece18e6f5..195f8c5f210 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_littlefs.a and b/tools/sdk/esp32c3/lib/libesp_littlefs.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_local_ctrl.a b/tools/sdk/esp32c3/lib/libesp_local_ctrl.a
index 79208e5d357..9df0a6ca3e6 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_local_ctrl.a and b/tools/sdk/esp32c3/lib/libesp_local_ctrl.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_netif.a b/tools/sdk/esp32c3/lib/libesp_netif.a
index 3deef9ca7dc..697c3645d07 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_netif.a and b/tools/sdk/esp32c3/lib/libesp_netif.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_phy.a b/tools/sdk/esp32c3/lib/libesp_phy.a
index b2b9885f59f..dc237e13710 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_phy.a and b/tools/sdk/esp32c3/lib/libesp_phy.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_pm.a b/tools/sdk/esp32c3/lib/libesp_pm.a
index f32146b8ca0..ee647250ec2 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_pm.a and b/tools/sdk/esp32c3/lib/libesp_pm.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_ringbuf.a b/tools/sdk/esp32c3/lib/libesp_ringbuf.a
index d08c234dc5b..ccd9d1a1f27 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_ringbuf.a and b/tools/sdk/esp32c3/lib/libesp_ringbuf.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_serial_slave_link.a b/tools/sdk/esp32c3/lib/libesp_serial_slave_link.a
index 3a5cbd2b6ad..c5740f366be 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_serial_slave_link.a and b/tools/sdk/esp32c3/lib/libesp_serial_slave_link.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_system.a b/tools/sdk/esp32c3/lib/libesp_system.a
index f8421329011..0ad8107ddb9 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_system.a and b/tools/sdk/esp32c3/lib/libesp_system.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_timer.a b/tools/sdk/esp32c3/lib/libesp_timer.a
index d74a2ac003b..29d0ea65fbc 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_timer.a and b/tools/sdk/esp32c3/lib/libesp_timer.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_websocket_client.a b/tools/sdk/esp32c3/lib/libesp_websocket_client.a
index d02f65a06fa..ea296d2c774 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_websocket_client.a and b/tools/sdk/esp32c3/lib/libesp_websocket_client.a differ
diff --git a/tools/sdk/esp32c3/lib/libesp_wifi.a b/tools/sdk/esp32c3/lib/libesp_wifi.a
index 3a89f878c94..4c73f013066 100644
Binary files a/tools/sdk/esp32c3/lib/libesp_wifi.a and b/tools/sdk/esp32c3/lib/libesp_wifi.a differ
diff --git a/tools/sdk/esp32c3/lib/libespcoredump.a b/tools/sdk/esp32c3/lib/libespcoredump.a
index 4af35b8b809..80c3fc77e9e 100644
Binary files a/tools/sdk/esp32c3/lib/libespcoredump.a and b/tools/sdk/esp32c3/lib/libespcoredump.a differ
diff --git a/tools/sdk/esp32c3/lib/libespnow.a b/tools/sdk/esp32c3/lib/libespnow.a
index 0d53db2568a..8c792259076 100644
Binary files a/tools/sdk/esp32c3/lib/libespnow.a and b/tools/sdk/esp32c3/lib/libespnow.a differ
diff --git a/tools/sdk/esp32c3/lib/libfatfs.a b/tools/sdk/esp32c3/lib/libfatfs.a
index 86f197e368c..f06478f76d1 100644
Binary files a/tools/sdk/esp32c3/lib/libfatfs.a and b/tools/sdk/esp32c3/lib/libfatfs.a differ
diff --git a/tools/sdk/esp32c3/lib/libfreemodbus.a b/tools/sdk/esp32c3/lib/libfreemodbus.a
index b46ecaac145..9ae7d61005f 100644
Binary files a/tools/sdk/esp32c3/lib/libfreemodbus.a and b/tools/sdk/esp32c3/lib/libfreemodbus.a differ
diff --git a/tools/sdk/esp32c3/lib/libfreertos.a b/tools/sdk/esp32c3/lib/libfreertos.a
index ac3c0fd50d1..acd068a9e05 100644
Binary files a/tools/sdk/esp32c3/lib/libfreertos.a and b/tools/sdk/esp32c3/lib/libfreertos.a differ
diff --git a/tools/sdk/esp32c3/lib/libhal.a b/tools/sdk/esp32c3/lib/libhal.a
index 12ad247f09c..ed34868e2e8 100644
Binary files a/tools/sdk/esp32c3/lib/libhal.a and b/tools/sdk/esp32c3/lib/libhal.a differ
diff --git a/tools/sdk/esp32c3/lib/libheap.a b/tools/sdk/esp32c3/lib/libheap.a
index 240bbab74e0..0ae3c7e01af 100644
Binary files a/tools/sdk/esp32c3/lib/libheap.a and b/tools/sdk/esp32c3/lib/libheap.a differ
diff --git a/tools/sdk/esp32c3/lib/liblog.a b/tools/sdk/esp32c3/lib/liblog.a
index 4be299b20b5..93460d356de 100644
Binary files a/tools/sdk/esp32c3/lib/liblog.a and b/tools/sdk/esp32c3/lib/liblog.a differ
diff --git a/tools/sdk/esp32c3/lib/liblwip.a b/tools/sdk/esp32c3/lib/liblwip.a
index 668266754ba..1051575b7bb 100644
Binary files a/tools/sdk/esp32c3/lib/liblwip.a and b/tools/sdk/esp32c3/lib/liblwip.a differ
diff --git a/tools/sdk/esp32c3/lib/libmbedcrypto.a b/tools/sdk/esp32c3/lib/libmbedcrypto.a
index b1cb253b4d7..3f13e9da4cb 100644
Binary files a/tools/sdk/esp32c3/lib/libmbedcrypto.a and b/tools/sdk/esp32c3/lib/libmbedcrypto.a differ
diff --git a/tools/sdk/esp32c3/lib/libmdns.a b/tools/sdk/esp32c3/lib/libmdns.a
index b3df4d78b13..49de62082ea 100644
Binary files a/tools/sdk/esp32c3/lib/libmdns.a and b/tools/sdk/esp32c3/lib/libmdns.a differ
diff --git a/tools/sdk/esp32c3/lib/libmesh.a b/tools/sdk/esp32c3/lib/libmesh.a
index ac0bceb375b..caf1f0c2200 100644
Binary files a/tools/sdk/esp32c3/lib/libmesh.a and b/tools/sdk/esp32c3/lib/libmesh.a differ
diff --git a/tools/sdk/esp32c3/lib/libmqtt.a b/tools/sdk/esp32c3/lib/libmqtt.a
index 7b2acfde2a4..a0b3d873517 100644
Binary files a/tools/sdk/esp32c3/lib/libmqtt.a and b/tools/sdk/esp32c3/lib/libmqtt.a differ
diff --git a/tools/sdk/esp32c3/lib/libnet80211.a b/tools/sdk/esp32c3/lib/libnet80211.a
index 79f91f7f17d..659d0c256b4 100644
Binary files a/tools/sdk/esp32c3/lib/libnet80211.a and b/tools/sdk/esp32c3/lib/libnet80211.a differ
diff --git a/tools/sdk/esp32c3/lib/libnewlib.a b/tools/sdk/esp32c3/lib/libnewlib.a
index e0ebeacbfd1..d11f7b0890e 100644
Binary files a/tools/sdk/esp32c3/lib/libnewlib.a and b/tools/sdk/esp32c3/lib/libnewlib.a differ
diff --git a/tools/sdk/esp32c3/lib/libnvs_flash.a b/tools/sdk/esp32c3/lib/libnvs_flash.a
index 72bb887799b..9d11816f2be 100644
Binary files a/tools/sdk/esp32c3/lib/libnvs_flash.a and b/tools/sdk/esp32c3/lib/libnvs_flash.a differ
diff --git a/tools/sdk/esp32c3/lib/libopenssl.a b/tools/sdk/esp32c3/lib/libopenssl.a
index 4019e41f7a4..a3187965100 100644
Binary files a/tools/sdk/esp32c3/lib/libopenssl.a and b/tools/sdk/esp32c3/lib/libopenssl.a differ
diff --git a/tools/sdk/esp32c3/lib/libpp.a b/tools/sdk/esp32c3/lib/libpp.a
index 4891f8b5c92..0da117ad274 100644
Binary files a/tools/sdk/esp32c3/lib/libpp.a and b/tools/sdk/esp32c3/lib/libpp.a differ
diff --git a/tools/sdk/esp32c3/lib/libprotocomm.a b/tools/sdk/esp32c3/lib/libprotocomm.a
index b604bd05cd1..c21544bf740 100644
Binary files a/tools/sdk/esp32c3/lib/libprotocomm.a and b/tools/sdk/esp32c3/lib/libprotocomm.a differ
diff --git a/tools/sdk/esp32c3/lib/libpthread.a b/tools/sdk/esp32c3/lib/libpthread.a
index b16b6ff4e64..4a83fa2779a 100644
Binary files a/tools/sdk/esp32c3/lib/libpthread.a and b/tools/sdk/esp32c3/lib/libpthread.a differ
diff --git a/tools/sdk/esp32c3/lib/libsdmmc.a b/tools/sdk/esp32c3/lib/libsdmmc.a
index 04cbb827fc5..8dd26b634d1 100644
Binary files a/tools/sdk/esp32c3/lib/libsdmmc.a and b/tools/sdk/esp32c3/lib/libsdmmc.a differ
diff --git a/tools/sdk/esp32c3/lib/libsmartconfig.a b/tools/sdk/esp32c3/lib/libsmartconfig.a
index bd9cc0d4cee..bf215058c15 100644
Binary files a/tools/sdk/esp32c3/lib/libsmartconfig.a and b/tools/sdk/esp32c3/lib/libsmartconfig.a differ
diff --git a/tools/sdk/esp32c3/lib/libspi_flash.a b/tools/sdk/esp32c3/lib/libspi_flash.a
index ac3b0ee32af..111a88afcc9 100644
Binary files a/tools/sdk/esp32c3/lib/libspi_flash.a and b/tools/sdk/esp32c3/lib/libspi_flash.a differ
diff --git a/tools/sdk/esp32c3/lib/libspiffs.a b/tools/sdk/esp32c3/lib/libspiffs.a
index 120571b2b86..0cc5fc8f091 100644
Binary files a/tools/sdk/esp32c3/lib/libspiffs.a and b/tools/sdk/esp32c3/lib/libspiffs.a differ
diff --git a/tools/sdk/esp32c3/lib/libtcp_transport.a b/tools/sdk/esp32c3/lib/libtcp_transport.a
index 42d386c8e1a..8ed74692157 100644
Binary files a/tools/sdk/esp32c3/lib/libtcp_transport.a and b/tools/sdk/esp32c3/lib/libtcp_transport.a differ
diff --git a/tools/sdk/esp32c3/lib/libtcpip_adapter.a b/tools/sdk/esp32c3/lib/libtcpip_adapter.a
index 316d8562b49..fd243c03dbc 100644
Binary files a/tools/sdk/esp32c3/lib/libtcpip_adapter.a and b/tools/sdk/esp32c3/lib/libtcpip_adapter.a differ
diff --git a/tools/sdk/esp32c3/lib/libvfs.a b/tools/sdk/esp32c3/lib/libvfs.a
index 4b644d301c0..bd784ad9cbb 100644
Binary files a/tools/sdk/esp32c3/lib/libvfs.a and b/tools/sdk/esp32c3/lib/libvfs.a differ
diff --git a/tools/sdk/esp32c3/lib/libwapi.a b/tools/sdk/esp32c3/lib/libwapi.a
index 549b78aad96..7fe0160c721 100644
Binary files a/tools/sdk/esp32c3/lib/libwapi.a and b/tools/sdk/esp32c3/lib/libwapi.a differ
diff --git a/tools/sdk/esp32c3/lib/libwear_levelling.a b/tools/sdk/esp32c3/lib/libwear_levelling.a
index 8e320e50db7..4001b7a1319 100644
Binary files a/tools/sdk/esp32c3/lib/libwear_levelling.a and b/tools/sdk/esp32c3/lib/libwear_levelling.a differ
diff --git a/tools/sdk/esp32c3/lib/libwifi_provisioning.a b/tools/sdk/esp32c3/lib/libwifi_provisioning.a
index 08e7562e415..41243a30389 100644
Binary files a/tools/sdk/esp32c3/lib/libwifi_provisioning.a and b/tools/sdk/esp32c3/lib/libwifi_provisioning.a differ
diff --git a/tools/sdk/esp32c3/lib/libwpa_supplicant.a b/tools/sdk/esp32c3/lib/libwpa_supplicant.a
index 60617dc9981..2191f1ee02b 100644
Binary files a/tools/sdk/esp32c3/lib/libwpa_supplicant.a and b/tools/sdk/esp32c3/lib/libwpa_supplicant.a differ
diff --git a/tools/sdk/esp32c3/sdkconfig b/tools/sdk/esp32c3/sdkconfig
index 85edbebed07..fac95fd42fb 100644
--- a/tools/sdk/esp32c3/sdkconfig
+++ b/tools/sdk/esp32c3/sdkconfig
@@ -92,6 +92,7 @@ CONFIG_ESPTOOLPY_BAUD_OTHER_VAL=115200
 # CONFIG_ESPTOOLPY_FLASHMODE_QOUT is not set
 CONFIG_ESPTOOLPY_FLASHMODE_DIO=y
 # CONFIG_ESPTOOLPY_FLASHMODE_DOUT is not set
+CONFIG_ESPTOOLPY_FLASH_SAMPLE_MODE_STR=y
 CONFIG_ESPTOOLPY_FLASHMODE="dio"
 # CONFIG_ESPTOOLPY_FLASHFREQ_120M is not set
 CONFIG_ESPTOOLPY_FLASHFREQ_80M=y
@@ -1044,6 +1045,7 @@ CONFIG_LWIP_TCPIP_RECVMBOX_SIZE=32
 CONFIG_LWIP_DHCP_DOES_ARP_CHECK=y
 # CONFIG_LWIP_DHCP_DISABLE_CLIENT_ID is not set
 # CONFIG_LWIP_DHCP_RESTORE_LAST_IP is not set
+CONFIG_LWIP_DHCP_OPTIONS_LEN=68
 
 #
 # DHCP server
@@ -1291,6 +1293,7 @@ CONFIG_MDNS_SERVICE_ADD_TIMEOUT_MS=2000
 # CONFIG_MDNS_STRICT_MODE is not set
 CONFIG_MDNS_TIMER_PERIOD_MS=100
 # CONFIG_MDNS_NETWORKING_SOCKET is not set
+CONFIG_MDNS_MULTIPLE_INSTANCE=y
 # end of mDNS
 
 #
@@ -1475,6 +1478,7 @@ CONFIG_WL_SECTOR_SIZE=4096
 #
 CONFIG_WIFI_PROV_SCAN_MAX_ENTRIES=16
 CONFIG_WIFI_PROV_AUTOSTOP_TIMEOUT=30
+# CONFIG_WIFI_PROV_BLE_BONDING is not set
 # end of Wi-Fi Provisioning Manager
 
 #
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/include/tusb_config.h b/tools/sdk/esp32s2/include/arduino_tinyusb/include/tusb_config.h
index a119e4d81c3..a5a0afd32dc 100755
--- a/tools/sdk/esp32s2/include/arduino_tinyusb/include/tusb_config.h
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/include/tusb_config.h
@@ -52,6 +52,10 @@ extern "C" {
 #   define CONFIG_TINYUSB_MIDI_ENABLED 0
 #endif
 
+#ifndef CONFIG_TINYUSB_VIDEO_ENABLED
+#   define CONFIG_TINYUSB_VIDEO_ENABLED 0
+#endif
+
 #ifndef CONFIG_TINYUSB_CUSTOM_CLASS_ENABLED
 #   define CONFIG_TINYUSB_CUSTOM_CLASS_ENABLED 0
 #endif
@@ -99,6 +103,7 @@ extern "C" {
 #define CFG_TUD_MSC 				CONFIG_TINYUSB_MSC_ENABLED
 #define CFG_TUD_HID 				CONFIG_TINYUSB_HID_ENABLED
 #define CFG_TUD_MIDI 				CONFIG_TINYUSB_MIDI_ENABLED
+#define CFG_TUD_VIDEO               CONFIG_TINYUSB_VIDEO_ENABLED
 #define CFG_TUD_CUSTOM_CLASS 		CONFIG_TINYUSB_CUSTOM_CLASS_ENABLED
 #define CFG_TUD_DFU_RUNTIME			CONFIG_TINYUSB_DFU_RT_ENABLED
 #define CFG_TUD_VENDOR 				CONFIG_TINYUSB_VENDOR_ENABLED
@@ -117,6 +122,10 @@ extern "C" {
 #define CFG_TUD_MIDI_RX_BUFSIZE		CONFIG_TINYUSB_MIDI_RX_BUFSIZE
 #define CFG_TUD_MIDI_TX_BUFSIZE		CONFIG_TINYUSB_MIDI_TX_BUFSIZE
 
+// The number of video streaming interfaces and  endpoint size
+#define CFG_TUD_VIDEO_STREAMING     CONFIG_TINYUSB_VIDEO_STREAMING_IFS
+#define CFG_TUD_VIDEO_STREAMING_EP_BUFSIZE  CONFIG_TINYUSB_VIDEO_STREAMING_BUFSIZE
+
 // VENDOR FIFO size of TX and RX
 #define CFG_TUD_VENDOR_RX_BUFSIZE 	CONFIG_TINYUSB_VENDOR_RX_BUFSIZE
 #define CFG_TUD_VENDOR_TX_BUFSIZE 	CONFIG_TINYUSB_VENDOR_TX_BUFSIZE
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/audio/audio.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/audio/audio.h
index f99061eaec6..6f9c1a6b582 100644
--- a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/audio/audio.h
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/audio/audio.h
@@ -494,18 +494,6 @@ typedef enum
 
 /// All remaining definitions are taken from the descriptor descriptions in the UAC2 main specification
 
-/// Isochronous End Point Attributes
-typedef enum
-{
-  TUSB_ISO_EP_ATT_NO_SYNC         = 0x00,
-  TUSB_ISO_EP_ATT_ASYNCHRONOUS    = 0x04,
-  TUSB_ISO_EP_ATT_ADAPTIVE        = 0x08,
-  TUSB_ISO_EP_ATT_SYNCHRONOUS     = 0x0C,
-  TUSB_ISO_EP_ATT_DATA            = 0x00, ///< Data End Point
-  TUSB_ISO_EP_ATT_EXPLICIT_FB     = 0x10, ///< Feedback End Point
-  TUSB_ISO_EP_ATT_IMPLICIT_FB     = 0x20, ///< Data endpoint that also serves as an implicit feedback
-} tusb_iso_ep_attribute_t;
-
 /// Audio Class-Control Values UAC2
 typedef enum
 {
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/cdc/cdc.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/cdc/cdc.h
index 5df47f70b4a..e345139eab0 100644
--- a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/cdc/cdc.h
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/cdc/cdc.h
@@ -58,31 +58,32 @@ typedef enum
 /// Communication Interface Subclass Codes
 typedef enum
 {
-  CDC_COMM_SUBCLASS_DIRECT_LINE_CONTROL_MODEL = 0x01  , ///< Direct Line Control Model  [USBPSTN1.2]
-  CDC_COMM_SUBCLASS_ABSTRACT_CONTROL_MODEL            , ///< Abstract Control Model  [USBPSTN1.2]
-  CDC_COMM_SUBCLASS_TELEPHONE_CONTROL_MODEL           , ///< Telephone Control Model  [USBPSTN1.2]
-  CDC_COMM_SUBCLASS_MULTICHANNEL_CONTROL_MODEL        , ///< Multi-Channel Control Model  [USBISDN1.2]
-  CDC_COMM_SUBCLASS_CAPI_CONTROL_MODEL                , ///< CAPI Control Model  [USBISDN1.2]
-  CDC_COMM_SUBCLASS_ETHERNET_CONTROL_MODEL            , ///< Ethernet Networking Control Model  [USBECM1.2]
-  CDC_COMM_SUBCLASS_ATM_NETWORKING_CONTROL_MODEL      , ///< ATM Networking Control Model  [USBATM1.2]
-  CDC_COMM_SUBCLASS_WIRELESS_HANDSET_CONTROL_MODEL    , ///< Wireless Handset Control Model  [USBWMC1.1]
-  CDC_COMM_SUBCLASS_DEVICE_MANAGEMENT                 , ///< Device Management  [USBWMC1.1]
-  CDC_COMM_SUBCLASS_MOBILE_DIRECT_LINE_MODEL          , ///< Mobile Direct Line Model  [USBWMC1.1]
-  CDC_COMM_SUBCLASS_OBEX                              , ///< OBEX  [USBWMC1.1]
-  CDC_COMM_SUBCLASS_ETHERNET_EMULATION_MODEL            ///< Ethernet Emulation Model  [USBEEM1.0]
+  CDC_COMM_SUBCLASS_DIRECT_LINE_CONTROL_MODEL      = 0x01 , ///< Direct Line Control Model         [USBPSTN1.2]
+  CDC_COMM_SUBCLASS_ABSTRACT_CONTROL_MODEL         = 0x02 , ///< Abstract Control Model            [USBPSTN1.2]
+  CDC_COMM_SUBCLASS_TELEPHONE_CONTROL_MODEL        = 0x03 , ///< Telephone Control Model           [USBPSTN1.2]
+  CDC_COMM_SUBCLASS_MULTICHANNEL_CONTROL_MODEL     = 0x04 , ///< Multi-Channel Control Model       [USBISDN1.2]
+  CDC_COMM_SUBCLASS_CAPI_CONTROL_MODEL             = 0x05 , ///< CAPI Control Model                [USBISDN1.2]
+  CDC_COMM_SUBCLASS_ETHERNET_CONTROL_MODEL         = 0x06 , ///< Ethernet Networking Control Model [USBECM1.2]
+  CDC_COMM_SUBCLASS_ATM_NETWORKING_CONTROL_MODEL   = 0x07 , ///< ATM Networking Control Model      [USBATM1.2]
+  CDC_COMM_SUBCLASS_WIRELESS_HANDSET_CONTROL_MODEL = 0x08 , ///< Wireless Handset Control Model    [USBWMC1.1]
+  CDC_COMM_SUBCLASS_DEVICE_MANAGEMENT              = 0x09 , ///< Device Management                 [USBWMC1.1]
+  CDC_COMM_SUBCLASS_MOBILE_DIRECT_LINE_MODEL       = 0x0A , ///< Mobile Direct Line Model          [USBWMC1.1]
+  CDC_COMM_SUBCLASS_OBEX                           = 0x0B , ///< OBEX                              [USBWMC1.1]
+  CDC_COMM_SUBCLASS_ETHERNET_EMULATION_MODEL       = 0x0C , ///< Ethernet Emulation Model          [USBEEM1.0]
+  CDC_COMM_SUBCLASS_NETWORK_CONTROL_MODEL          = 0x0D   ///< Network Control Model             [USBNCM1.0]
 } cdc_comm_sublcass_type_t;
 
 /// Communication Interface Protocol Codes
 typedef enum
 {
-  CDC_COMM_PROTOCOL_NONE                   = 0x00 , ///< No specific protocol
-  CDC_COMM_PROTOCOL_ATCOMMAND                     , ///< AT Commands: V.250 etc
-  CDC_COMM_PROTOCOL_ATCOMMAND_PCCA_101            , ///< AT Commands defined by PCCA-101
-  CDC_COMM_PROTOCOL_ATCOMMAND_PCCA_101_AND_ANNEXO , ///< AT Commands defined by PCCA-101 & Annex O
-  CDC_COMM_PROTOCOL_ATCOMMAND_GSM_707             , ///< AT Commands defined by GSM 07.07
-  CDC_COMM_PROTOCOL_ATCOMMAND_3GPP_27007          , ///< AT Commands defined by 3GPP 27.007
-  CDC_COMM_PROTOCOL_ATCOMMAND_CDMA                , ///< AT Commands defined by TIA for CDMA
-  CDC_COMM_PROTOCOL_ETHERNET_EMULATION_MODEL        ///< Ethernet Emulation Model
+  CDC_COMM_PROTOCOL_NONE                          = 0x00 , ///< No specific protocol
+  CDC_COMM_PROTOCOL_ATCOMMAND                     = 0x01 , ///< AT Commands: V.250 etc
+  CDC_COMM_PROTOCOL_ATCOMMAND_PCCA_101            = 0x02 , ///< AT Commands defined by PCCA-101
+  CDC_COMM_PROTOCOL_ATCOMMAND_PCCA_101_AND_ANNEXO = 0x03 , ///< AT Commands defined by PCCA-101 & Annex O
+  CDC_COMM_PROTOCOL_ATCOMMAND_GSM_707             = 0x04 , ///< AT Commands defined by GSM 07.07
+  CDC_COMM_PROTOCOL_ATCOMMAND_3GPP_27007          = 0x05 , ///< AT Commands defined by 3GPP 27.007
+  CDC_COMM_PROTOCOL_ATCOMMAND_CDMA                = 0x06 , ///< AT Commands defined by TIA for CDMA
+  CDC_COMM_PROTOCOL_ETHERNET_EMULATION_MODEL      = 0x07   ///< Ethernet Emulation Model
 } cdc_comm_protocol_type_t;
 
 //------------- SubType Descriptor in COMM Functional Descriptor -------------//
@@ -114,7 +115,8 @@ typedef enum
   CDC_FUNC_DESC_COMMAND_SET                                      = 0x16 , ///< Command Set Functional Descriptor
   CDC_FUNC_DESC_COMMAND_SET_DETAIL                               = 0x17 , ///< Command Set Detail Functional Descriptor
   CDC_FUNC_DESC_TELEPHONE_CONTROL_MODEL                          = 0x18 , ///< Telephone Control Model Functional Descriptor
-  CDC_FUNC_DESC_OBEX_SERVICE_IDENTIFIER                          = 0x19   ///< OBEX Service Identifier Functional Descriptor
+  CDC_FUNC_DESC_OBEX_SERVICE_IDENTIFIER                          = 0x19 , ///< OBEX Service Identifier Functional Descriptor
+  CDC_FUNC_DESC_NCM                                              = 0x1A , ///< NCM Functional Descriptor
 }cdc_func_desc_type_t;
 
 //--------------------------------------------------------------------+
@@ -122,7 +124,8 @@ typedef enum
 //--------------------------------------------------------------------+
 
 // SUBCLASS code of Data Interface is not used and should/must be zero
-/// Data Interface Protocol Codes
+
+// Data Interface Protocol Codes
 typedef enum{
   CDC_DATA_PROTOCOL_ISDN_BRI                               = 0x30, ///< Physical interface protocol for ISDN BRI
   CDC_DATA_PROTOCOL_HDLC                                   = 0x31, ///< HDLC
@@ -147,7 +150,6 @@ typedef enum
 {
   CDC_REQUEST_SEND_ENCAPSULATED_COMMAND                    = 0x00, ///< is used to issue a command in the format of the supported control protocol of the Communications Class interface
   CDC_REQUEST_GET_ENCAPSULATED_RESPONSE                    = 0x01, ///< is used to request a response in the format of the supported control protocol of the Communications Class interface.
-
   CDC_REQUEST_SET_COMM_FEATURE                             = 0x02,
   CDC_REQUEST_GET_COMM_FEATURE                             = 0x03,
   CDC_REQUEST_CLEAR_COMM_FEATURE                           = 0x04,
@@ -194,21 +196,18 @@ typedef enum
 // Management Elemenent Notification (Notification Endpoint)
 //--------------------------------------------------------------------+
 
-/// Communication Interface Management Element Notification Codes
+/// 6.3 Notification Codes
 typedef enum
 {
-  NETWORK_CONNECTION               = 0x00, ///< This notification allows the device to notify the host about network connection status.
-  RESPONSE_AVAILABLE               = 0x01, ///< This notification allows the device to notify the hostthat a response is available. This response can be retrieved with a subsequent \ref CDC_REQUEST_GET_ENCAPSULATED_RESPONSE request.
-
-  AUX_JACK_HOOK_STATE              = 0x08,
-  RING_DETECT                      = 0x09,
-
-  SERIAL_STATE                     = 0x20,
-
-  CALL_STATE_CHANGE                = 0x28,
-  LINE_STATE_CHANGE                = 0x29,
-  CONNECTION_SPEED_CHANGE          = 0x2A, ///< This notification allows the device to inform the host-networking driver that a change in either the upstream or the downstream bit rate of the connection has occurred
-  MDLM_SEMANTIC_MODEL_NOTIFICATION = 0x40,
+  CDC_NOTIF_NETWORK_CONNECTION               = 0x00, ///< This notification allows the device to notify the host about network connection status.
+  CDC_NOTIF_RESPONSE_AVAILABLE               = 0x01, ///< This notification allows the device to notify the hostthat a response is available. This response can be retrieved with a subsequent \ref CDC_REQUEST_GET_ENCAPSULATED_RESPONSE request.
+  CDC_NOTIF_AUX_JACK_HOOK_STATE              = 0x08,
+  CDC_NOTIF_RING_DETECT                      = 0x09,
+  CDC_NOTIF_SERIAL_STATE                     = 0x20,
+  CDC_NOTIF_CALL_STATE_CHANGE                = 0x28,
+  CDC_NOTIF_LINE_STATE_CHANGE                = 0x29,
+  CDC_NOTIF_CONNECTION_SPEED_CHANGE          = 0x2A, ///< This notification allows the device to inform the host-networking driver that a change in either the upstream or the downstream bit rate of the connection has occurred
+  CDC_NOTIF_MDLM_SEMANTIC_MODEL_NOTIFICATION = 0x40,
 }cdc_notification_request_t;
 
 //--------------------------------------------------------------------+
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/cdc/cdc_device.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/cdc/cdc_device.h
index 7ff757addd6..fbc7162a366 100644
--- a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/cdc/cdc_device.h
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/cdc/cdc_device.h
@@ -82,7 +82,7 @@ int32_t  tud_cdc_n_read_char       (uint8_t itf);
 void     tud_cdc_n_read_flush      (uint8_t itf);
 
 // Get a byte from FIFO at the specified position without removing it
-bool     tud_cdc_n_peek            (uint8_t itf, uint8_t* u8);
+bool     tud_cdc_n_peek            (uint8_t itf, uint8_t* ui8);
 
 // Write bytes to TX FIFO, data may remain in the FIFO for a while
 uint32_t tud_cdc_n_write           (uint8_t itf, void const* buffer, uint32_t bufsize);
@@ -116,7 +116,7 @@ static inline uint32_t tud_cdc_available       (void);
 static inline int32_t  tud_cdc_read_char       (void);
 static inline uint32_t tud_cdc_read            (void* buffer, uint32_t bufsize);
 static inline void     tud_cdc_read_flush      (void);
-static inline bool     tud_cdc_peek            (uint8_t* u8);
+static inline bool     tud_cdc_peek            (uint8_t* ui8);
 
 static inline uint32_t tud_cdc_write_char      (char ch);
 static inline uint32_t tud_cdc_write           (void const* buffer, uint32_t bufsize);
@@ -206,9 +206,9 @@ static inline void tud_cdc_read_flush (void)
   tud_cdc_n_read_flush(0);
 }
 
-static inline bool tud_cdc_peek (uint8_t* u8)
+static inline bool tud_cdc_peek (uint8_t* ui8)
 {
-  return tud_cdc_n_peek(0, u8);
+  return tud_cdc_n_peek(0, ui8);
 }
 
 static inline uint32_t tud_cdc_write_char (char ch)
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/net/ncm.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/net/ncm.h
new file mode 100644
index 00000000000..96ba11fbc5c
--- /dev/null
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/net/ncm.h
@@ -0,0 +1,69 @@
+/*
+ * The MIT License (MIT)
+ *
+ * Copyright (c) 2021, Ha Thach (tinyusb.org)
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ *
+ * This file is part of the TinyUSB stack.
+ */
+
+
+#ifndef _TUSB_NCM_H_
+#define _TUSB_NCM_H_
+
+#include "common/tusb_common.h"
+
+#ifdef __cplusplus
+ extern "C" {
+#endif
+
+// Table 4.3 Data Class Interface Protocol Codes
+typedef enum
+{
+  NCM_DATA_PROTOCOL_NETWORK_TRANSFER_BLOCK = 0x01
+} ncm_data_interface_protocol_code_t;
+
+
+// Table 6.2 Class-Specific Request Codes for Network Control Model subclass
+typedef enum
+{
+  NCM_SET_ETHERNET_MULTICAST_FILTERS               = 0x40,
+  NCM_SET_ETHERNET_POWER_MANAGEMENT_PATTERN_FILTER = 0x41,
+  NCM_GET_ETHERNET_POWER_MANAGEMENT_PATTERN_FILTER = 0x42,
+  NCM_SET_ETHERNET_PACKET_FILTER                   = 0x43,
+  NCM_GET_ETHERNET_STATISTIC                       = 0x44,
+  NCM_GET_NTB_PARAMETERS                           = 0x80,
+  NCM_GET_NET_ADDRESS                              = 0x81,
+  NCM_SET_NET_ADDRESS                              = 0x82,
+  NCM_GET_NTB_FORMAT                               = 0x83,
+  NCM_SET_NTB_FORMAT                               = 0x84,
+  NCM_GET_NTB_INPUT_SIZE                           = 0x85,
+  NCM_SET_NTB_INPUT_SIZE                           = 0x86,
+  NCM_GET_MAX_DATAGRAM_SIZE                        = 0x87,
+  NCM_SET_MAX_DATAGRAM_SIZE                        = 0x88,
+  NCM_GET_CRC_MODE                                 = 0x89,
+  NCM_SET_CRC_MODE                                 = 0x8A,
+} ncm_request_code_t;
+
+#ifdef __cplusplus
+ }
+#endif
+
+#endif
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/net/net_device.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/net/net_device.h
index f030f3077ac..6e294465b28 100644
--- a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/net/net_device.h
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/net/net_device.h
@@ -30,14 +30,36 @@
 
 #include "class/cdc/cdc.h"
 
+#if CFG_TUD_ECM_RNDIS && CFG_TUD_NCM
+#error "Cannot enable both ECM_RNDIS and NCM network drivers"
+#endif
+
+#include "ncm.h"
+
 /* declared here, NOT in usb_descriptors.c, so that the driver can intelligently ZLP as needed */
 #define CFG_TUD_NET_ENDPOINT_SIZE (TUD_OPT_HIGH_SPEED ? 512 : 64)
 
-/* Maximum Tranmission Unit (in bytes) of the network, including Ethernet header */
+/* Maximum Transmission Unit (in bytes) of the network, including Ethernet header */
 #ifndef CFG_TUD_NET_MTU
 #define CFG_TUD_NET_MTU           1514
 #endif
 
+#ifndef CFG_TUD_NCM_IN_NTB_MAX_SIZE
+#define CFG_TUD_NCM_IN_NTB_MAX_SIZE 3200
+#endif
+
+#ifndef CFG_TUD_NCM_OUT_NTB_MAX_SIZE
+#define CFG_TUD_NCM_OUT_NTB_MAX_SIZE 3200
+#endif
+
+#ifndef CFG_TUD_NCM_MAX_DATAGRAMS_PER_NTB
+#define CFG_TUD_NCM_MAX_DATAGRAMS_PER_NTB 8
+#endif
+
+#ifndef CFG_TUD_NCM_ALIGNMENT
+#define CFG_TUD_NCM_ALIGNMENT 4
+#endif
+
 #ifdef __cplusplus
  extern "C" {
 #endif
@@ -46,8 +68,18 @@
 // Application API
 //--------------------------------------------------------------------+
 
-// client must provide this: initialize any network state back to the beginning
-void tud_network_init_cb(void);
+// indicate to network driver that client has finished with the packet provided to network_recv_cb()
+void tud_network_recv_renew(void);
+
+// poll network driver for its ability to accept another packet to transmit
+bool tud_network_can_xmit(uint16_t size);
+
+// if network_can_xmit() returns true, network_xmit() can be called once
+void tud_network_xmit(void *ref, uint16_t arg);
+
+//--------------------------------------------------------------------+
+// Application Callbacks (WEAK is optional)
+//--------------------------------------------------------------------+
 
 // client must provide this: return false if the packet buffer was not accepted
 bool tud_network_recv_cb(const uint8_t *src, uint16_t size);
@@ -55,18 +87,19 @@ bool tud_network_recv_cb(const uint8_t *src, uint16_t size);
 // client must provide this: copy from network stack packet pointer to dst
 uint16_t tud_network_xmit_cb(uint8_t *dst, void *ref, uint16_t arg);
 
+//------------- ECM/RNDIS -------------//
+
+// client must provide this: initialize any network state back to the beginning
+void tud_network_init_cb(void);
+
 // client must provide this: 48-bit MAC address
 // TODO removed later since it is not part of tinyusb stack
 extern const uint8_t tud_network_mac_address[6];
 
-// indicate to network driver that client has finished with the packet provided to network_recv_cb()
-void tud_network_recv_renew(void);
-
-// poll network driver for its ability to accept another packet to transmit
-bool tud_network_can_xmit(void);
+//------------- NCM -------------//
 
-// if network_can_xmit() returns true, network_xmit() can be called once
-void tud_network_xmit(void *ref, uint16_t arg);
+// callback to client providing optional indication of internal state of network driver
+void tud_network_link_state_cb(bool state);
 
 //--------------------------------------------------------------------+
 // INTERNAL USBD-CLASS DRIVER API
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/vendor/vendor_device.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/vendor/vendor_device.h
index 844693c68b7..d71c2a3e911 100644
--- a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/vendor/vendor_device.h
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/vendor/vendor_device.h
@@ -44,7 +44,7 @@ bool     tud_vendor_n_mounted         (uint8_t itf);
 
 uint32_t tud_vendor_n_available       (uint8_t itf);
 uint32_t tud_vendor_n_read            (uint8_t itf, void* buffer, uint32_t bufsize);
-bool     tud_vendor_n_peek            (uint8_t itf, uint8_t* u8);
+bool     tud_vendor_n_peek            (uint8_t itf, uint8_t* ui8);
 void     tud_vendor_n_read_flush      (uint8_t itf);
 
 uint32_t tud_vendor_n_write           (uint8_t itf, void const* buffer, uint32_t bufsize);
@@ -59,7 +59,7 @@ uint32_t tud_vendor_n_write_str       (uint8_t itf, char const* str);
 static inline bool     tud_vendor_mounted         (void);
 static inline uint32_t tud_vendor_available       (void);
 static inline uint32_t tud_vendor_read            (void* buffer, uint32_t bufsize);
-static inline bool     tud_vendor_peek            (uint8_t* u8);
+static inline bool     tud_vendor_peek            (uint8_t* ui8);
 static inline void     tud_vendor_read_flush      (void);
 static inline uint32_t tud_vendor_write           (void const* buffer, uint32_t bufsize);
 static inline uint32_t tud_vendor_write_str       (char const* str);
@@ -96,9 +96,9 @@ static inline uint32_t tud_vendor_read (void* buffer, uint32_t bufsize)
   return tud_vendor_n_read(0, buffer, bufsize);
 }
 
-static inline bool tud_vendor_peek (uint8_t* u8)
+static inline bool tud_vendor_peek (uint8_t* ui8)
 {
-  return tud_vendor_n_peek(0, u8);
+  return tud_vendor_n_peek(0, ui8);
 }
 
 static inline void tud_vendor_read_flush(void)
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/video/video.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/video/video.h
new file mode 100644
index 00000000000..b24eb0bcc00
--- /dev/null
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/video/video.h
@@ -0,0 +1,472 @@
+/*
+ * The MIT License (MIT)
+ *
+ * Copyright (c) 2021 Koji KITAYAMA
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ *
+ * This file is part of the TinyUSB stack.
+ */
+
+#ifndef TUSB_VIDEO_H_
+#define TUSB_VIDEO_H_
+
+#include "common/tusb_common.h"
+
+// Table 3-19 Color Matching Descriptor
+typedef enum {
+  VIDEO_COLOR_PRIMARIES_UNDEFINED = 0x00,
+  VIDEO_COLOR_PRIMARIES_BT709, // sRGB (default)
+  VIDEO_COLOR_PRIMARIES_BT470_2M,
+  VIDEO_COLOR_PRIMARIES_BT470_2BG,
+  VIDEO_COLOR_PRIMARIES_SMPTE170M,
+  VIDEO_COLOR_PRIMARIES_SMPTE240M,
+} video_color_primaries_t;
+
+// Table 3-19 Color Matching Descriptor
+typedef enum {
+  VIDEO_COLOR_XFER_CH_UNDEFINED = 0x00,
+  VIDEO_COLOR_XFER_CH_BT709, // default
+  VIDEO_COLOR_XFER_CH_BT470_2M,
+  VIDEO_COLOR_XFER_CH_BT470_2BG,
+  VIDEO_COLOR_XFER_CH_SMPTE170M,
+  VIDEO_COLOR_XFER_CH_SMPTE240M,
+  VIDEO_COLOR_XFER_CH_LINEAR,
+  VIDEO_COLOR_XFER_CH_SRGB,
+} video_color_transfer_characteristics_t;
+
+// Table 3-19 Color Matching Descriptor
+typedef enum {
+  VIDEO_COLOR_COEF_UNDEFINED = 0x00,
+  VIDEO_COLOR_COEF_BT709,
+  VIDEO_COLOR_COEF_FCC,
+  VIDEO_COLOR_COEF_BT470_2BG,
+  VIDEO_COLOR_COEF_SMPTE170M, // BT.601 default
+  VIDEO_COLOR_COEF_SMPTE240M,
+} video_color_matrix_coefficients_t;
+
+/* 4.2.1.2 Request Error Code Control */
+typedef enum {
+  VIDEO_ERROR_NONE = 0, /* The request succeeded. */
+  VIDEO_ERROR_NOT_READY,
+  VIDEO_ERROR_WRONG_STATE,
+  VIDEO_ERROR_POWER,
+  VIDEO_ERROR_OUT_OF_RANGE,
+  VIDEO_ERROR_INVALID_UNIT,
+  VIDEO_ERROR_INVALID_CONTROL,
+  VIDEO_ERROR_INVALID_REQUEST,
+  VIDEO_ERROR_INVALID_VALUE_WITHIN_RANGE,
+  VIDEO_ERROR_UNKNOWN = 0xFF,
+} video_error_code_t;
+
+/* A.2 Interface Subclass */
+typedef enum {
+  VIDEO_SUBCLASS_UNDEFINED = 0x00,
+  VIDEO_SUBCLASS_CONTROL,
+  VIDEO_SUBCLASS_STREAMING,
+  VIDEO_SUBCLASS_INTERFACE_COLLECTION,
+} video_subclass_type_t;
+
+/* A.3 Interface Protocol */
+typedef enum {
+  VIDEO_ITF_PROTOCOL_UNDEFINED = 0x00,
+  VIDEO_ITF_PROTOCOL_15,
+} video_interface_protocol_code_t;
+
+/* A.5 Class-Specific VideoControl Interface Descriptor Subtypes */
+typedef enum {
+  VIDEO_CS_ITF_VC_UNDEFINED = 0x00,
+  VIDEO_CS_ITF_VC_HEADER,
+  VIDEO_CS_ITF_VC_INPUT_TERMINAL,
+  VIDEO_CS_ITF_VC_OUTPUT_TERMINAL,
+  VIDEO_CS_ITF_VC_SELECTOR_UNIT,
+  VIDEO_CS_ITF_VC_PROCESSING_UNIT,
+  VIDEO_CS_ITF_VC_EXTENSION_UNIT,
+  VIDEO_CS_ITF_VC_ENCODING_UNIT,
+  VIDEO_CS_ITF_VC_MAX,
+} video_cs_vc_interface_subtype_t;
+
+/* A.6 Class-Specific VideoStreaming Interface Descriptor Subtypes */
+typedef enum {
+  VIDEO_CS_ITF_VS_UNDEFINED             = 0x00,
+  VIDEO_CS_ITF_VS_INPUT_HEADER          = 0x01,
+  VIDEO_CS_ITF_VS_OUTPUT_HEADER         = 0x02,
+  VIDEO_CS_ITF_VS_STILL_IMAGE_FRAME     = 0x03,
+  VIDEO_CS_ITF_VS_FORMAT_UNCOMPRESSED   = 0x04,
+  VIDEO_CS_ITF_VS_FRAME_UNCOMPRESSED    = 0x05,
+  VIDEO_CS_ITF_VS_FORMAT_MJPEG          = 0x06,
+  VIDEO_CS_ITF_VS_FRAME_MJPEG           = 0x07,
+  VIDEO_CS_ITF_VS_FORMAT_MPEG2TS        = 0x0A,
+  VIDEO_CS_ITF_VS_FORMAT_DV             = 0x0C,
+  VIDEO_CS_ITF_VS_COLORFORMAT           = 0x0D,
+  VIDEO_CS_ITF_VS_FORMAT_FRAME_BASED    = 0x10,
+  VIDEO_CS_ITF_VS_FRAME_FRAME_BASED     = 0x11,
+  VIDEO_CS_ITF_VS_FORMAT_STREAM_BASED   = 0x12,
+  VIDEO_CS_ITF_VS_FORMAT_H264           = 0x13,
+  VIDEO_CS_ITF_VS_FRAME_H264            = 0x14,
+  VIDEO_CS_ITF_VS_FORMAT_H264_SIMULCAST = 0x15,
+  VIDEO_CS_ITF_VS_FORMAT_VP8            = 0x16,
+  VIDEO_CS_ITF_VS_FRAME_VP8             = 0x17,
+  VIDEO_CS_ITF_VS_FORMAT_VP8_SIMULCAST  = 0x18,
+} video_cs_vs_interface_subtype_t;
+
+/* A.7. Class-Specific Endpoint Descriptor Subtypes */
+typedef enum {
+  VIDEO_CS_EP_UNDEFINED = 0x00,
+  VIDEO_CS_EP_GENERAL,
+  VIDEO_CS_EP_ENDPOINT,
+  VIDEO_CS_EP_INTERRUPT
+} video_cs_ep_subtype_t;
+
+/* A.8 Class-Specific Request Codes */
+typedef enum {
+  VIDEO_REQUEST_UNDEFINED   = 0x00,
+  VIDEO_REQUEST_SET_CUR     = 0x01,
+  VIDEO_REQUEST_SET_CUR_ALL = 0x11,
+  VIDEO_REQUEST_GET_CUR     = 0x81,
+  VIDEO_REQUEST_GET_MIN     = 0x82,
+  VIDEO_REQUEST_GET_MAX     = 0x83,
+  VIDEO_REQUEST_GET_RES     = 0x84,
+  VIDEO_REQUEST_GET_LEN     = 0x85,
+  VIDEO_REQUEST_GET_INFO    = 0x86,
+  VIDEO_REQUEST_GET_DEF     = 0x87,
+  VIDEO_REQUEST_GET_CUR_ALL = 0x91,
+  VIDEO_REQUEST_GET_MIN_ALL = 0x92,
+  VIDEO_REQUEST_GET_MAX_ALL = 0x93,
+  VIDEO_REQUEST_GET_RES_ALL = 0x94,
+  VIDEO_REQUEST_GET_DEF_ALL = 0x97
+} video_control_request_t;
+
+/* A.9.1 VideoControl Interface Control Selectors */
+typedef enum {
+  VIDEO_VC_CTL_UNDEFINED = 0x00,
+  VIDEO_VC_CTL_VIDEO_POWER_MODE,
+  VIDEO_VC_CTL_REQUEST_ERROR_CODE,
+} video_interface_control_selector_t;
+
+/* A.9.8 VideoStreaming Interface Control Selectors */
+typedef enum {
+  VIDEO_VS_CTL_UNDEFINED = 0x00,
+  VIDEO_VS_CTL_PROBE,
+  VIDEO_VS_CTL_COMMIT,
+  VIDEO_VS_CTL_STILL_PROBE,
+  VIDEO_VS_CTL_STILL_COMMIT,
+  VIDEO_VS_CTL_STILL_IMAGE_TRIGGER,
+  VIDEO_VS_CTL_STREAM_ERROR_CODE,
+  VIDEO_VS_CTL_GENERATE_KEY_FRAME,
+  VIDEO_VS_CTL_UPDATE_FRAME_SEGMENT,
+  VIDEO_VS_CTL_SYNCH_DELAY_CONTROL,
+} video_interface_streaming_selector_t;
+
+/* B. Terminal Types */
+typedef enum {
+  // Terminal
+  VIDEO_TT_VENDOR_SPECIFIC         = 0x0100,
+  VIDEO_TT_STREAMING               = 0x0101,
+
+  // Input
+  VIDEO_ITT_VENDOR_SPECIFIC        = 0x0200,
+  VIDEO_ITT_CAMERA                 = 0x0201,
+  VIDEO_ITT_MEDIA_TRANSPORT_INPUT  = 0x0202,
+
+  // Output
+  VIDEO_OTT_VENDOR_SPECIFIC        = 0x0300,
+  VIDEO_OTT_DISPLAY                = 0x0301,
+  VIDEO_OTT_MEDIA_TRANSPORT_OUTPUT = 0x0302,
+
+  // External
+  VIDEO_ETT_VENDOR_SPEIFIC         = 0x0400,
+  VIDEO_ETT_COMPOSITE_CONNECTOR    = 0x0401,
+  VIDEO_ETT_SVIDEO_CONNECTOR       = 0x0402,
+  VIDEO_ETT_COMPONENT_CONNECTOR    = 0x0403,
+} video_terminal_type_t;
+
+//--------------------------------------------------------------------+
+// Descriptors
+//--------------------------------------------------------------------+
+
+/* 2.3.4.2 */
+typedef struct TU_ATTR_PACKED {
+  uint8_t  bLength;
+  uint8_t  bDescriptorType;
+  uint8_t  bDescriptorSubType;
+  uint16_t bcdUVC;
+  uint16_t wTotalLength;
+  uint32_t dwClockFrequency;
+  uint8_t  bInCollection;
+  uint8_t  baInterfaceNr[];
+} tusb_desc_cs_video_ctl_itf_hdr_t;
+
+/* 2.4.3.3 */
+typedef struct TU_ATTR_PACKED {
+  uint8_t bHeaderLength;
+  union {
+    uint8_t bmHeaderInfo;
+    struct {
+      uint8_t FrameID:              1;
+      uint8_t EndOfFrame:           1;
+      uint8_t PresentationTime:     1;
+      uint8_t SourceClockReference: 1;
+      uint8_t PayloadSpecific:      1;
+      uint8_t StillImage:           1;
+      uint8_t Error:                1;
+      uint8_t EndOfHeader:          1;
+    };
+  };
+} tusb_video_payload_header_t;
+
+/* 3.9.2.1 */
+typedef struct TU_ATTR_PACKED {
+  uint8_t  bLength;
+  uint8_t  bDescriptorType;
+  uint8_t  bDescriptorSubType;
+  uint8_t  bNumFormats;
+  uint16_t wTotalLength;
+  uint8_t  bEndpointAddress;
+  uint8_t  bmInfo;
+  uint8_t  bTerminalLink;
+  uint8_t  bStillCaptureMethod;
+  uint8_t  bTriggerSupport;
+  uint8_t  bTriggerUsage;
+  uint8_t  bControlSize;
+  uint8_t  bmaControls[];
+} tusb_desc_cs_video_stm_itf_in_hdr_t;
+
+/* 3.9.2.2 */
+typedef struct TU_ATTR_PACKED {
+  uint8_t  bLength;
+  uint8_t  bDescriptorType;
+  uint8_t  bDescriptorSubType;
+  uint8_t  bNumFormats;
+  uint16_t wTotalLength;
+  uint8_t  bEndpointAddress;
+  uint8_t  bTerminalLink;
+  uint8_t  bControlSize;
+  uint8_t  bmaControls[];
+} tusb_desc_cs_video_stm_itf_out_hdr_t;
+
+typedef struct TU_ATTR_PACKED {
+  uint8_t  bLength;
+  uint8_t  bDescriptorType;
+  uint8_t  bDescriptorSubType;
+  uint8_t  bNumFormats;
+  uint16_t wTotalLength;
+  uint8_t  bEndpointAddress;
+  union {
+    struct {
+      uint8_t  bmInfo;
+      uint8_t  bTerminalLink;
+      uint8_t  bStillCaptureMethod;
+      uint8_t  bTriggerSupport;
+      uint8_t  bTriggerUsage;
+      uint8_t  bControlSize;
+      uint8_t  bmaControls[];
+    } input;
+    struct {
+      uint8_t  bEndpointAddress;
+      uint8_t  bTerminalLink;
+      uint8_t  bControlSize;
+      uint8_t  bmaControls[];
+    } output;
+  };
+} tusb_desc_cs_video_stm_itf_hdr_t;
+
+typedef struct TU_ATTR_PACKED {
+  uint8_t bLength;
+  uint8_t bDescriptorType;
+  uint8_t bDescriptorSubType;
+  uint8_t bFormatIndex;
+  uint8_t bNumFrameDescriptors;
+  uint8_t guidFormat[16];
+  uint8_t bBitsPerPixel;
+  uint8_t bDefaultFrameIndex;
+  uint8_t bAspectRatioX;
+  uint8_t bAspectRatioY;
+  uint8_t bmInterlaceFlags;
+  uint8_t bCopyProtect;
+} tusb_desc_cs_video_fmt_uncompressed_t;
+
+typedef struct TU_ATTR_PACKED {
+  uint8_t  bLength;
+  uint8_t  bDescriptorType;
+  uint8_t  bDescriptorSubType;
+  uint8_t  bFrameIndex;
+  uint8_t  bmCapabilities;
+  uint16_t wWidth;
+  uint16_t wHeight;
+  uint32_t dwMinBitRate;
+  uint32_t dwMaxBitRate;
+  uint32_t dwMaxVideoFrameBufferSize; /* deprecated */
+  uint32_t dwDefaultFrameInterval;
+  uint8_t  bFrameIntervalType;
+  uint32_t dwFrameInterval[];
+} tusb_desc_cs_video_frm_uncompressed_t;
+
+//--------------------------------------------------------------------+
+// Requests
+//--------------------------------------------------------------------+
+
+/* 4.3.1.1 */
+typedef struct TU_ATTR_PACKED {
+  union {
+    uint8_t bmHint;
+    struct TU_ATTR_PACKED {
+      uint16_t dwFrameInterval: 1;
+      uint16_t wKeyFrameRatel : 1;
+      uint16_t wPFrameRate    : 1;
+      uint16_t wCompQuality   : 1;
+      uint16_t wCompWindowSize: 1;
+      uint16_t                : 0;
+    } Hint;
+  };
+  uint8_t  bFormatIndex;
+  uint8_t  bFrameIndex;
+  uint32_t dwFrameInterval;
+  uint16_t wKeyFrameRate;
+  uint16_t wPFrameRate;
+  uint16_t wCompQuality;
+  uint16_t wCompWindowSize;
+  uint16_t wDelay;
+  uint32_t dwMaxVideoFrameSize;
+  uint32_t dwMaxPayloadTransferSize;
+  uint32_t dwClockFrequency;
+  union {
+    uint8_t bmFramingInfo;
+    struct TU_ATTR_PACKED {
+      uint8_t FrameID   : 1;
+      uint8_t EndOfFrame: 1;
+      uint8_t EndOfSlice: 1;
+      uint8_t           : 0;
+    } FramingInfo;
+  };
+  uint8_t  bPreferedVersion;
+  uint8_t  bMinVersion;
+  uint8_t  bMaxVersion;
+  uint8_t  bUsage;
+  uint8_t  bBitDepthLuma;
+  uint8_t  bmSettings;
+  uint8_t  bMaxNumberOfRefFramesPlus1;
+  uint16_t bmRateControlModes;
+  uint64_t bmLayoutPerStream;
+} video_probe_and_commit_control_t;
+
+TU_VERIFY_STATIC( sizeof(video_probe_and_commit_control_t) == 48, "size is not correct");
+
+#define TUD_VIDEO_DESC_IAD_LEN                    8
+#define TUD_VIDEO_DESC_STD_VC_LEN                 9
+#define TUD_VIDEO_DESC_CS_VC_LEN                  12
+#define TUD_VIDEO_DESC_INPUT_TERM_LEN             8
+#define TUD_VIDEO_DESC_OUTPUT_TERM_LEN            9
+#define TUD_VIDEO_DESC_STD_VS_LEN                 9
+#define TUD_VIDEO_DESC_CS_VS_IN_LEN               13
+#define TUD_VIDEO_DESC_CS_VS_OUT_LEN              9
+#define TUD_VIDEO_DESC_CS_VS_FMT_UNCOMPR_LEN      27
+#define TUD_VIDEO_DESC_CS_VS_FRM_UNCOMPR_CONT_LEN 38
+#define TUD_VIDEO_DESC_CS_VS_FRM_UNCOMPR_DISC_LEN 26
+#define TUD_VIDEO_DESC_CS_VS_COLOR_MATCHING_LEN   6
+
+/* 2.2 compression formats */
+#define TUD_VIDEO_GUID_YUY2   0x59,0x55,0x59,0x32,0x00,0x00,0x10,0x00,0x00,0x80,0x71,0x9B,0x38,0x00,0xAA,0x00
+#define TUD_VIDEO_GUID_NV12   0x4E,0x56,0x31,0x32,0x00,0x00,0x10,0x00,0x00,0x80,0x71,0x9B,0x38,0x00,0xAA,0x00
+#define TUD_VIDEO_GUID_M420   0x4D,0x34,0x32,0x30,0x00,0x00,0x10,0x00,0x00,0x80,0x71,0x9B,0x38,0x00,0xAA,0x00
+#define TUD_VIDEO_GUID_I420   0x49,0x34,0x32,0x30,0x00,0x00,0x10,0x00,0x00,0x80,0x71,0x9B,0x38,0x00,0xAA,0x00
+
+#define TUD_VIDEO_DESC_IAD(_firstitfs, _nitfs, _stridx) \
+  TUD_VIDEO_DESC_IAD_LEN, TUSB_DESC_INTERFACE_ASSOCIATION, \
+  _firstitfs, _nitfs, TUSB_CLASS_VIDEO, VIDEO_SUBCLASS_INTERFACE_COLLECTION, \
+  VIDEO_ITF_PROTOCOL_UNDEFINED, _stridx
+
+#define TUD_VIDEO_DESC_STD_VC(_itfnum, _nEPs, _stridx) \
+  TUD_VIDEO_DESC_STD_VC_LEN, TUSB_DESC_INTERFACE, _itfnum, /* fixed to zero */ 0x00, \
+  _nEPs, TUSB_CLASS_VIDEO, VIDEO_SUBCLASS_CONTROL, VIDEO_ITF_PROTOCOL_15, _stridx
+
+/* 3.7.2 */
+#define TUD_VIDEO_DESC_CS_VC(_bcdUVC, _totallen, _clkfreq, ...)	\
+  TUD_VIDEO_DESC_CS_VC_LEN + (TU_ARGS_NUM(__VA_ARGS__)), TUSB_DESC_CS_INTERFACE, VIDEO_CS_ITF_VC_HEADER, \
+  U16_TO_U8S_LE(_bcdUVC), U16_TO_U8S_LE((_totallen) + TUD_VIDEO_DESC_CS_VC_LEN + (TU_ARGS_NUM(__VA_ARGS__))), \
+  U32_TO_U8S_LE(_clkfreq), TU_ARGS_NUM(__VA_ARGS__), __VA_ARGS__
+
+/* 3.7.2.1 */
+#define TUD_VIDEO_DESC_INPUT_TERM(_tid, _tt, _at, _stridx) \
+  TUD_VIDEO_DESC_INPUT_TERM_LEN, TUSB_DESC_CS_INTERFACE, VIDEO_CS_ITF_VC_INPUT_TERMINAL, \
+    _tid, U16_TO_U8S_LE(_tt), _at, _stridx
+
+/* 3.7.2.2 */
+#define TUD_VIDEO_DESC_OUTPUT_TERM(_tid, _tt, _at, _srcid, _stridx) \
+  TUD_VIDEO_DESC_OUTPUT_TERM_LEN, TUSB_DESC_CS_INTERFACE, VIDEO_CS_ITF_VC_OUTPUT_TERMINAL, \
+    _tid, U16_TO_U8S_LE(_tt), _at, _srcid, _stridx
+
+/* 3.9.1 */
+#define TUD_VIDEO_DESC_STD_VS(_itfnum, _alt, _epn, _stridx) \
+  TUD_VIDEO_DESC_STD_VS_LEN, TUSB_DESC_INTERFACE, _itfnum, _alt, \
+  _epn, TUSB_CLASS_VIDEO, VIDEO_SUBCLASS_STREAMING, VIDEO_ITF_PROTOCOL_15, _stridx
+
+/* 3.9.2.1 */
+#define TUD_VIDEO_DESC_CS_VS_INPUT(_numfmt, _totallen, _ep, _inf, _termlnk, _sticaptmeth, _trgspt, _trgusg, ...) \
+  TUD_VIDEO_DESC_CS_VS_IN_LEN + (_numfmt) * (TU_ARGS_NUM(__VA_ARGS__)), TUSB_DESC_CS_INTERFACE, \
+  VIDEO_CS_ITF_VS_INPUT_HEADER, _numfmt, \
+  U16_TO_U8S_LE((_totallen) + TUD_VIDEO_DESC_CS_VS_IN_LEN + (_numfmt) * (TU_ARGS_NUM(__VA_ARGS__))), \
+  _ep, _inf, _termlnk, _sticaptmeth, _trgspt, _trgusg, (TU_ARGS_NUM(__VA_ARGS__)), __VA_ARGS__
+
+/* 3.9.2.2 */
+#define TUD_VIDEO_DESC_CS_VS_OUTPUT(_numfmt, _totallen, _ep, _inf, _termlnk, ...) \
+  TUD_VIDEO_DESC_CS_VS_OUT_LEN + (_numfmt) * (TU_ARGS_NUM(__VA_ARGS__)), TUSB_DESC_CS_INTERFACE, \
+  VIDEO_CS_ITF_VS_OUTPUT_HEADER, _numfmt, \
+  U16_TO_U8S_LE((_totallen) + TUD_VIDEO_DESC_CS_VS_OUT_LEN + (_numfmt) * (TU_ARGS_NUM(__VA_ARGS__))), \
+  _ep, _inf, _termlnk, (TU_ARGS_NUM(__VA_ARGS__)), __VA_ARGS__
+
+/* Uncompressed 3.1.1 */
+#define TUD_VIDEO_GUID(_g0,_g1,_g2,_g3,_g4,_g5,_g6,_g7,_g8,_g9,_g10,_g11,_g12,_g13,_g14,_g15) _g0,_g1,_g2,_g3,_g4,_g5,_g6,_g7,_g8,_g9,_g10,_g11,_g12,_g13,_g14,_g15
+
+#define TUD_VIDEO_DESC_CS_VS_FMT_UNCOMPR(_fmtidx, _numfrmdesc, \
+  _guid, _bitsperpix, _frmidx, _asrx, _asry, _interlace, _cp) \
+  TUD_VIDEO_DESC_CS_VS_FMT_UNCOMPR_LEN, TUSB_DESC_CS_INTERFACE, VIDEO_CS_ITF_VS_FORMAT_UNCOMPRESSED, \
+  _fmtidx, _numfrmdesc, TUD_VIDEO_GUID(_guid), \
+  _bitsperpix, _frmidx, _asrx,  _asry, _interlace, _cp
+
+/* Uncompressed 3.1.2 Table 3-3 */
+#define TUD_VIDEO_DESC_CS_VS_FRM_UNCOMPR_CONT(_frmidx, _cap, _width, _height, _minbr, _maxbr, _maxfrmbufsz, _frminterval, _minfrminterval, _maxfrminterval, _frmintervalstep) \
+  TUD_VIDEO_DESC_CS_VS_FRM_UNCOMPR_CONT_LEN, TUSB_DESC_CS_INTERFACE, VIDEO_CS_ITF_VS_FRAME_UNCOMPRESSED, \
+  _frmidx, _cap, U16_TO_U8S_LE(_width), U16_TO_U8S_LE(_height), U32_TO_U8S_LE(_minbr), U32_TO_U8S_LE(_maxbr), \
+  U32_TO_U8S_LE(_maxfrmbufsz), U32_TO_U8S_LE(_frminterval), 0, \
+  U32_TO_U8S_LE(_minfrminterval), U32_TO_U8S_LE(_maxfrminterval), U32_TO_U8S_LE(_frmintervalstep)
+
+/* Uncompressed 3.1.2 Table 3-4 */
+#define TUD_VIDEO_DESC_CS_VS_FRM_UNCOMPR_DISC(_frmidx, _cap, _width, _height, _minbr, _maxbr, _maxfrmbufsz, _frminterval, ...) \
+  TUD_VIDEO_DESC_CS_VS_FRM_UNCOMPR_DISC_LEN + (TU_ARGS_NUM(__VA_ARGS__)) * 4, \
+  TUSB_DESC_CS_INTERFACE, VIDEO_CS_ITF_VS_FRAME_UNCOMPRESSED, \
+  _frmidx, _cap, U16_TO_U8S_LE(_width), U16_TO_U8S_LE(_height), U32_TO_U8S_LE(_minbr), U32_TO_U8S_LE(_maxbr), \
+  U32_TO_U8S_LE(_maxfrmbufsz), U32_TO_U8S_LE(_frminterval), (TU_ARGS_NUM(__VA_ARGS__)), __VA_ARGS__
+
+/* 3.9.2.6 */
+#define TUD_VIDEO_DESC_CS_VS_COLOR_MATCHING(_color, _trns, _mat) \
+  TUD_VIDEO_DESC_CS_VS_COLOR_MATCHING_LEN, \
+  TUSB_DESC_CS_INTERFACE, VIDEO_CS_ITF_VS_COLORFORMAT, \
+  _color, _trns, _mat
+
+/* 3.10.1.1 */
+#define TUD_VIDEO_DESC_EP_ISO(_ep, _epsize, _ep_interval) \
+  7, TUSB_DESC_ENDPOINT, _ep, TUSB_XFER_ISOCHRONOUS | TUSB_ISO_EP_ATT_ASYNCHRONOUS,\
+  U16_TO_U8S_LE(_epsize), _ep_interval
+
+/* 3.10.1.2 */
+#define TUD_VIDEO_DESC_EP_BULK(_ep, _epsize, _ep_interval) \
+  7, TUSB_DESC_ENDPOINT, _ep, TUSB_XFER_BULK, U16_TO_U8S_LE(_epsize), _ep_interval
+
+#endif
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/video/video_device.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/video/video_device.h
new file mode 100644
index 00000000000..ee2fcb9d513
--- /dev/null
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/class/video/video_device.h
@@ -0,0 +1,97 @@
+/*
+ * The MIT License (MIT)
+ *
+ * Copyright (c) 2019 Ha Thach (tinyusb.org)
+ * Copyright (c) 2021 Koji KITAYAMA
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ *
+ * This file is part of the TinyUSB stack.
+ */
+
+#ifndef TUSB_VIDEO_DEVICE_H_
+#define TUSB_VIDEO_DEVICE_H_
+
+#include "common/tusb_common.h"
+#include "video.h"
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+//--------------------------------------------------------------------+
+// Application API (Multiple Ports)
+// CFG_TUD_VIDEO > 1
+//--------------------------------------------------------------------+
+
+/** Return true if streaming
+ *
+ * @param[in] ctl_idx    Destination control interface index
+ * @param[in] stm_idx    Destination streaming interface index */
+bool tud_video_n_streaming(uint_fast8_t ctl_idx, uint_fast8_t stm_idx);
+
+/** Transfer a frame
+ *
+ * @param[in] ctl_idx    Destination control interface index
+ * @param[in] stm_idx    Destination streaming interface index
+ * @param[in] buffer     Frame buffer. The caller must not use this buffer until the operation is completed.
+ * @param[in] bufsize    Byte size of the frame buffer */
+bool tud_video_n_frame_xfer(uint_fast8_t ctl_idx, uint_fast8_t stm_idx, void *buffer, size_t bufsize);
+
+/*------------- Optional callbacks -------------*/
+/** Invoked when compeletion of a frame transfer
+ *
+ * @param[in] ctl_idx    Destination control interface index
+ * @param[in] stm_idx    Destination streaming interface index */
+TU_ATTR_WEAK void tud_video_frame_xfer_complete_cb(uint_fast8_t ctl_idx, uint_fast8_t stm_idx);
+
+//--------------------------------------------------------------------+
+// Application Callback API (weak is optional)
+//--------------------------------------------------------------------+
+
+/** Invoked when SET_POWER_MODE request received
+ *
+ * @param[in] ctl_idx    Destination control interface index
+ * @param[in] stm_idx    Destination streaming interface index
+ * @return video_error_code_t */
+TU_ATTR_WEAK int tud_video_power_mode_cb(uint_fast8_t ctl_idx, uint8_t power_mod);
+
+/** Invoked when VS_COMMIT_CONTROL(SET_CUR) request received
+ *
+ * @param[in] ctl_idx     Destination control interface index
+ * @param[in] stm_idx     Destination streaming interface index
+ * @param[in] parameters  Video streaming parameters
+ * @return video_error_code_t */
+TU_ATTR_WEAK int tud_video_commit_cb(uint_fast8_t ctl_idx, uint_fast8_t stm_idx,
+                                     video_probe_and_commit_control_t const *parameters);
+
+//--------------------------------------------------------------------+
+// INTERNAL USBD-CLASS DRIVER API
+//--------------------------------------------------------------------+
+void     videod_init           (void);
+void     videod_reset          (uint8_t rhport);
+uint16_t videod_open           (uint8_t rhport, tusb_desc_interface_t const * itf_desc, uint16_t max_len);
+bool     videod_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb_control_request_t const * request);
+bool     videod_xfer_cb        (uint8_t rhport, uint8_t ep_addr, xfer_result_t result, uint32_t xferred_bytes);
+
+#ifdef __cplusplus
+ }
+#endif
+
+#endif
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/common/tusb_common.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/common/tusb_common.h
index 1899b35cc3d..fc4d3bc62d3 100644
--- a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/common/tusb_common.h
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/common/tusb_common.h
@@ -38,18 +38,18 @@
 #define TU_MIN(_x, _y)        ( ( (_x) < (_y) ) ? (_x) : (_y) )
 #define TU_MAX(_x, _y)        ( ( (_x) > (_y) ) ? (_x) : (_y) )
 
-#define TU_U16_HIGH(u16)      ((uint8_t) (((u16) >> 8) & 0x00ff))
-#define TU_U16_LOW(u16)       ((uint8_t) ((u16)       & 0x00ff))
-#define U16_TO_U8S_BE(u16)    TU_U16_HIGH(u16), TU_U16_LOW(u16)
-#define U16_TO_U8S_LE(u16)    TU_U16_LOW(u16), TU_U16_HIGH(u16)
+#define TU_U16_HIGH(_u16)      ((uint8_t) (((_u16) >> 8) & 0x00ff))
+#define TU_U16_LOW(_u16)       ((uint8_t) ((_u16)       & 0x00ff))
+#define U16_TO_U8S_BE(_u16)    TU_U16_HIGH(_u16), TU_U16_LOW(_u16)
+#define U16_TO_U8S_LE(_u16)    TU_U16_LOW(_u16), TU_U16_HIGH(_u16)
 
-#define TU_U32_BYTE3(u32)     ((uint8_t) ((((uint32_t) u32) >> 24) & 0x000000ff)) // MSB
-#define TU_U32_BYTE2(u32)     ((uint8_t) ((((uint32_t) u32) >> 16) & 0x000000ff))
-#define TU_U32_BYTE1(u32)     ((uint8_t) ((((uint32_t) u32) >>  8) & 0x000000ff))
-#define TU_U32_BYTE0(u32)     ((uint8_t) (((uint32_t)  u32)        & 0x000000ff)) // LSB
+#define TU_U32_BYTE3(_u32)     ((uint8_t) ((((uint32_t) _u32) >> 24) & 0x000000ff)) // MSB
+#define TU_U32_BYTE2(_u32)     ((uint8_t) ((((uint32_t) _u32) >> 16) & 0x000000ff))
+#define TU_U32_BYTE1(_u32)     ((uint8_t) ((((uint32_t) _u32) >>  8) & 0x000000ff))
+#define TU_U32_BYTE0(_u32)     ((uint8_t) (((uint32_t)  _u32)        & 0x000000ff)) // LSB
 
-#define U32_TO_U8S_BE(u32)    TU_U32_BYTE3(u32), TU_U32_BYTE2(u32), TU_U32_BYTE1(u32), TU_U32_BYTE0(u32)
-#define U32_TO_U8S_LE(u32)    TU_U32_BYTE0(u32), TU_U32_BYTE1(u32), TU_U32_BYTE2(u32), TU_U32_BYTE3(u32)
+#define U32_TO_U8S_BE(_u32)    TU_U32_BYTE3(_u32), TU_U32_BYTE2(_u32), TU_U32_BYTE1(_u32), TU_U32_BYTE0(_u32)
+#define U32_TO_U8S_LE(_u32)    TU_U32_BYTE0(_u32), TU_U32_BYTE1(_u32), TU_U32_BYTE2(_u32), TU_U32_BYTE3(_u32)
 
 #define TU_BIT(n)             (1UL << (n))
 
@@ -105,16 +105,13 @@ TU_ATTR_ALWAYS_INLINE static inline uint16_t tu_u16(uint8_t high, uint8_t low)
   return (uint16_t) ((((uint16_t) high) << 8) | low);
 }
 
-TU_ATTR_ALWAYS_INLINE static inline uint8_t tu_u32_byte3(uint32_t u32) { return TU_U32_BYTE3(u32); }
-TU_ATTR_ALWAYS_INLINE static inline uint8_t tu_u32_byte2(uint32_t u32) { return TU_U32_BYTE2(u32); }
-TU_ATTR_ALWAYS_INLINE static inline uint8_t tu_u32_byte1(uint32_t u32) { return TU_U32_BYTE1(u32); }
-TU_ATTR_ALWAYS_INLINE static inline uint8_t tu_u32_byte0(uint32_t u32) { return TU_U32_BYTE0(u32); }
+TU_ATTR_ALWAYS_INLINE static inline uint8_t tu_u32_byte3(uint32_t ui32) { return TU_U32_BYTE3(ui32); }
+TU_ATTR_ALWAYS_INLINE static inline uint8_t tu_u32_byte2(uint32_t ui32) { return TU_U32_BYTE2(ui32); }
+TU_ATTR_ALWAYS_INLINE static inline uint8_t tu_u32_byte1(uint32_t ui32) { return TU_U32_BYTE1(ui32); }
+TU_ATTR_ALWAYS_INLINE static inline uint8_t tu_u32_byte0(uint32_t ui32) { return TU_U32_BYTE0(ui32); }
 
-TU_ATTR_ALWAYS_INLINE static inline uint16_t tu_u32_high16(uint32_t u32) { return (uint16_t) (u32 >> 16); }
-TU_ATTR_ALWAYS_INLINE static inline uint16_t tu_u32_low16 (uint32_t u32) { return (uint16_t) (u32 & 0x0000ffffu); }
-
-TU_ATTR_ALWAYS_INLINE static inline uint8_t tu_u16_high(uint16_t u16) { return TU_U16_HIGH(u16); }
-TU_ATTR_ALWAYS_INLINE static inline uint8_t tu_u16_low (uint16_t u16) { return TU_U16_LOW(u16); }
+TU_ATTR_ALWAYS_INLINE static inline uint8_t tu_u16_high(uint16_t ui16) { return TU_U16_HIGH(ui16); }
+TU_ATTR_ALWAYS_INLINE static inline uint8_t tu_u16_low (uint16_t ui16) { return TU_U16_LOW(ui16); }
 
 //------------- Bits -------------//
 TU_ATTR_ALWAYS_INLINE static inline uint32_t tu_bit_set  (uint32_t value, uint8_t pos) { return value | TU_BIT(pos);                  }
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/common/tusb_types.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/common/tusb_types.h
index f26983a7483..897e89b3943 100644
--- a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/common/tusb_types.h
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/common/tusb_types.h
@@ -69,6 +69,18 @@ typedef enum
   TUSB_DIR_IN_MASK = 0x80
 }tusb_dir_t;
 
+/// Isochronous End Point Attributes
+typedef enum
+{
+  TUSB_ISO_EP_ATT_NO_SYNC         = 0x00,
+  TUSB_ISO_EP_ATT_ASYNCHRONOUS    = 0x04,
+  TUSB_ISO_EP_ATT_ADAPTIVE        = 0x08,
+  TUSB_ISO_EP_ATT_SYNCHRONOUS     = 0x0C,
+  TUSB_ISO_EP_ATT_DATA            = 0x00, ///< Data End Point
+  TUSB_ISO_EP_ATT_EXPLICIT_FB     = 0x10, ///< Feedback End Point
+  TUSB_ISO_EP_ATT_IMPLICIT_FB     = 0x20, ///< Data endpoint that also serves as an implicit feedback
+}tusb_iso_ep_attribute_t;
+
 /// USB Descriptor Types
 typedef enum
 {
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/device/dcd_attr.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/device/dcd_attr.h
index a35fc0ac502..44f2cc1d6d0 100644
--- a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/device/dcd_attr.h
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/device/dcd_attr.h
@@ -63,6 +63,9 @@
 #elif TU_CHECK_MCU(MKL25ZXX) || TU_CHECK_MCU(K32L2BXX)
   #define DCD_ATTR_ENDPOINT_MAX   16
 
+#elif TU_CHECK_MCU(MM32F327X)
+  #define DCD_ATTR_ENDPOINT_MAX   16
+
 //------------- Nordic -------------//
 #elif TU_CHECK_MCU(NRF5X)
   // 8 CBI + 1 ISO
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/device/usbd.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/device/usbd.h
index 638d9309470..8d02de1ff1f 100644
--- a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/device/usbd.h
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/device/usbd.h
@@ -178,17 +178,18 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
   0x9C, 0xD2, 0x65, 0x9D, 0x9E, 0x64, 0x8A, 0x9F
 
 //--------------------------------------------------------------------+
-// Configuration & Interface Descriptor Templates
+// Configuration Descriptor Templates
 //--------------------------------------------------------------------+
 
-//------------- Configuration -------------//
 #define TUD_CONFIG_DESC_LEN   (9)
 
 // Config number, interface count, string index, total length, attribute, power in mA
 #define TUD_CONFIG_DESCRIPTOR(config_num, _itfcount, _stridx, _total_len, _attribute, _power_ma) \
   9, TUSB_DESC_CONFIGURATION, U16_TO_U8S_LE(_total_len), _itfcount, config_num, _stridx, TU_BIT(7) | _attribute, (_power_ma)/2
 
-//------------- CDC -------------//
+//--------------------------------------------------------------------+
+// CDC Descriptor Templates
+//--------------------------------------------------------------------+
 
 // Length of template descriptor: 66 bytes
 #define TUD_CDC_DESC_LEN  (8+9+5+5+4+5+7+9+7+7)
@@ -217,7 +218,9 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
   /* Endpoint In */\
   7, TUSB_DESC_ENDPOINT, _epin, TUSB_XFER_BULK, U16_TO_U8S_LE(_epsize), 0
 
-//------------- MSC -------------//
+//--------------------------------------------------------------------+
+// MSC Descriptor Templates
+//--------------------------------------------------------------------+
 
 // Length of template descriptor: 23 bytes
 #define TUD_MSC_DESC_LEN    (9 + 7 + 7)
@@ -231,7 +234,10 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
   /* Endpoint In */\
   7, TUSB_DESC_ENDPOINT, _epin, TUSB_XFER_BULK, U16_TO_U8S_LE(_epsize), 0
 
-//------------- HID -------------//
+
+//--------------------------------------------------------------------+
+// HID Descriptor Templates
+//--------------------------------------------------------------------+
 
 // Length of template descriptor: 25 bytes
 #define TUD_HID_DESC_LEN    (9 + 9 + 7)
@@ -261,8 +267,10 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
   /* Endpoint In */\
   7, TUSB_DESC_ENDPOINT, _epin, TUSB_XFER_INTERRUPT, U16_TO_U8S_LE(_epsize), _ep_interval
 
-//------------- MIDI -------------//
-// MIDI v1.0 is based on Audio v1.0
+//--------------------------------------------------------------------+
+// MIDI Descriptor Templates
+// Note: MIDI v1.0 is based on Audio v1.0
+//--------------------------------------------------------------------+
 
 #define TUD_MIDI_DESC_HEAD_LEN (9 + 9 + 9 + 7)
 #define TUD_MIDI_DESC_HEAD(_itfnum,  _stridx, _numcables) \
@@ -319,7 +327,9 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
   TUD_MIDI_DESC_EP(_epin, _epsize, 1),\
   TUD_MIDI_JACKID_OUT_EMB(1)
 
-//------------- AUDIO -------------//
+//--------------------------------------------------------------------+
+// Audio v2.0 Descriptor Templates
+//--------------------------------------------------------------------+
 
 /* Standard Interface Association Descriptor (IAD) */
 #define TUD_AUDIO_DESC_IAD_LEN 8
@@ -551,7 +561,10 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
     ((((_maxFrequency + ((CFG_TUSB_RHPORT0_MODE & OPT_MODE_HIGH_SPEED) ? 7999 : 999)) / ((CFG_TUSB_RHPORT0_MODE & OPT_MODE_HIGH_SPEED) ? 8000 : 1000)) + 1) * _nBytesPerSample * _nChannels)
 
 
-//------------- TUD_USBTMC/USB488 -------------//
+//--------------------------------------------------------------------+
+// USBTMC/USB488 Descriptor Templates
+//--------------------------------------------------------------------+
+
 #define TUD_USBTMC_APP_CLASS    (TUSB_CLASS_APPLICATION_SPECIFIC)
 #define TUD_USBTMC_APP_SUBCLASS 0x03u
 
@@ -581,8 +594,10 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
 
 #define TUD_USBTMC_INT_DESCRIPTOR_LEN (7u)
 
+//--------------------------------------------------------------------+
+// Vendor Descriptor Templates
+//--------------------------------------------------------------------+
 
-//------------- Vendor -------------//
 #define TUD_VENDOR_DESC_LEN  (9+7+7)
 
 // Interface number, string index, EP Out & IN address, EP size
@@ -594,7 +609,10 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
   /* Endpoint In */\
   7, TUSB_DESC_ENDPOINT, _epin, TUSB_XFER_BULK, U16_TO_U8S_LE(_epsize), 0
 
-//------------- DFU Runtime -------------//
+//--------------------------------------------------------------------+
+// DFU Runtime Descriptor Templates
+//--------------------------------------------------------------------+
+
 #define TUD_DFU_APP_CLASS    (TUSB_CLASS_APPLICATION_SPECIFIC)
 #define TUD_DFU_APP_SUBCLASS (APP_SUBCLASS_DFU_RUNTIME)
 
@@ -609,6 +627,10 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
   /* Function */ \
   9, DFU_DESC_FUNCTIONAL, _attr, U16_TO_U8S_LE(_timeout), U16_TO_U8S_LE(_xfer_size), U16_TO_U8S_LE(0x0101)
 
+//--------------------------------------------------------------------+
+// DFU Descriptor Templates
+//--------------------------------------------------------------------+
+
 // Length of template descriptor: 9 bytes + number of alternatives * 9
 #define TUD_DFU_DESC_LEN(_alt_count)    (9 + (_alt_count) * 9)
 
@@ -654,8 +676,9 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
   _TUD_DFU_ALT(_itfnum, _alt_count, _stridx),      \
   _TUD_DFU_ALT_7(_itfnum, _alt_count+1, _stridx+1)
 
-
-//------------- CDC-ECM -------------//
+//--------------------------------------------------------------------+
+// CDC-ECM Descriptor Templates
+//--------------------------------------------------------------------+
 
 // Length of template descriptor: 71 bytes
 #define TUD_CDC_ECM_DESC_LEN  (8+9+5+5+13+7+9+9+7+7)
@@ -684,8 +707,9 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
   /* Endpoint Out */\
   7, TUSB_DESC_ENDPOINT, _epout, TUSB_XFER_BULK, U16_TO_U8S_LE(_epsize), 0
 
-
-//------------- RNDIS -------------//
+//--------------------------------------------------------------------+
+// RNDIS Descriptor Templates
+//--------------------------------------------------------------------+
 
 #if 0
 /* Windows XP */
@@ -726,7 +750,10 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
   /* Endpoint Out */\
   7, TUSB_DESC_ENDPOINT, _epout, TUSB_XFER_BULK, U16_TO_U8S_LE(_epsize), 0
 
-//------------- BT Radio -------------//
+//--------------------------------------------------------------------+
+// Bluetooth Radio Descriptor Templates
+//--------------------------------------------------------------------+
+
 #define TUD_BT_APP_CLASS                    (TUSB_CLASS_WIRELESS_CONTROLLER)
 #define TUD_BT_APP_SUBCLASS                 0x01
 #define TUD_BT_PROTOCOL_PRIMARY_CONTROLLER  0x01
@@ -777,10 +804,44 @@ TU_ATTR_WEAK bool tud_vendor_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb
 
 // BT Primary controller descriptor
 // Interface number, string index, attributes, event endpoint, event endpoint size, interval, data in, data out, data endpoint size, iso endpoint sizes
+// TODO BTH should also use IAD like CDC for composite device
 #define TUD_BTH_DESCRIPTOR(_itfnum, _stridx, _ep_evt, _ep_evt_size, _ep_evt_interval, _ep_in, _ep_out, _ep_size,...) \
   TUD_BTH_PRI_ITF(_itfnum, _stridx, _ep_evt, _ep_evt_size, _ep_evt_interval, _ep_in, _ep_out, _ep_size) \
   TUD_BTH_ISO_ITFS(_itfnum + 1, _ep_in + 1, _ep_out + 1, __VA_ARGS__)
 
+//--------------------------------------------------------------------+
+// CDC-NCM Descriptor Templates
+//--------------------------------------------------------------------+
+
+// Length of template descriptor
+#define TUD_CDC_NCM_DESC_LEN  (8+9+5+5+13+6+7+9+9+7+7)
+
+// CDC-ECM Descriptor Template
+// Interface number, description string index, MAC address string index, EP notification address and size, EP data address (out, in), and size, max segment size.
+#define TUD_CDC_NCM_DESCRIPTOR(_itfnum, _desc_stridx, _mac_stridx, _ep_notif, _ep_notif_size, _epout, _epin, _epsize, _maxsegmentsize) \
+  /* Interface Association */\
+  8, TUSB_DESC_INTERFACE_ASSOCIATION, _itfnum, 2, TUSB_CLASS_CDC, CDC_COMM_SUBCLASS_NETWORK_CONTROL_MODEL, 0, 0,\
+  /* CDC Control Interface */\
+  9, TUSB_DESC_INTERFACE, _itfnum, 0, 1, TUSB_CLASS_CDC, CDC_COMM_SUBCLASS_NETWORK_CONTROL_MODEL, 0, _desc_stridx,\
+  /* CDC-NCM Header */\
+  5, TUSB_DESC_CS_INTERFACE, CDC_FUNC_DESC_HEADER, U16_TO_U8S_LE(0x0110),\
+  /* CDC-NCM Union */\
+  5, TUSB_DESC_CS_INTERFACE, CDC_FUNC_DESC_UNION, _itfnum, (uint8_t)((_itfnum) + 1),\
+  /* CDC-NCM Functional Descriptor */\
+  13, TUSB_DESC_CS_INTERFACE, CDC_FUNC_DESC_ETHERNET_NETWORKING, _mac_stridx, 0, 0, 0, 0, U16_TO_U8S_LE(_maxsegmentsize), U16_TO_U8S_LE(0), 0, \
+  /* CDC-NCM Functional Descriptor */\
+  6, TUSB_DESC_CS_INTERFACE, CDC_FUNC_DESC_NCM, U16_TO_U8S_LE(0x0100), 0, \
+  /* Endpoint Notification */\
+  7, TUSB_DESC_ENDPOINT, _ep_notif, TUSB_XFER_INTERRUPT, U16_TO_U8S_LE(_ep_notif_size), 50,\
+  /* CDC Data Interface (default inactive) */\
+  9, TUSB_DESC_INTERFACE, (uint8_t)((_itfnum)+1), 0, 0, TUSB_CLASS_CDC_DATA, 0, NCM_DATA_PROTOCOL_NETWORK_TRANSFER_BLOCK, 0,\
+  /* CDC Data Interface (alternative active) */\
+  9, TUSB_DESC_INTERFACE, (uint8_t)((_itfnum)+1), 1, 2, TUSB_CLASS_CDC_DATA, 0, NCM_DATA_PROTOCOL_NETWORK_TRANSFER_BLOCK, 0,\
+  /* Endpoint In */\
+  7, TUSB_DESC_ENDPOINT, _epin, TUSB_XFER_BULK, U16_TO_U8S_LE(_epsize), 0,\
+  /* Endpoint Out */\
+  7, TUSB_DESC_ENDPOINT, _epout, TUSB_XFER_BULK, U16_TO_U8S_LE(_epsize), 0
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/tusb.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/tusb.h
index b52f8839ac8..0d29e106c9b 100644
--- a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/tusb.h
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/tusb.h
@@ -76,9 +76,13 @@
     #include "class/msc/msc_device.h"
   #endif
 
-#if CFG_TUD_AUDIO
-  #include "class/audio/audio_device.h"
-#endif
+  #if CFG_TUD_AUDIO
+    #include "class/audio/audio_device.h"
+  #endif
+
+  #if CFG_TUD_VIDEO
+    #include "class/video/video_device.h"
+  #endif
 
   #if CFG_TUD_MIDI
     #include "class/midi/midi_device.h"
@@ -100,7 +104,7 @@
     #include "class/dfu/dfu_device.h"
   #endif
 
-  #if CFG_TUD_NET
+  #if CFG_TUD_ECM_RNDIS || CFG_TUD_NCM
     #include "class/net/net_device.h"
   #endif
 
diff --git a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/tusb_option.h b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/tusb_option.h
index ad2a26a6654..b2ebfdf66de 100644
--- a/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/tusb_option.h
+++ b/tools/sdk/esp32s2/include/arduino_tinyusb/tinyusb/src/tusb_option.h
@@ -241,6 +241,10 @@
   #define CFG_TUD_AUDIO           0
 #endif
 
+#ifndef CFG_TUD_VIDEO
+  #define CFG_TUD_VIDEO           0
+#endif
+
 #ifndef CFG_TUD_MIDI
   #define CFG_TUD_MIDI            0
 #endif
@@ -261,14 +265,23 @@
   #define CFG_TUD_DFU             0
 #endif
 
-#ifndef CFG_TUD_NET
-  #define CFG_TUD_NET             0
-#endif
-
 #ifndef CFG_TUD_BTH
   #define CFG_TUD_BTH             0
 #endif
 
+#ifndef CFG_TUD_ECM_RNDIS
+  #ifdef CFG_TUD_NET
+    #warning "CFG_TUD_NET is renamed to CFG_TUD_ECM_RNDIS"
+    #define CFG_TUD_ECM_RNDIS   CFG_TUD_NET
+  #else
+    #define CFG_TUD_ECM_RNDIS   0
+  #endif
+#endif
+
+#ifndef CFG_TUD_NCM
+  #define CFG_TUD_NCM         0
+#endif
+
 //--------------------------------------------------------------------
 // HOST OPTIONS
 //--------------------------------------------------------------------
@@ -280,10 +293,34 @@
   #ifndef CFG_TUH_ENUMERATION_BUFSIZE
     #define CFG_TUH_ENUMERATION_BUFSIZE 256
   #endif
-
-  //------------- CLASS -------------//
 #endif // TUSB_OPT_HOST_ENABLED
 
+//------------- CLASS -------------//
+
+#ifndef CFG_TUH_HUB
+#define CFG_TUH_HUB    0
+#endif
+
+#ifndef CFG_TUH_CDC
+#define CFG_TUH_CDC    0
+#endif
+
+#ifndef CFG_TUH_HID
+#define CFG_TUH_HID    0
+#endif
+
+#ifndef CFG_TUH_MIDI
+#define CFG_TUH_MIDI   0
+#endif
+
+#ifndef CFG_TUH_MSC
+#define CFG_TUH_MSC    0
+#endif
+
+#ifndef CFG_TUH_VENDOR
+#define CFG_TUH_VENDOR 0
+#endif
+
 //--------------------------------------------------------------------+
 // Port Specific
 // TUP stand for TinyUSB Port (can be renamed)
diff --git a/tools/sdk/esp32s2/include/asio/port/include/esp_asio_config.h b/tools/sdk/esp32s2/include/asio/port/include/esp_asio_config.h
index cba316527e6..3f3a9b03ed4 100644
--- a/tools/sdk/esp32s2/include/asio/port/include/esp_asio_config.h
+++ b/tools/sdk/esp32s2/include/asio/port/include/esp_asio_config.h
@@ -18,6 +18,11 @@
 #  define ASIO_NO_TYPEID
 # endif   // CONFIG_COMPILER_RTTI
 
+//
+// Supress OpenSSL deprecation warning, when building ASIO
+//
+#define ESP_OPENSSL_SUPPRESS_LEGACY_WARNING
+
 //
 // LWIP compatibility inet and address macros/functions
 //
diff --git a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_block_internal.h b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_block_internal.h
index 9abe81557fa..b7ad0a554cc 100644
--- a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_block_internal.h
+++ b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_block_internal.h
@@ -191,6 +191,9 @@ int coap_handle_response_get_block(coap_context_t *context,
 void coap_block_delete_lg_xmit(coap_session_t *session,
                                coap_lg_xmit_t *lg_xmit);
 
+coap_tick_t coap_block_check_lg_xmit_timeouts(coap_session_t *session,
+                                              coap_tick_t now);
+
 /**
  * The function that does all the work for the coap_add_data_large*()
  * functions.
diff --git a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_dtls.h b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_dtls.h
index cbd369dfce6..fc30445431d 100644
--- a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_dtls.h
+++ b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_dtls.h
@@ -27,6 +27,12 @@ typedef struct coap_dtls_pki_t coap_dtls_pki_t;
 #ifndef COAP_DTLS_HINT_LENGTH
 #define COAP_DTLS_HINT_LENGTH 128
 #endif
+#ifndef COAP_DTLS_MAX_PSK_IDENTITY
+#define COAP_DTLS_MAX_PSK_IDENTITY 64
+#endif
+#ifndef COAP_DTLS_MAX_PSK
+#define COAP_DTLS_MAX_PSK 64
+#endif
 
 typedef enum coap_dtls_role_t {
   COAP_DTLS_ROLE_CLIENT, /**< Internal function invoked for client */
diff --git a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_event.h b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_event.h
index 89b2a63967c..b6ea60524a2 100644
--- a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_event.h
+++ b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_event.h
@@ -24,34 +24,34 @@
  * Scalar type to represent different events, e.g. DTLS events or
  * retransmission timeouts.
  */
- typedef unsigned int coap_event_t;
-
+typedef enum coap_event_t {
 /**
  * (D)TLS events for COAP_PROTO_DTLS and COAP_PROTO_TLS
  */
-#define COAP_EVENT_DTLS_CLOSED        0x0000
-#define COAP_EVENT_DTLS_CONNECTED     0x01DE
-#define COAP_EVENT_DTLS_RENEGOTIATE   0x01DF
-#define COAP_EVENT_DTLS_ERROR         0x0200
+  COAP_EVENT_DTLS_CLOSED       = 0x0000,
+  COAP_EVENT_DTLS_CONNECTED    = 0x01DE,
+  COAP_EVENT_DTLS_RENEGOTIATE  = 0x01DF,
+  COAP_EVENT_DTLS_ERROR        = 0x0200,
 
 /**
  * TCP events for COAP_PROTO_TCP and COAP_PROTO_TLS
  */
-#define COAP_EVENT_TCP_CONNECTED      0x1001
-#define COAP_EVENT_TCP_CLOSED         0x1002
-#define COAP_EVENT_TCP_FAILED         0x1003
+  COAP_EVENT_TCP_CONNECTED     = 0x1001,
+  COAP_EVENT_TCP_CLOSED        = 0x1002,
+  COAP_EVENT_TCP_FAILED        = 0x1003,
 
 /**
  * CSM exchange events for reliable protocols only
  */
-#define COAP_EVENT_SESSION_CONNECTED  0x2001
-#define COAP_EVENT_SESSION_CLOSED     0x2002
-#define COAP_EVENT_SESSION_FAILED     0x2003
+  COAP_EVENT_SESSION_CONNECTED = 0x2001,
+  COAP_EVENT_SESSION_CLOSED    = 0x2002,
+  COAP_EVENT_SESSION_FAILED    = 0x2003,
 
 /**
- * BLOCK2 receive errors
+ * (Q-)BLOCK receive errors
  */
-#define COAP_EVENT_PARTIAL_BLOCK      0x3001
+  COAP_EVENT_PARTIAL_BLOCK     = 0x3001
+} coap_event_t;
 
 /**
  * Type for event handler functions that can be registered with a CoAP
diff --git a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_time.h b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_time.h
index baa8650eaff..99d117a4eb7 100644
--- a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_time.h
+++ b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/coap_time.h
@@ -88,7 +88,11 @@ COAP_STATIC_INLINE uint64_t coap_ticks_to_rt_us(coap_tick_t t) {
 #elif defined(RIOT_VERSION)
 #include <xtimer.h>
 
+#ifdef XTIMER_HZ
 #define COAP_TICKS_PER_SECOND (XTIMER_HZ)
+#else /* XTIMER_HZ */
+#define COAP_TICKS_PER_SECOND (XTIMER_HZ_BASE)
+#endif /* XTIMER_HZ */
 
 typedef uint64_t coap_tick_t;
 typedef int64_t coap_tick_diff_t;
diff --git a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/net.h b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/net.h
index 577a0b5d5a5..ea5a2cba1bb 100644
--- a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/net.h
+++ b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/net.h
@@ -15,6 +15,7 @@
 #include <stdlib.h>
 #include <string.h>
 #ifndef _WIN32
+#include <sys/select.h>
 #include <sys/time.h>
 #endif
 #include <time.h>
diff --git a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/pdu.h b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/pdu.h
index a22869b69ed..8031a1c2c40 100644
--- a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/pdu.h
+++ b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/pdu.h
@@ -299,7 +299,6 @@ typedef enum coap_pdu_code_t {
   COAP_REQUEST_CODE_PATCH  = COAP_REQUEST_PATCH,
   COAP_REQUEST_CODE_IPATCH = COAP_REQUEST_IPATCH,
 
-  COAP_RESPONSE_CODE_OK                         = COAP_RESPONSE_CODE(200),
   COAP_RESPONSE_CODE_CREATED                    = COAP_RESPONSE_CODE(201),
   COAP_RESPONSE_CODE_DELETED                    = COAP_RESPONSE_CODE(202),
   COAP_RESPONSE_CODE_VALID                      = COAP_RESPONSE_CODE(203),
diff --git a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/resource.h b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/resource.h
index 2cd9aea48fa..5cf7751f67e 100644
--- a/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/resource.h
+++ b/tools/sdk/esp32s2/include/coap/libcoap/include/coap3/resource.h
@@ -83,7 +83,8 @@ typedef void (*coap_method_handler_t)
  * variable of coap_str_const_t has to point to constant text, or point to data
  * within the allocated coap_str_const_t parameter.
  *
- * @param uri_path The string URI path of the new resource.
+ * @param uri_path The string URI path of the new resource. The leading '/' is
+ *                 not normally required - e.g. just "full/path/for/resource".
  * @param flags    Flags for memory management (in particular release of
  *                 memory). Possible values:@n
  *
diff --git a/tools/sdk/esp32s2/include/config/sdkconfig.h b/tools/sdk/esp32s2/include/config/sdkconfig.h
index 48660a5006f..7bad1c4fa50 100644
--- a/tools/sdk/esp32s2/include/config/sdkconfig.h
+++ b/tools/sdk/esp32s2/include/config/sdkconfig.h
@@ -30,6 +30,7 @@
 #define CONFIG_BOOT_ROM_LOG_ALWAYS_ON 1
 #define CONFIG_ESPTOOLPY_BAUD_OTHER_VAL 115200
 #define CONFIG_ESPTOOLPY_FLASHMODE_QIO 1
+#define CONFIG_ESPTOOLPY_FLASH_SAMPLE_MODE_STR 1
 #define CONFIG_ESPTOOLPY_FLASHMODE "dio"
 #define CONFIG_ESPTOOLPY_FLASHFREQ_80M 1
 #define CONFIG_ESPTOOLPY_FLASHFREQ "80m"
@@ -70,7 +71,7 @@
 #define CONFIG_TINYUSB_CDC_TX_BUFSIZE 64
 #define CONFIG_TINYUSB_MSC_ENABLED 1
 #define CONFIG_TINYUSB_DESC_MSC_STRING "Espressif MSC Device"
-#define CONFIG_TINYUSB_MSC_BUFSIZE 512
+#define CONFIG_TINYUSB_MSC_BUFSIZE 4096
 #define CONFIG_TINYUSB_HID_ENABLED 1
 #define CONFIG_TINYUSB_DESC_HID_STRING "Espressif HID Device"
 #define CONFIG_TINYUSB_HID_BUFSIZE 64
@@ -78,6 +79,10 @@
 #define CONFIG_TINYUSB_DESC_MIDI_STRING "Espressif MIDI Device"
 #define CONFIG_TINYUSB_MIDI_RX_BUFSIZE 64
 #define CONFIG_TINYUSB_MIDI_TX_BUFSIZE 64
+#define CONFIG_TINYUSB_VIDEO_ENABLED 1
+#define CONFIG_TINYUSB_DESC_VIDEO_STRING "Espressif VIDEO Device"
+#define CONFIG_TINYUSB_VIDEO_STREAMING_BUFSIZE 64
+#define CONFIG_TINYUSB_VIDEO_STREAMING_IFS 1
 #define CONFIG_TINYUSB_DFU_RT_ENABLED 1
 #define CONFIG_TINYUSB_DESC_DFU_RT_STRING "Espressif DFU Device"
 #define CONFIG_TINYUSB_VENDOR_ENABLED 1
@@ -289,6 +294,7 @@
 #define CONFIG_LWIP_GARP_TMR_INTERVAL 60
 #define CONFIG_LWIP_TCPIP_RECVMBOX_SIZE 32
 #define CONFIG_LWIP_DHCP_RESTORE_LAST_IP 1
+#define CONFIG_LWIP_DHCP_OPTIONS_LEN 68
 #define CONFIG_LWIP_DHCPS 1
 #define CONFIG_LWIP_DHCPS_LEASE_UNIT 60
 #define CONFIG_LWIP_DHCPS_MAX_STATION_NUM 8
@@ -401,6 +407,7 @@
 #define CONFIG_MDNS_TASK_AFFINITY 0x0
 #define CONFIG_MDNS_SERVICE_ADD_TIMEOUT_MS 2000
 #define CONFIG_MDNS_TIMER_PERIOD_MS 100
+#define CONFIG_MDNS_MULTIPLE_INSTANCE 1
 #define CONFIG_MQTT_PROTOCOL_311 1
 #define CONFIG_MQTT_TRANSPORT_SSL 1
 #define CONFIG_MQTT_TRANSPORT_WEBSOCKET 1
@@ -575,5 +582,5 @@
 #define CONFIG_USB_MSC_BUFSIZE CONFIG_TINYUSB_MSC_BUFSIZE
 #define CONFIG_USB_MSC_ENABLED CONFIG_TINYUSB_MSC_ENABLED
 #define CONFIG_WARN_WRITE_STRINGS CONFIG_COMPILER_WARN_WRITE_STRINGS
-#define CONFIG_ARDUINO_IDF_COMMIT "3e370c4296"
+#define CONFIG_ARDUINO_IDF_COMMIT "b86fe0c66c"
 #define CONFIG_ARDUINO_IDF_BRANCH "master"
diff --git a/tools/sdk/esp32s2/include/driver/esp32s2/include/driver/touch_sensor.h b/tools/sdk/esp32s2/include/driver/esp32s2/include/driver/touch_sensor.h
index 178e3b1a9dd..d90a56e992c 100644
--- a/tools/sdk/esp32s2/include/driver/esp32s2/include/driver/touch_sensor.h
+++ b/tools/sdk/esp32s2/include/driver/esp32s2/include/driver/touch_sensor.h
@@ -293,7 +293,7 @@ esp_err_t touch_pad_reset_benchmark(touch_pad_t touch_num);
  * @return
  *     - ESP_OK Success
  */
-esp_err_t touch_pad_filter_set_config(touch_filter_config_t *filter_info);
+esp_err_t touch_pad_filter_set_config(const touch_filter_config_t *filter_info);
 
 /**
  * @brief get parameter of touch sensor filter and detection algorithm.
@@ -331,7 +331,7 @@ esp_err_t touch_pad_filter_disable(void);
  * @return
  *     - ESP_OK Success
  */
-esp_err_t touch_pad_denoise_set_config(touch_pad_denoise_t *denoise);
+esp_err_t touch_pad_denoise_set_config(const touch_pad_denoise_t *denoise);
 
 /**
  * @brief get parameter of denoise pad (TOUCH_PAD_NUM0).
@@ -380,7 +380,7 @@ esp_err_t touch_pad_denoise_read_data(uint32_t *data);
  * @return
  *     - ESP_OK Success
  */
-esp_err_t touch_pad_waterproof_set_config(touch_pad_waterproof_t *waterproof);
+esp_err_t touch_pad_waterproof_set_config(const touch_pad_waterproof_t *waterproof);
 
 /**
  * @brief get parameter of waterproof function.
diff --git a/tools/sdk/esp32s2/include/driver/include/driver/rmt.h b/tools/sdk/esp32s2/include/driver/include/driver/rmt.h
index a7e2aad5f24..bc07954a080 100644
--- a/tools/sdk/esp32s2/include/driver/include/driver/rmt.h
+++ b/tools/sdk/esp32s2/include/driver/include/driver/rmt.h
@@ -856,16 +856,35 @@ esp_err_t rmt_remove_channel_from_group(rmt_channel_t channel);
 
 #if SOC_RMT_SUPPORT_TX_LOOP_COUNT
 /**
- * @brief Set loop count for RMT TX channel
+ * @brief Set loop count threshold value for RMT TX channel
+ *
+ * When tx loop count reaches this value, an ISR callback will notify user
  *
  * @param channel RMT channel
- * @param count loop count
+ * @param count loop count, 1 ~ 1023
  * @return
  *      - ESP_ERR_INVALID_ARG Parameter error
  *      - ESP_OK Success
  */
 esp_err_t rmt_set_tx_loop_count(rmt_channel_t channel, uint32_t count);
-#endif
+
+/**
+ * @brief Enable or disable the feature that when loop count reaches the threshold, RMT will stop transmitting.
+ *
+ * - When the loop auto-stop feature is enabled will halt RMT transmission after the loop count reaches a certain threshold
+ * - When disabled, the RMT transmission continue indefinitely until halted by the users
+ *
+ * @note The auto-stop feature is implemented in hardware on particular targets (i.e. those with SOC_RMT_SUPPORT_TX_LOOP_AUTOSTOP defined).
+ *       Otherwise, the auto-stop feature is implemented in software via the interrupt.
+ *
+ * @param channel RMT channel
+ * @param en enable bit
+ * @return
+ *      - ESP_ERR_INVALID_ARG Parameter error
+ *      - ESP_OK Success
+ */
+esp_err_t rmt_enable_tx_loop_autostop(rmt_channel_t channel, bool en);
+#endif // SOC_RMT_SUPPORT_TX_LOOP_COUNT
 
 /**
 * @brief Reset RMT TX/RX memory index.
diff --git a/tools/sdk/esp32s2/include/esp-face/include/dl_define.hpp b/tools/sdk/esp32s2/include/esp-face/include/dl_define.hpp
index 16eb0881562..7809768a5bc 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/dl_define.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/dl_define.hpp
@@ -10,7 +10,7 @@
 #define DL_LOG_LAYER_LATENCY 0 /*<! - 1: print the latency of each parts of layer */
                                /*<! - 0: mute */
 
-#if CONFIG_SPIRAM_SUPPORT || CONFIG_ESP32_SPIRAM_SUPPORT || CONFIG_ESP32S3_SPIRAM_SUPPORT
+#if CONFIG_SPIRAM_SUPPORT || CONFIG_ESP32_SPIRAM_SUPPORT || CONFIG_ESP32S2_SPIRAM_SUPPORT || CONFIG_ESP32S3_SPIRAM_SUPPORT
 #define DL_SPIRAM_SUPPORT 1
 #else
 #define DL_SPIRAM_SUPPORT 0
@@ -83,8 +83,17 @@ namespace dl
 
     typedef enum
     {
-        PADDING_VALID,     /*<! no padding >*/
-        PADDING_SAME,      /*<! SAME in TensorFlow style >*/
-        PADDING_SAME_MXNET /*<! SAME in MXNET style >*/
+        PADDING_NOT_SET,
+        PADDING_VALID,      /*<! no padding >*/
+        PADDING_SAME_BEGIN,  /*<! SAME in MXNET style >*/
+        PADDING_SAME_END,   /*<! SAME in TensorFlow style >*/
     } padding_type_t;
-} // namespace dl
\ No newline at end of file
+    
+    typedef enum
+    {
+        CONSTANT,
+        EDGE, 
+        REFLECT,
+        SYMMETRIC,
+    } padding_mode_t;
+} // namespace dl
diff --git a/tools/sdk/esp32s2/include/esp-face/include/image/dl_image.hpp b/tools/sdk/esp32s2/include/esp-face/include/image/dl_image.hpp
index c7fecdc49cc..4a974df063a 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/image/dl_image.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/image/dl_image.hpp
@@ -370,11 +370,70 @@ namespace dl
          */
         uint32_t get_moving_point_number(uint8_t *f1, uint8_t *f2, const uint32_t height, const uint32_t width, const uint32_t stride, const uint32_t threshold = 5);
 
-
+        /**
+         * @brief Apply an affine transformation to an image.
+         * 
+         * @tparam T 
+         * @param input     the input image.
+         * @param output    the output image.
+         * @param M_inv     the inverse transformation matrix.
+         */
         template <typename T>
         void warp_affine(dl::Tensor<T> *input, dl::Tensor<T> *output, dl::math::Matrix<float> *M_inv);
+
+        /**
+         * @brief Apply an affine transformation to an image.
+         * 
+         * @tparam T 
+         * @param input    the pointer of the input image.
+         * @param shape    the shape of the input image.
+         * @param output   the output image.
+         * @param M_inv    the inverse transformation matrix.
+         */
         template <typename T>
         void warp_affine(uint16_t *input, std::vector<int> shape, dl::Tensor<T> *output, dl::math::Matrix<float> *M_inv);
 
+        /**
+         * @brief Get the otsu thresh object.
+         * 
+         * @param image  the gray image.
+         * @return uint8_t the otsu thresh.
+         */
+        uint8_t get_otsu_thresh(Tensor<uint8_t> &image);
+
+        /**
+         * @brief Convert RGB image to gray image
+         * 
+         * @param image  input image
+         * @param bgr    true: the image is in BGR format
+         *               false: the image is in RGB format
+         * @return Tensor<uint8_t>* output image in gray format
+         */
+        Tensor<uint8_t> *rgb2gray(Tensor<uint8_t> &image, bool bgr = false);
+
+        /**
+         * @brief Convert RGB image to LAB image
+         * 
+         * @param image  input image
+         * @param bgr    true: the image is in BGR format
+         *               false: the image is in RGB format 
+         * @param fast   true: use the fast alogrithm， but the accuracy will be reduced
+         *               false: do not use the fast alogrithm
+         * @return Tensor<uint8_t>* output image in LAB foramt
+         */
+        Tensor<uint8_t> *rgb2lab(Tensor<uint8_t> &image, bool bgr = false, bool fast = true);
+
+        /**
+         * @brief Convert RGB image to HSV image
+         * 
+         * @param image   input image
+         * @param bgr     true: the image is in BGR format
+         *                false: the image is in RGB format 
+         * @param fast    true: use the fast alogrithm， but the accuracy will be reduced
+         *                false: do not use the fast alogrithm
+         * @return Tensor<uint8_t>* output image in HSV format
+         */
+        Tensor<uint8_t> *rgb2hsv(Tensor<uint8_t> &image, bool bgr = false, bool fast = true);
+
     } // namespace image
 } // namespace dl
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_add2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_add2d.hpp
index f0c5964b3d1..c43282b42de 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_add2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_add2d.hpp
@@ -25,7 +25,8 @@ namespace dl
             const int output_exponent;               /*<! exponent of output >*/
             Tensor<feature_t> *output;               /*<! output ptr of add2d >*/
             bool inplace;                            /*<! true: the output will store to input0
-                                                          false: the output will store to a seperate memeory >*/
+                                                          false: the output will store to a separate memory >*/
+            std::vector<int> output_shape;           /*<! output shape of add2d >*/
 
         public:
             /**
@@ -35,19 +36,21 @@ namespace dl
              * @param activation      activation of add2d, if you don't specify anything, no activation is applied
              * @param name            name of add2d
              * @param inplace         true: the output will store to input0
-             *                        false: the output will store to a seperate memeory
+             *                        false: the output will store to a separate memory
              */
-            Add2D(const int output_exponent, const Activation<feature_t> *activation = NULL, const char *name = NULL, bool inplace = false) : Layer(name), activation(activation), output_exponent(output_exponent), output(NULL)
-            {
-                this->inplace = inplace;
-            }
+            Add2D(const int output_exponent, const Activation<feature_t> *activation = NULL, const char *name = "Add2D", bool inplace = false) : Layer(name),
+                                                                                                                                                 activation(activation),
+                                                                                                                                                 output_exponent(output_exponent),
+                                                                                                                                                 output(NULL),
+                                                                                                                                                 inplace(inplace),
+                                                                                                                                                 output_shape({}) {}
 
             /**
              * @brief Destroy the Add2D object
              */
             ~Add2D()
             {
-                if((!this->inplace) && (this->output != NULL))
+                if ((!this->inplace) && (this->output != NULL))
                 {
                     delete this->output;
                 }
@@ -59,10 +62,12 @@ namespace dl
              * 
              * @param input0 as one input
              * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1)
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
             {
                 assert(input0.is_same_shape(input1));
+                this->output_shape = input0.shape;
 
                 if (!this->inplace)
                 {
@@ -78,6 +83,11 @@ namespace dl
                 {
                     this->output = &input0;
                 }
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -105,7 +115,11 @@ namespace dl
                 if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output->apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
                     this->output->set_exponent(this->output_exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -116,6 +130,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::add2d(*this->output, input0, input1, this->activation, assign_core, this->output_exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "add2d");
                 }
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_avg_pool2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_avg_pool2d.hpp
index f78b262beb9..8a9aaa8dfbe 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_avg_pool2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_avg_pool2d.hpp
@@ -24,23 +24,26 @@ namespace dl
             std::vector<int> filter_shape;     /*<! filter shape in [filter_height, filter_width] >*/
             const int stride_y;                /*<! stride in height >*/
             const int stride_x;                /*<! stride in width >*/
-            const padding_type_t padding_type; /*<! one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET >*/
+            const padding_type_t padding_type; /*<! one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN >*/
             std::vector<int> padding;          /*<! padding size needed in [top, bottom, left, right] of this operation >*/
-            Tensor<feature_t> *output;          /*<! output ptr of AvgPool2D >*/
+            Tensor<feature_t> *output;         /*<! output ptr of AvgPool2D >*/
+            std::vector<int> output_shape;     /*<! output shape of AvgPool2D >*/
 
         public:
-
             /**
              * @brief Construct a new AvgPool2D object.
              * 
              * @param output_exponent exponent of output
              * @param filter_shape    filter shape in [filter_height, filter_width]
-             * @param padding_type    one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+             * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN or PADDING_NOT_SET,
              *                        - PADDING_VALID means no padding
-             *                        PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+             *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
              *                        such that output has the same height/width dimension as the input,
-             *                        - PADDING_SAME results padding in TensorFlow style
-             *                        - PADDING_SAME_MXNET results padding in MXNET style
+             *                        - PADDING_SAME_END results padding in TensorFlow style
+             *                        - PADDING_SAME_BEGIN results padding in MXNET style
+             *                        - PADDING_NOT_SET means padding with the specific "padding" value below. 
+             * @param padding         if padding_type is PADDING_NOT_SET, this value will be used as padding size. 
+             *                        the shape must be 4, the value of each position is: [padding top, padding bottom, padding left, padding right]
              * @param stride_y        stride in height
              * @param stride_x        stride in width
              * @param name            name of layer
@@ -48,16 +51,23 @@ namespace dl
             AvgPool2D(const int output_exponent,
                       const std::vector<int> filter_shape,
                       const padding_type_t padding_type = PADDING_VALID,
+                      std::vector<int> padding = {},
                       const int stride_y = 1,
                       const int stride_x = 1,
-                      const char *name = NULL) : Layer(name),
-                                                 output_exponent(output_exponent),
-                                                 filter_shape(filter_shape),
-                                                 stride_y(stride_y),
-                                                 stride_x(stride_x),
-                                                 padding_type(padding_type)
+                      const char *name = "AvgPool2D") : Layer(name),
+                                                        output_exponent(output_exponent),
+                                                        filter_shape(filter_shape),
+                                                        padding_type(padding_type),
+                                                        padding(padding),
+                                                        stride_y(stride_y),
+                                                        stride_x(stride_x),
+                                                        output_shape({})
             {
                 this->output = new Tensor<feature_t>;
+                if (this->padding_type == PADDING_NOT_SET)
+                {
+                    assert(this->padding.size() == 4);
+                }
             }
 
             /**
@@ -66,7 +76,7 @@ namespace dl
             */
             ~AvgPool2D()
             {
-                if(this->output != NULL)
+                if (this->output != NULL)
                 {
                     delete this->output;
                 }
@@ -76,20 +86,31 @@ namespace dl
              * @brief Update output shape and padding.
              * 
              * @param input as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
                 assert(input.shape[0] > 0);
                 assert(input.shape[1] > 0);
-                std::vector<int> output_shape = nn::get_output_shape(input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
-                this->output->set_shape(output_shape);
+                assert(input.shape.size() == 3);
+
+                this->output_shape = nn::get_output_shape(input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type, false, this->padding);
+                this->output->set_shape(this->output_shape);
                 this->output->set_exponent(this->output_exponent);
 
-                this->padding = nn::get_pad_size(output_shape, input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
-                input.set_padding_size(this->padding);
+                if (this->padding_type != PADDING_NOT_SET)
+                {
+                    this->padding = nn::get_pad_size(this->output_shape, input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
+                }
+
                 this->output->free_element();
-            }
 
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
 
             /**
              * @brief Get the output
@@ -108,7 +129,6 @@ namespace dl
              * @param autoload_enable one of true or false, 
              *                        - true: load input and output from PSRAM to CACHE automatically
              *                        - false: do not
-             * @param assign_core     not effective yet
              * @return AvgPool2D result
              */
             Tensor<feature_t> &call(Tensor<feature_t> &input, uint8_t autoload_enable = 0)
@@ -116,7 +136,11 @@ namespace dl
                 DL_LOG_LAYER_LATENCY_INIT();
 
                 DL_LOG_LAYER_LATENCY_START();
-                this->output->apply_element();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
                 this->output->set_exponent(this->output_exponent);
                 DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_base.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_base.hpp
index 5c7d28d52b1..b265b454538 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_base.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_base.hpp
@@ -1,6 +1,7 @@
 #pragma once
 #include "dl_tool.hpp"
 #include "dl_tool_cache.hpp"
+#include <iostream>
 
 namespace dl
 {
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_concat.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_concat.hpp
new file mode 100644
index 00000000000..35ebe652e53
--- /dev/null
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_concat.hpp
@@ -0,0 +1,139 @@
+#pragma once
+
+#include <assert.h>
+#include <vector>
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+#include "dl_nn_concat.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Concat(input1, input2, input3, ...).
+         * 
+         * @tparam feature_t support all kinds of integer and float data type
+         */
+        template <typename feature_t>
+        class Concat : Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            int axis;                      /*<! The axis along which the Tensor will be concatenated. >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Concat >*/
+            std::vector<int> output_shape; /*<! output shape of Concat >*/
+        public:
+            /**
+             * @brief Construct a new Concat object.
+             * 
+             * @param name name of layer
+             * @param axis The axis along which the Tensor will be concatenated.
+             */
+            Concat(int axis, const char *name = "Concat") : Layer(name), axis(axis), output_shape({})
+            {
+                this->output = new Tensor<feature_t>;
+            }
+
+            /**
+             * @brief Destroy the Concat object
+             */
+            ~Concat()
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Collect inputs' channel and memory offset, called in Model.build().
+             * 
+             * @param args pointers of concatenated Tensor
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(std::vector<Tensor<feature_t> *> args, bool print_shape = false)
+            {
+                assert(args.size() > 1);
+                int shape_size = args[0]->shape.size();
+
+                if (this->axis < 0)
+                {
+                    this->axis = shape_size + this->axis;
+                }
+                assert((this->axis < shape_size) && (this->axis > -1));
+
+                int output_shape_axis = args[0]->shape[this->axis];
+
+                for (int i = 1; i < args.size(); i++)
+                {
+                    assert(shape_size == args[i]->shape.size());
+                    assert(args[i]->exponent == args[i - 1]->exponent);
+                    output_shape_axis += args[i]->shape[this->axis];
+
+                    for (int j = 0; j < shape_size; j++)
+                    {
+                        if (j != this->axis)
+                        {
+                            assert(args[i]->shape[j] == args[i - 1]->shape[j]);
+                        }
+                    }
+                }
+
+                this->output_exponent = args[0]->exponent;
+                this->output_shape = args[0]->shape;
+                this->output_shape[this->axis] = output_shape_axis;
+
+                this->output->set_shape(this->output_shape);
+                this->output->set_exponent(this->output_exponent);
+                this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Call Concat operation
+             * 
+             * @param inputs                the pointers of inputs 
+             * @param free_inputs           true: free the inputs after call
+             *                              false: do not free inputs
+             * @return Tensor<feature_t>&   concat result
+             */
+            Tensor<feature_t> &call(std::vector<Tensor<feature_t> *> inputs, bool free_inputs = false)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
+                this->output->set_exponent(this->output_exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                DL_LOG_LAYER_LATENCY_START();
+                nn::concat(*this->output, inputs, this->axis, free_inputs);
+                DL_LOG_LAYER_LATENCY_END(this->name, "concat");
+                return *this->output;
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>&  Concat result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_conv2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_conv2d.hpp
index a7c2229db09..038dd6cd79f 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_conv2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_conv2d.hpp
@@ -13,8 +13,11 @@ namespace dl
          * @tparam feature_t supports int16_t and int8_t,
          *         - int16_t: stands for operation in int16_t quantize
          *         - int8_t: stands for operation in int8_t quantize
+         * @tparam bias_t supports int16_t and int8_t, must specify when using int8 per-channel quantization
+         *         - int16_t: for int16 quantization and int8 per-channel quantization
+         *         - int8_t: for int8 per-tensor quantization
          */
-        template <typename feature_t>
+        template <typename feature_t, typename bias_t = feature_t>
         class Conv2D : public Layer
         {
         private:
@@ -22,14 +25,14 @@ namespace dl
             const Filter<feature_t> *filter;         /*<! filter of Conv2D >*/
             const int stride_y;                      /*<! stride in height >*/
             const int stride_x;                      /*<! stride in width >*/
-            const padding_type_t padding_type;       /*<! one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET >*/
-            const Bias<feature_t> *bias;             /*<! bias of Conv2D, if you don't specify anything, no bias is added >*/
+            const padding_type_t padding_type;       /*<! one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN >*/
+            const Bias<bias_t> *bias;                /*<! bias of Conv2D, if you don't specify anything, no bias is added >*/
             const Activation<feature_t> *activation; /*<! activation of Conv2D, if you don't specify anything, no activation is applied >*/
             std::vector<int> padding;                /*<! padding size needed in [top, bottom, left, right] of this operation >*/
-            Tensor<feature_t> *output;              /*<! output ptr of Conv2D >*/
+            Tensor<feature_t> *output;               /*<! output ptr of Conv2D >*/
+            std::vector<int> output_shape;           /*<! output shape of Conv2D >*/
 
         public:
-
             /**
              * @brief Construct a new Conv2D object.
              * 
@@ -37,33 +40,43 @@ namespace dl
              * @param filter          filter of Conv2D
              * @param bias            bias of Conv2D, if you don't specify anything, no bias is added
              * @param activation      activation of Conv2D, if you don't specify anything, no activation is applied
-             * @param padding_type    one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+             * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN or PADDING_NOT_SET,
              *                        - PADDING_VALID means no padding
-             *                        PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+             *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
              *                        such that output has the same height/width dimension as the input,
-             *                        - PADDING_SAME results padding in TensorFlow style
-             *                        - PADDING_SAME_MXNET results padding in MXNET style
+             *                        - PADDING_SAME_END results padding in TensorFlow style
+             *                        - PADDING_SAME_BEGIN results padding in MXNET style
+             *                        - PADDING_NOT_SET means padding with the specific "padding" value below. 
+             * @param padding         if padding_type is PADDING_NOT_SET, this value will be used as padding size. 
+             *                        the shape must be 4, the value of each position is: [padding top, padding bottom, padding left, padding right]
              * @param stride_y        stride in height
              * @param stride_x        stride in width
              * @param name            name of layer
              */
             Conv2D(const int output_exponent,
                    const Filter<feature_t> *filter,
-                   const Bias<feature_t> *bias = NULL,
+                   const Bias<bias_t> *bias = NULL,
                    const Activation<feature_t> *activation = NULL,
                    const padding_type_t padding_type = PADDING_VALID,
+                   std::vector<int> padding = {},
                    const int stride_y = 1,
                    const int stride_x = 1,
-                   const char *name = NULL) : Layer(name),
-                                              output_exponent(output_exponent),
-                                              filter(filter),
-                                              stride_y(stride_y),
-                                              stride_x(stride_x),
-                                              padding_type(padding_type),
-                                              bias(bias),
-                                              activation(activation)
+                   const char *name = "Conv2D") : Layer(name),
+                                                  output_exponent(output_exponent),
+                                                  filter(filter),
+                                                  stride_y(stride_y),
+                                                  stride_x(stride_x),
+                                                  padding_type(padding_type),
+                                                  bias(bias),
+                                                  activation(activation),
+                                                  padding(padding),
+                                                  output_shape({})
             {
                 this->output = new Tensor<feature_t>;
+                if (this->padding_type == PADDING_NOT_SET)
+                {
+                    assert(this->padding.size() == 4);
+                }
             }
 
             /**
@@ -82,19 +95,30 @@ namespace dl
              * @brief Update output padding and input padding.
              * 
              * @param input as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
                 assert(input.shape[0] > 0);
                 assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
+                assert(this->filter->shape.size() == 4);
+                assert(input.shape[2] == this->filter->shape[2]);
 
-                std::vector<int> output_shape = nn::get_output_shape(input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type, true);
-                this->output->set_shape(output_shape);
+                this->output_shape = nn::get_output_shape(input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type, true, this->padding);
+                this->output->set_shape(this->output_shape);
                 this->output->set_exponent(this->output_exponent);
                 this->output->free_element();
+                if (this->padding_type != PADDING_NOT_SET)
+                {
+                    this->padding = nn::get_pad_size(this->output_shape, input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type);
+                }
 
-                this->padding = nn::get_pad_size(output_shape, input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type);
-                input.set_padding_size(this->padding);
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -122,7 +146,11 @@ namespace dl
                 DL_LOG_LAYER_LATENCY_INIT();
 
                 DL_LOG_LAYER_LATENCY_START();
-                this->output->apply_element();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
                 this->output->set_exponent(this->output_exponent);
                 DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -153,5 +181,6 @@ namespace dl
                 dl::tool::cache::preload_func((uint32_t)(this->filter->element), size);
             }
         };
+
     } // namespace layer
 } // namespace dl
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_depthwise_conv2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_depthwise_conv2d.hpp
index 37475353209..30b2c2a6c77 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_depthwise_conv2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_depthwise_conv2d.hpp
@@ -13,8 +13,11 @@ namespace dl
          * @tparam feature_t supports int16_t and int8_t,
          *         - int16_t: stands for operation in int16_t quantize
          *         - int8_t: stands for operation in int8_t quantize
+         * @tparam bias_t supports int16_t and int8_t, must specify when using int8 per-channel quantization
+         *         - int16_t: for int16 quantization and int8 per-channel quantization
+         *         - int8_t: for int8 per-tensor quantization
          */
-        template <typename feature_t>
+        template <typename feature_t, typename bias_t = feature_t>
         class DepthwiseConv2D : public Layer
         {
         private:
@@ -22,14 +25,14 @@ namespace dl
             const Filter<feature_t> *filter;         /*<! filter of DepthwiseConv2D >*/
             const int stride_y;                      /*<! stride in height >*/
             const int stride_x;                      /*<! stride in width >*/
-            const padding_type_t padding_type;       /*<! one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET >*/
-            const Bias<feature_t> *bias;             /*<! bias of DepthwiseConv2D, if you don't specify anything, no bias is added >*/
+            const padding_type_t padding_type;       /*<! one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN >*/
+            const Bias<bias_t> *bias;                /*<! bias of DepthwiseConv2D, if you don't specify anything, no bias is added >*/
             const Activation<feature_t> *activation; /*<! activation of DepthwiseConv2D, if you don't specify anything, no activation is applied >*/
             std::vector<int> padding;                /*<! padding size needed in [top, bottom, left, right] of this operation >*/
             Tensor<feature_t> *output;               /*<! output ptr of DepthwiseConv2D >*/
+            std::vector<int> output_shape;           /*<! output shape of DepthwiseConv2D >*/
 
         public:
-
             /**
              * @brief Construct a new DepthwiseConv2D object.
              * 
@@ -37,40 +40,50 @@ namespace dl
              * @param filter          filter of DepthwiseConv2D
              * @param bias            bias of DepthwiseConv2D, if you don't specify anything, no bias is added
              * @param activation      activation of DepthwiseConv2D, if you don't specify anything, no activation is applied
-             * @param padding_type    one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+             * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN or PADDING_NOT_SET,
              *                        - PADDING_VALID means no padding
-             *                        PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
-             *                        such that output has the same height/width dimension as the input
-             *                        - PADDING_SAME results padding in TensorFlow style
-             *                        - PADDING_SAME_MXNET results padding in MXNET style
+             *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
+             *                        such that output has the same height/width dimension as the input,
+             *                        - PADDING_SAME_END results padding in TensorFlow style
+             *                        - PADDING_SAME_BEGIN results padding in MXNET style
+             *                        - PADDING_NOT_SET means padding with the specific "padding" value below. 
+             * @param padding         if padding_type is PADDING_NOT_SET, this value will be used as padding size. 
+             *                        the shape must be 4, the value of each position is: [padding top, padding bottom, padding left, padding right]
              * @param stride_y        - stride in height
              * @param stride_x        - stride in width
              * @param name            name of layer
              */
             DepthwiseConv2D(const int output_exponent,
                             const Filter<feature_t> *filter,
-                            const Bias<feature_t> *bias = NULL,
+                            const Bias<bias_t> *bias = NULL,
                             const Activation<feature_t> *activation = NULL,
                             const padding_type_t padding_type = PADDING_VALID,
+                            std::vector<int> padding = {},
                             const int stride_y = 1,
                             const int stride_x = 1,
-                            const char *name = NULL) : Layer(name),
-                                                       output_exponent(output_exponent),
-                                                       filter(filter),
-                                                       stride_y(stride_y),
-                                                       stride_x(stride_x),
-                                                       padding_type(padding_type),
-                                                       bias(bias),
-                                                       activation(activation)
+                            const char *name = "DepthwiseConv2D") : Layer(name),
+                                                                    output_exponent(output_exponent),
+                                                                    filter(filter),
+                                                                    stride_y(stride_y),
+                                                                    stride_x(stride_x),
+                                                                    padding_type(padding_type),
+                                                                    bias(bias),
+                                                                    activation(activation),
+                                                                    padding(padding),
+                                                                    output_shape({})
             {
                 this->output = new Tensor<feature_t>;
+                if (this->padding_type == PADDING_NOT_SET)
+                {
+                    assert(this->padding.size() == 4);
+                }
             }
 
             /**
              * @brief Destroy the DepthwiseConv2D object.
              * 
              */
-            ~DepthwiseConv2D() 
+            ~DepthwiseConv2D()
             {
                 if (this->output != NULL)
                 {
@@ -82,19 +95,31 @@ namespace dl
              * @brief Update output shape and padding.
              * 
              * @param input as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
                 assert(input.shape[0] > 0);
                 assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
+                assert(this->filter->shape.size() == 4);
+                assert(input.shape[2] == this->filter->shape[2]);
 
-                std::vector<int> output_shape = nn::get_output_shape(input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type);
-                this->output->set_shape(output_shape);
+                this->output_shape = nn::get_output_shape(input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type, false, this->padding);
+                this->output->set_shape(this->output_shape);
                 this->output->set_exponent(this->output_exponent);
 
-                this->padding = nn::get_pad_size(output_shape, input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type);
-                input.set_padding_size(this->padding);
+                if (this->padding_type != PADDING_NOT_SET)
+                {
+                    this->padding = nn::get_pad_size(this->output_shape, input.shape, this->filter->shape_with_dilation, this->stride_y, this->stride_x, this->padding_type);
+                }
                 this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -122,7 +147,12 @@ namespace dl
                 DL_LOG_LAYER_LATENCY_INIT();
 
                 DL_LOG_LAYER_LATENCY_START();
-                this->output->apply_element();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+
+                this->output->malloc_element();
                 this->output->set_exponent(this->output_exponent);
                 DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_expand_dims.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_expand_dims.hpp
new file mode 100644
index 00000000000..a59bed183fb
--- /dev/null
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_expand_dims.hpp
@@ -0,0 +1,128 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief 
+         * 
+         * @tparam feature_t 
+         */
+        template <typename feature_t>
+        class ExpandDims : public Layer
+        {
+        private:
+            std::vector<int> output_shape; /*<! output shape of ExpandDims >*/
+            std::vector<int> axis;         /*<! position where the new axis is placed. >*/
+            Tensor<feature_t> *output;     /*<! output ptr of ExpandDims >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+
+        public:
+            int output_exponent;
+
+            /**
+             * @brief Construct a new ExpandDims object
+             * 
+             * @param axis position where the new axis is placed.
+             * @param name name of layer
+             * @param inplace true: the output will store to input
+             *                false: the output will store to a separate memory
+             */
+            ExpandDims(std::vector<int> axis, const char *name = "ExpandDims", bool inplace = false) : Layer(name),
+                                                                                                       axis(axis), inplace(inplace), output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the ExpandDims object
+             * 
+             */
+            ~ExpandDims()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief  Update output shape.
+            * 
+            * @param input as an input.
+            * @param print_shape  whether to print the output shape.
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->expand_dims(this->axis);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                    this->output->expand_dims(this->axis);
+                }
+                this->output_shape = this->output->shape;
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& ExpandDims result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief call ExpandDims opeartion
+             * 
+             * @param input 
+             * @return Tensor<feature_t>& ExpandDims result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->copy_element(input, true);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "ExpandDims");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_shape(this->output_shape);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "ExpandDims");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_flatten.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_flatten.hpp
new file mode 100644
index 00000000000..70ae483a07f
--- /dev/null
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_flatten.hpp
@@ -0,0 +1,120 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief 
+         * 
+         * @tparam feature_t 
+         */
+        template <typename feature_t>
+        class Flatten : public Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Flatten >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of Flatten >*/
+
+        public:
+            /**
+             * @brief Construct a new Flatten object
+             * 
+             * @param name name of layer
+             * @param inplace true: the output will store to input0
+             *                false: the output will store to a separate memory
+             */
+            Flatten(const char *name = "Flatten", bool inplace = false) : Layer(name), inplace(inplace), output_shape({})
+            {}
+
+            /**
+             * @brief Destroy the Flatten object
+             * 
+             */
+            ~Flatten()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief  Update output shape.
+            * 
+            * @param input as an input
+            * @param print_shape  whether to print the output shape. 
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                this->output_shape = {input.get_size()};
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                }
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Flatten result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Flatten operation.
+             * 
+             * @param input as an input
+             * @return Tensor<feature_t>& Flatten result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->flatten();
+                    this->output->copy_element(input, true);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "flatten");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->flatten();
+                    DL_LOG_LAYER_LATENCY_END(this->name, "flatten");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_fullyconnected.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_fullyconnected.hpp
new file mode 100644
index 00000000000..afa7e5befba
--- /dev/null
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_fullyconnected.hpp
@@ -0,0 +1,167 @@
+#pragma once
+
+#include "dl_nn_fully_connected.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief Activation(FullyConnected(input, filter) + bias).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @tparam bias_t supports int16_t and int8_t, must specify when using int8 per-channel quantization
+         *         - int16_t: for int16 quantization and int8 per-channel quantization
+         *         - int8_t: for int8 per-tensor quantization
+         */
+        template <typename feature_t, typename bias_t = feature_t>
+        class FullyConnected : public Layer
+        {
+        private:
+            const int output_exponent;               /*<! exponent of output >*/
+            const bool flatten;                      /*<! true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+                                                          false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim] >*/
+            const Filter<feature_t> *filter;         /*<! filter of FullyConnected >*/
+            const Bias<bias_t> *bias;                /*<! bias of FullyConnected, if you don't specify anything, no bias is added >*/
+            const Activation<feature_t> *activation; /*<! activation of FullyConnected, if you don't specify anything, no activation is applied >*/
+            Tensor<feature_t> *output;               /*<! output ptr of FullyConnected >*/
+            std::vector<int> output_shape;           /*<! output shape of FullyConnected >*/
+
+        public:
+            /**
+             * @brief Construct a new FullyConnected object.
+             * 
+             * @param output_exponent exponent of output
+             * @param filter          filter of FullyConnected
+             * @param bias            bias of FullyConnected, if you don't specify anything, no bias is added
+             * @param activation      activation of FullyConnected, if you don't specify anything, no activation is applied
+             * @param flatten         true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+                                      false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+             * @param name            name of layer
+             */
+            FullyConnected(const int output_exponent,
+                           const Filter<feature_t> *filter,
+                           const Bias<bias_t> *bias = NULL,
+                           const Activation<feature_t> *activation = NULL,
+                           const bool flatten = true,
+                           const char *name = "FullyConnected") : Layer(name),
+                                                                  output_exponent(output_exponent),
+                                                                  flatten(flatten),
+                                                                  filter(filter),
+                                                                  bias(bias),
+                                                                  activation(activation),
+                                                                  output_shape({})
+            {
+                this->output = new Tensor<feature_t>;
+            }
+
+            /**
+             * @brief Destroy the FullyConnected object.
+             * 
+             */
+            ~FullyConnected()
+            {
+                if (this->output != NULL)
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+             * @brief Update output padding and input padding.
+             * 
+             * @param input as an input
+             * @param print_shape  whether to print the output shape.
+             */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                assert(this->filter->shape.size() == 4);
+                assert(this->filter->shape[0] == 1);
+                assert(this->filter->shape[1] == 1);
+                if (this->flatten)
+                {
+                    assert(input.get_size() == this->filter->shape[2]);
+                    this->output_shape = {this->filter->shape[3]};
+                }
+                else
+                {
+                    assert(input.shape.back() == this->filter->shape[2]);
+                    this->output_shape = input.shape;
+                    this->output_shape[this->output_shape.size() - 1] = this->filter->shape[3];
+                }
+                this->output->set_shape(this->output_shape);
+                this->output->set_exponent(this->output_exponent);
+                this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& FullyConnected result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call FullyConnected operation
+             * 
+             * @param input           as an input.
+             * @param autoload_enable one of true or false, 
+             *                        - true: load input and output from PSRAM to CACHE automatically
+             *                        - false: do not
+             * @param assign_core     not effective yet
+             * @return FullyConnected result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input, bool autoload_enable = false, const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                DL_LOG_LAYER_LATENCY_START();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
+                this->output->set_exponent(this->output_exponent);
+                DL_LOG_LAYER_LATENCY_END(this->name, "apply");
+
+                if (autoload_enable)
+                {
+                    dl::tool::cache::autoload_func((uint32_t)(this->output->element), this->output->get_size() * sizeof(feature_t),
+                                                   (uint32_t)(input.element), input.get_size() * sizeof(feature_t));
+                }
+
+                DL_LOG_LAYER_LATENCY_START();
+                nn::fully_connected(*this->output, input, *(this->filter), this->bias, this->activation, this->flatten, assign_core);
+                DL_LOG_LAYER_LATENCY_END(this->name, "fully_connected");
+                return *this->output;
+            }
+
+            /**
+             * @brief Preload the filter to Cache.
+             * NOTE: Call this layer's preload() before previous layer's call() such that filter could be loaded while previous layer is doing calculation.
+             */
+            void preload()
+            {
+                size_t size = sizeof(feature_t);
+                int shape_size = this->filter->shape.size();
+                for (int i = 0; i < shape_size; ++i)
+                {
+                    size *= filter->shape[i];
+                }
+                dl::tool::cache::preload_func((uint32_t)(this->filter->element), size);
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_global_avg_pool2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_global_avg_pool2d.hpp
index 044c5f61847..93f2d30ac89 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_global_avg_pool2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_global_avg_pool2d.hpp
@@ -20,8 +20,9 @@ namespace dl
         class GlobalAveragePool2D : public Layer
         {
         private:
-            const int output_exponent; /*<! exponent of output >*/
-            Tensor<feature_t> *output; /*<! output ptr of GlobalAveragePool2D >*/
+            const int output_exponent;     /*<! exponent of output >*/
+            std::vector<int> output_shape; /*<! output shape of GlobalAveragePool2D >*/
+            Tensor<feature_t> *output;     /*<! output ptr of GlobalAveragePool2D >*/
         public:
             /**
              * @brief Construct a new GlobalAveragePool2D object.
@@ -29,8 +30,9 @@ namespace dl
              * @param output_exponent exponent of output
              * @param name            name of layer
              */
-            GlobalAveragePool2D(const int output_exponent, const char *name = NULL) : Layer(name),
-                                                                                      output_exponent(output_exponent)
+            GlobalAveragePool2D(const int output_exponent, const char *name = "GlobalAveragePool2D") : Layer(name),
+                                                                                                       output_exponent(output_exponent),
+                                                                                                       output_shape({})
 
             {
                 this->output = new Tensor<feature_t>;
@@ -52,17 +54,26 @@ namespace dl
              * @brief Update output shape.
              * 
              * @param input as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
                 assert(input.shape[0] > 0);
                 assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
 
                 std::vector<int> output_shape(input.shape.size(), 1);
                 output_shape[2] = input.shape[2];
-                this->output->set_shape(output_shape);
+                this->output_shape = output_shape;
+                this->output->set_shape(this->output_shape);
                 this->output->set_exponent(this->output_exponent);
                 this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -90,7 +101,11 @@ namespace dl
                 DL_LOG_LAYER_LATENCY_INIT();
 
                 DL_LOG_LAYER_LATENCY_START();
-                this->output->apply_element();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
                 this->output->set_exponent(this->output_exponent);
                 DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_global_max_pool2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_global_max_pool2d.hpp
index aa146823b78..f9b7f73ff8c 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_global_max_pool2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_global_max_pool2d.hpp
@@ -20,15 +20,15 @@ namespace dl
         class GlobalMaxPool2D : public Layer
         {
         private:
-            Tensor<feature_t> *output;  /*<! output ptr of GlobalMaxPool2D >*/
+            Tensor<feature_t> *output;     /*<! output ptr of GlobalMaxPool2D >*/
+            std::vector<int> output_shape; /*<! output shape of GlobalMaxPool2D >*/
         public:
-
             /**
              * @brief Construct a new GlobalMaxPool2D object.
              * 
              * @param name         name of layer
              */
-            GlobalMaxPool2D(const char *name = NULL) : Layer(name)
+            GlobalMaxPool2D(const char *name = "GlobalMaxPool2D") : Layer(name), output_shape({})
             {
                 this->output = new Tensor<feature_t>;
             }
@@ -49,17 +49,26 @@ namespace dl
              * @brief Update output shape and exponent.
              * 
              * @param input as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
                 assert(input.shape[0] > 0);
                 assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
                 this->output->set_exponent(input.exponent);
 
                 std::vector<int> output_shape(input.shape.size(), 1);
                 output_shape[2] = input.shape[2];
-                this->output->set_shape(output_shape);
+                this->output_shape = output_shape;
+                this->output->set_shape(this->output_shape);
                 this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -87,7 +96,11 @@ namespace dl
                 DL_LOG_LAYER_LATENCY_INIT();
 
                 DL_LOG_LAYER_LATENCY_START();
-                this->output->apply_element();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
                 this->output->set_exponent(input.exponent);
                 DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_leakyrelu.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_leakyrelu.hpp
index f18d9b1c522..a972e135006 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_leakyrelu.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_leakyrelu.hpp
@@ -2,7 +2,7 @@
 
 #include "dl_constant.hpp"
 #include "dl_variable.hpp"
-#include "dl_nn_LeakyReLU.hpp"
+#include "dl_nn_leakyrelu.hpp"
 #include "dl_layer_base.hpp"
 
 namespace dl
@@ -20,13 +20,13 @@ namespace dl
         class LeakyReLU : public Layer
         {
         private:
-            feature_t activation_alpha; /*<! quantized alpha >*/
-            int activation_exponent;    /*<! exponent of quantized alpha >*/
-            Tensor<feature_t> *output;  /*<! output ptr of leakyrelu>*/
-            bool inplace;               /*<! true: the output will store to input0
-                                             false: the output will store to a seperate memeory >*/
+            feature_t activation_alpha;    /*<! quantized alpha >*/
+            int activation_exponent;       /*<! exponent of quantized alpha >*/
+            Tensor<feature_t> *output;     /*<! output ptr of leakyrelu>*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of leakyrelu >*/
         public:
-
             /**
              * @brief Construct a new LeakyReLU object
              * 
@@ -34,9 +34,9 @@ namespace dl
              * @param activation_exponent  exponent of quantized alpha
              * @param name                 name of leakyrelu
              * @param inplace              true: the output will store to input0
-             *                             false: the output will store to a seperate memeory
+             *                             false: the output will store to a separate memory
              */
-            LeakyReLU(const int activation_alpha, const int activation_exponent, const char *name = NULL, bool inplace = false) : Layer(name), output(NULL)
+            LeakyReLU(const int activation_alpha, const int activation_exponent, const char *name = "LeakyReLU", bool inplace = false) : Layer(name), output(NULL), output_shape({})
             {
                 this->activation_alpha = activation_alpha;
                 this->activation_exponent = activation_exponent;
@@ -47,7 +47,7 @@ namespace dl
              * @brief Destroy the LeakyReLU object
              * 
              */
-            ~LeakyReLU() 
+            ~LeakyReLU()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -59,24 +59,32 @@ namespace dl
              * @brief Update output shape and exponent
              * 
              * @param input       as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
-                if(!this->inplace)
+                this->output_shape = input.shape;
+                if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
-                    }  
-                    this->output->set_shape(input.shape);
+                    }
+                    this->output->set_shape(this->output_shape);
                     this->output->set_exponent(input.exponent);
                     this->output->free_element();
                 }
                 else
                 {
                     this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
                 }
-                
             }
 
             /**
@@ -100,10 +108,14 @@ namespace dl
             {
                 DL_LOG_LAYER_LATENCY_INIT();
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output->apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
                     this->output->set_exponent(input.exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -114,6 +126,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::leakyrelu<true>(*this->output, input, this->activation_alpha, this->activation_exponent, assign_core);
                     DL_LOG_LAYER_LATENCY_END(this->name, "leakyrelu");
                 }
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_max2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_max2d.hpp
index 8a775a2e0f0..93bc5899443 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_max2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_max2d.hpp
@@ -22,28 +22,28 @@ namespace dl
         class Max2D : public Layer
         {
         private:
-            Tensor<feature_t> *output;  /*<! output ptr of max2d >*/
-            bool inplace;               /*<! true: the output will store to input0
-                                             false: the output will store to a seperate memeory >*/
+            Tensor<feature_t> *output;     /*<! output ptr of max2d >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of max2d >*/
         public:
-            
             /**
              * @brief Construct a new Max2D object.
              * 
              * @param name            name of max2d
              * @param inplace         true: the output will store to input0
-             *                        false: the output will store to a seperate memeory
+             *                        false: the output will store to a separate memory
              */
-            Max2D(const char *name = NULL, bool inplace = false) : Layer(name), output(NULL)
+            Max2D(const char *name = "Max2D", bool inplace = false) : Layer(name),
+                                                                      output(NULL), inplace(inplace), output_shape({})
             {
-                this->inplace = inplace;
             }
 
             /**
              * @brief Destroy the Max2D object
              * 
              */
-            ~Max2D() 
+            ~Max2D()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -58,24 +58,34 @@ namespace dl
              * 
              * @param input0 as one input
              * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1)
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
             {
                 assert(input0.is_same_shape(input1));
                 assert(input0.exponent == input1.exponent);
+                this->output_shape = input0.shape;
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
                     }
                     this->output->set_exponent(this->output_exponent);
-                    this->output->set_shape(input0.shape);
+                    this->output->set_shape(this->output_shape);
                     this->output->free_element();
                 }
                 else
+                {
                     this->output = &input0;
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -100,10 +110,14 @@ namespace dl
             {
                 DL_LOG_LAYER_LATENCY_INIT();
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output->apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
                     this->output->set_exponent(input0.exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -114,6 +128,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::max2d<true>(*this->output, input0, input1, assign_core);
                     DL_LOG_LAYER_LATENCY_END(this->name, "max2d");
                 }
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_max_pool2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_max_pool2d.hpp
index f836a983b34..629aa87f515 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_max_pool2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_max_pool2d.hpp
@@ -23,44 +23,54 @@ namespace dl
             std::vector<int> filter_shape;     /*<! filter shape in [filter_height, filter_width] >*/
             const int stride_y;                /*<! stride in height >*/
             const int stride_x;                /*<! stride in width >*/
-            const padding_type_t padding_type; /*<! one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET >*/
+            const padding_type_t padding_type; /*<! one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN >*/
             std::vector<int> padding;          /*<! padding size needed in [top, bottom, left, right] of this operation >*/
             Tensor<feature_t> *output;         /*<! output ptr of MaxPool2D >*/
+            std::vector<int> output_shape;     /*<! output shape of MaxPool2D >*/
 
         public:
-
             /**
              * @brief Construct a new MaxPool2D object.
              * 
              * @param filter_shape filter shape in [filter_height, filter_width]
-             * @param padding_type one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+             * @param padding_type one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN or PADDING_NOT_SET,
              *                     - PADDING_VALID means no padding
-             *                     PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+             *                     PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
              *                     such that output has the same height/width dimension as the input,
-             *                     - PADDING_SAME results padding in TensorFlow style
-             *                     - PADDING_SAME_MXNET results padding in MXNET style
+             *                     - PADDING_SAME_END results padding in TensorFlow style
+             *                     - PADDING_SAME_BEGIN results padding in MXNET style
+             *                     - PADDING_NOT_SET means padding with the specific "padding" value below. 
+             * @param padding      if padding_type is PADDING_NOT_SET, this value will be used as padding size. 
+             *                     the shape must be 4, the value of each position is: [padding top, padding bottom, padding left, padding right]
              * @param stride_y     stride in height
              * @param stride_x     stride in width
              * @param name         name of layer
              */
             MaxPool2D(const std::vector<int> filter_shape,
                       const padding_type_t padding_type = PADDING_VALID,
+                      std::vector<int> padding = {},
                       const int stride_y = 1,
                       const int stride_x = 1,
-                      const char *name = NULL) : Layer(name),
-                                                 filter_shape(filter_shape),
-                                                 stride_y(stride_y),
-                                                 stride_x(stride_x),
-                                                 padding_type(padding_type)
+                      const char *name = "MaxPool2D") : Layer(name),
+                                                        filter_shape(filter_shape),
+                                                        padding_type(padding_type),
+                                                        padding(padding),
+                                                        stride_y(stride_y),
+                                                        stride_x(stride_x),
+                                                        output_shape({})
             {
                 this->output = new Tensor<feature_t>;
+                if (this->padding_type == PADDING_NOT_SET)
+                {
+                    assert(this->padding.size() == 4);
+                }
             }
 
             /**
              * @brief Destroy the MaxPool2D object.
              * 
              */
-            ~MaxPool2D() 
+            ~MaxPool2D()
             {
                 if (this->output != NULL)
                 {
@@ -72,18 +82,29 @@ namespace dl
              * @brief Update output shape and padding.
              * 
              * @param input as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
                 assert(input.shape[0] > 0);
                 assert(input.shape[1] > 0);
+                assert(input.shape.size() == 3);
+
                 this->output->set_exponent(input.exponent);
-                std::vector<int> output_shape = nn::get_output_shape(input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
-                this->output->set_shape(output_shape);
+                this->output_shape = nn::get_output_shape(input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type, false, this->padding);
+                this->output->set_shape(this->output_shape);
 
-                this->padding = nn::get_pad_size(output_shape, input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
-                input.set_padding_size(this->padding);
+                if (this->padding_type != PADDING_NOT_SET)
+                {
+                    this->padding = nn::get_pad_size(this->output_shape, input.shape, filter_shape, this->stride_y, this->stride_x, this->padding_type);
+                }
                 this->output->free_element();
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -111,7 +132,11 @@ namespace dl
                 DL_LOG_LAYER_LATENCY_INIT();
 
                 DL_LOG_LAYER_LATENCY_START();
-                this->output->apply_element();
+                if (this->output->shape != this->output_shape)
+                {
+                    this->output->set_shape(this->output_shape);
+                }
+                this->output->malloc_element();
                 this->output->set_exponent(input.exponent);
                 DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_min2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_min2d.hpp
index 71d39c9b285..e38fbf3d0d2 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_min2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_min2d.hpp
@@ -22,28 +22,28 @@ namespace dl
         class Min2D : public Layer
         {
         private:
-            Tensor<feature_t> *output;  /*<! output of ptr min2d>*/ 
-            bool inplace;               /*<! true: the output will store to input0
-                                             false: the output will store to a seperate memeory >*/ 
-        public:    
-
+            Tensor<feature_t> *output;     /*<! output of ptr min2d>*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of min2d >*/
+        public:
             /**
              * @brief Construct a new Min2D object
              * 
              * @param name            name of min2d
              * @param inplace         true: the output will store to input0
-             *                        false: the output will store to a seperate memeory
+             *                        false: the output will store to a separate memory
              */
-            Min2D(const char *name = NULL, bool inplace = false) : Layer(name), output(NULL)
-            {
-                this->inplace = inplace;
-            }
+            Min2D(const char *name = "Min2D", bool inplace = false) : Layer(name),
+                                                                      output(NULL),
+                                                                      inplace(inplace),
+                                                                      output_shape({}) {}
 
             /**
              * @brief Destroy the Min2D object
              * 
              */
-            ~Min2D() 
+            ~Min2D()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -58,25 +58,34 @@ namespace dl
              * 
              * @param input0 as one input
              * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1)
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
             {
                 assert(input0.is_same_shape(input1));
                 assert(input0.exponent == input1.exponent);
+                this->output_shape = input0.shape;
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
                     }
-                    this->output->set_shape(input0.shape);
+                    this->output->set_shape(this->output_shape);
                     this->output->set_exponent(input0.exponent);
                     this->output->free_element();
                 }
                 else
+                {
                     this->output = &input0;
-                
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -101,10 +110,14 @@ namespace dl
             {
                 DL_LOG_LAYER_LATENCY_INIT();
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output->apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
                     this->output->set_exponent(input0.exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -115,6 +128,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::min2d<true>(*this->output, input0, input1, assign_core);
                     DL_LOG_LAYER_LATENCY_END(this->name, "min2d");
                 }
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_mul2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_mul2d.hpp
index 0edfc9a93f0..21bcca7a81e 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_mul2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_mul2d.hpp
@@ -21,14 +21,13 @@ namespace dl
         class Mul2D : public Layer
         {
         private:
-            const int output_exponent;              /*<! exponent of output >*/
+            const int output_exponent;               /*<! exponent of output >*/
             const Activation<feature_t> *activation; /*<! activation of Mul2D, if you don't specify anything, no activation is applied >*/
-            Tensor<feature_t> *output;  /*<! output ptr of Mul2D >*/
-            bool inplace;               /*<! true: the output will store to input0
-                                             false: the output will store to a seperate memeory >*/ 
+            Tensor<feature_t> *output;               /*<! output ptr of Mul2D >*/
+            bool inplace;                            /*<! true: the output will store to input0
+                                                            false: the output will store to a separate memory >*/
+            std::vector<int> output_shape;           /*<! output shape of Mul2D >*/
         public:
-            const int output_exponent; /*<! exponent of output >*/
-
             /**
              * @brief Construct a new Mul2D object.
              * 
@@ -36,18 +35,24 @@ namespace dl
              * @param activation      activation of Mul2D, if you don't specify anything, no activation is applied
              * @param name            name of layer
              * @param inplace         true: the output will store to input0
-             *                        false: the output will store to a seperate memeory
+             *                        false: the output will store to a separate memory
              */
-            Mul2D(const int output_exponent, const Activation<feature_t> *activation = NULL, const char *name = NULL, bool inplace = false) : Layer(name), 
-                                                                                            output_exponent(output_exponent),activation(activation), output(NULL)
+            Mul2D(const int output_exponent,
+                  const Activation<feature_t> *activation = NULL,
+                  const char *name = "Mul2D",
+                  bool inplace = false) : Layer(name),
+                                          output_exponent(output_exponent),
+                                          activation(activation),
+                                          output(NULL),
+                                          inplace(inplace),
+                                          output_shape({})
             {
-                this->inplace = inplace;
             }
 
             /**
              * @brief Destroy the Multiply2D object.
              */
-            ~Mul2D() 
+            ~Mul2D()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -61,24 +66,34 @@ namespace dl
              * 
              * @param input0 as one input
              * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1)
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
             {
                 assert(input0.is_same_shape(input1));
+                this->output_shape = input0.shape;
 
                 if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
                     }
                     this->output->set_exponent(this->output_exponent);
-                    this->output->set_shape(input0.shape);
+                    this->output->set_shape(this->output_shape);
                     this->output->free_element();
                 }
-                    
+
                 else
+                {
                     this->output = &input0;
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -106,7 +121,11 @@ namespace dl
                 if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output->apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
                     this->output->set_exponent(this->output_exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -117,6 +136,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::mul2d<true>(*this->output, input0, input1, this->activation, assign_core);
                     DL_LOG_LAYER_LATENCY_END(this->name, "mul2d");
                 }
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_prelu.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_prelu.hpp
index 4781a669130..96168a783b1 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_prelu.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_prelu.hpp
@@ -24,9 +24,9 @@ namespace dl
             int activation_exponent;       /*<! exponent of quantized alpha elements >*/
             Tensor<feature_t> *output;     /*<! output ptr of prelu >*/
             bool inplace;                  /*<! true: the output will store to input0
-                                                false: the output will store to a seperate memeory >*/ 
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of prelu >*/
         public:
-
             /**
              * @brief Construct a new PReLU object
              * 
@@ -34,20 +34,25 @@ namespace dl
              * @param activation_exponent  exponent of quantized alpha elements
              * @param name                 name of prelu
              * @param inplace              true: the output will store to input0
-             *                             false: the output will store to a seperate memeory
+             *                             false: the output will store to a separate memory
              */
-            PReLU(const feature_t *activation_element, const int activation_exponent = 0, const char *name = NULL, bool inplace = false) : Layer(name), output(NULL)
+            PReLU(const feature_t *activation_element,
+                  const int activation_exponent = 0,
+                  const char *name = NULL,
+                  bool inplace = "PReLU") : Layer(name),
+                                            activation_element(activation_element),
+                                            activation_exponent(activation_exponent),
+                                            output(NULL),
+                                            inplace(inplace),
+                                            output_shape({})
             {
-                this->activation_element = activation_element;
-                this->activation_exponent = activation_exponent;
-                this->inplace = inplace;
             }
 
             /**
              * @brief Destroy the PReLU object
              * 
              */
-            ~PReLU() 
+            ~PReLU()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -59,23 +64,31 @@ namespace dl
              * @brief Update output shape and exponent
              * 
              * @param input       as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
-                if(!this->inplace)
+                this->output_shape = input.shape;
+                if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
                     }
                     this->output->set_exponent(input.exponent);
-                    this->output->set_shape(input.shape);
+                    this->output->set_shape(this->output_shape);
                     this->output->free_element();
                 }
                 else
                 {
                     this->output = &input;
                 }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -99,11 +112,15 @@ namespace dl
             {
                 DL_LOG_LAYER_LATENCY_INIT();
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     this->output->set_exponent(input.exponent);
-                    this->output->apply_element();
+                    this->output->malloc_element();
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
                     DL_LOG_LAYER_LATENCY_START();
@@ -113,6 +130,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::prelu(*this->output, input, this->activation_element, this->activation_exponent, assign_core);
                     DL_LOG_LAYER_LATENCY_END(this->name, "leakyrelu");
                 }
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_relu.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_relu.hpp
index e70663b798c..7dd29d4a178 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_relu.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_relu.hpp
@@ -21,29 +21,28 @@ namespace dl
         class ReLU : public Layer
         {
         private:
-            Tensor<feature_t> *output;  /*<! output ptr of relu >*/
-            bool inplace;               /*<! true: the output will store to input0
-                                             false: the output will store to a seperate memeory >*/ 
+            Tensor<feature_t> *output;     /*<! output ptr of relu >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of relu >*/
         public:
-
-
             /**
              * @brief Construct a new ReLU object
              * 
              * @param name            name of relu
              * @param inplace         true: the output will store to input0
-             *                        false: the output will store to a seperate memeory
+             *                        false: the output will store to a separate memory
              */
-            ReLU(const char *name = NULL, bool inplace = false) : Layer(name), output(NULL)
+            ReLU(const char *name = "ReLU", bool inplace = false) : Layer(name),
+                                                                    output(NULL), inplace(inplace), output_shape({})
             {
-                this->inplace = inplace;
             }
 
             /**
              * @brief Destroy the ReLU object
              * 
              */
-            ~ReLU() 
+            ~ReLU()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -55,23 +54,31 @@ namespace dl
              * @brief Update output shape and exponent
              * 
              * @param input       as an input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input)
+            void build(Tensor<feature_t> &input, bool print_shape = false)
             {
-                if(!this->inplace)
+                this->output_shape = input.shape;
+                if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
                     }
                     this->output->set_exponent(input.exponent);
-                    this->output->set_shape(input.shape);
+                    this->output->set_shape(this->output_shape);
                     this->output->free_element();
                 }
                 else
                 {
                     this->output = &input;
                 }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -95,10 +102,14 @@ namespace dl
             {
                 DL_LOG_LAYER_LATENCY_INIT();
 
-                if(!this->inplace)
+                if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output->apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output->malloc_element();
                     this->output->set_exponent(input.exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -109,6 +120,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::relu(*this->output, input, assign_core);
                     DL_LOG_LAYER_LATENCY_END(this->name, "relu");
                 }
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_reshape.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_reshape.hpp
new file mode 100644
index 00000000000..3f2ed72b6e0
--- /dev/null
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_reshape.hpp
@@ -0,0 +1,124 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief  Reshape(input)
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         */
+        template <typename feature_t>
+        class Reshape : public Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Reshape >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> output_shape; /*<! output shape of Reshape >*/
+        public:
+            /**
+             * @brief Construct a new Reshape object
+             * 
+             * @param shape      the target shape
+             * @param name       name of Reshape layer
+             * @param inplace    true: the output will store to input0
+             *                   false: the output will store to a separate memory
+             */
+            Reshape(std::vector<int> shape, const char *name = "Reshape", bool inplace = false) : Layer(name),
+                                                                                                  output_shape(shape), inplace(inplace)
+            {
+            }
+
+            /**
+             * @brief Destroy the Reshape object
+             * 
+             */
+            ~Reshape()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief             Update output shape and exponent
+            * 
+            * @param input        as an input
+            * @param print_shape  whether to print the output shape.
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Reshape result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Reshape operation.
+             * 
+             * @param input  as an input
+             * @return Tensor<feature_t>& Reshape result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->reshape(this->output_shape);
+                    this->output->copy_element(input, true);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "reshape");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->reshape(this->output_shape);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "reshape");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_squeeze.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_squeeze.hpp
new file mode 100644
index 00000000000..cee92f22764
--- /dev/null
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_squeeze.hpp
@@ -0,0 +1,127 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief 
+         * 
+         * @tparam feature_t 
+         */
+        template <typename feature_t>
+        class Squeeze : public Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Squeeze >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                 false: the output will store to a separate memory >*/
+            int axis;                      /*<! the dim to to be remove. make sure the length of the dim is equal to 1.
+                                            if axis == INT32_MAX, all the dims with length==1 will be removed. >*/
+            std::vector<int> output_shape; /*<! output shape of AvgPool2D >*/
+        public:
+            /**
+             * @brief Construct a new Squeeze object
+             * 
+             * @param axis the dim to to be remove. make sure the length of the dim is equal to 1.
+             *             if axis == INT32_MAX, all the dims with length==1 will be removed.
+             * @param name      name of Squeeze layer
+             * @param inplace   true: the output will store to input0
+             *                  false: the output will store to a separate memory
+             */
+            Squeeze(int axis = INT32_MAX, const char *name = "Squeeze", bool inplace = false) : Layer(name), axis(axis), inplace(inplace), output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the Squeeze object
+             * 
+             */
+            ~Squeeze()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief Update output shape and exponent
+            * 
+            * @param input as an input
+            * @param print_shape  whether to print the output shape.
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(input.shape);
+                    this->output->squeeze(this->axis);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(input.shape);
+                    this->output->squeeze(this->axis);
+                }
+                this->output_shape = this->output->shape;
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Squeeze result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief  Call Squeeze operation.
+             * 
+             * @param input as an input
+             * @return Tensor<feature_t>& Squeeze result
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->copy_element(input, true);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "Squeeze");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_shape(this->output_shape);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "Squeeze");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_sub2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_sub2d.hpp
index e3453c08b97..da03b4aad85 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_sub2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_sub2d.hpp
@@ -21,13 +21,13 @@ namespace dl
         class Sub2D : public Layer
         {
         private:
-            const int output_exponent; /*<! exponent of output >*/
-            const Activation<feature_t> *activation; /*<! activation of Mul2D, if you don't specify anything, no activation is applied >*/
-            Tensor<feature_t> *output;  /*<! output ptr of Sub2D >*/
-            bool inplace;               /*<! true: the output will store to input0
-                                             false: the output will store to a seperate memeory >*/ 
+            const int output_exponent;               /*<! exponent of output >*/
+            const Activation<feature_t> *activation; /*<! activation of Sub2D, if you don't specify anything, no activation is applied >*/
+            Tensor<feature_t> *output;               /*<! output ptr of Sub2D >*/
+            bool inplace;                            /*<! true: the output will store to input0
+                                                            false: the output will store to a separate memory >*/
+            std::vector<int> output_shape;           /*<! output shape of Sub2D >*/
         public:
-
             /**
              * @brief Construct a new Sub2D object.
              * 
@@ -35,18 +35,17 @@ namespace dl
              * @param activation      activation of Mul2D, if you don't specify anything, no activation is applied
              * @param name            name of layer
              * @param inplace         true: the output will store to input0
-             *                        false: the output will store to a seperate memeory
+             *                        false: the output will store to a separate memory
              */
-            Sub2D(const int output_exponent, const Activation<feature_t> *activation = NULL, const char *name = NULL, bool inplace = false) : Layer(name), 
-                                                        output_exponent(output_exponent), activation(activation), output(NULL)
+            Sub2D(const int output_exponent, const Activation<feature_t> *activation = NULL, const char *name = "Sub2D", bool inplace = false) : Layer(name),
+                                                                                                                                                 output_exponent(output_exponent), activation(activation), output(NULL), inplace(inplace), output_shape({})
             {
-                this->inplace = inplace;
             }
 
             /**
              * @brief Destroy the Sub2D object.
              */
-            ~Sub2D() 
+            ~Sub2D()
             {
                 if ((!this->inplace) && (this->output != NULL))
                 {
@@ -60,22 +59,32 @@ namespace dl
              * 
              * @param input0 as one input
              * @param input1 as another input
+             * @param print_shape  whether to print the output shape.
              */
-            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1)
+            void build(Tensor<feature_t> &input0, Tensor<feature_t> &input1, bool print_shape = false)
             {
                 assert(input0.is_same_shape(input1));
+                this->output_shape = input0.shape;
                 if (!this->inplace)
                 {
-                    if(this->output != NULL)
+                    if (this->output != NULL)
                     {
                         this->output = new Tensor<feature_t>;
                     }
                     this->output->set_exponent(this->output_exponent);
-                    this->output->set_shape(input0.shape);
+                    this->output->set_shape(this->output_shape);
                     this->output->free_element();
-                }    
+                }
                 else
+                {
                     this->output = &input0;
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
             }
 
             /**
@@ -103,7 +112,11 @@ namespace dl
                 if (!this->inplace)
                 {
                     DL_LOG_LAYER_LATENCY_START();
-                    this->output.apply_element();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
+                    this->output.malloc_element();
                     this->output->set_exponent(input0.exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "apply");
 
@@ -114,6 +127,10 @@ namespace dl
                 else
                 {
                     DL_LOG_LAYER_LATENCY_START();
+                    if (this->output->shape != this->output_shape)
+                    {
+                        this->output->set_shape(this->output_shape);
+                    }
                     nn::sub2d<true>(this->output, input0, input1, this->activation, assign_core, this->output_exponent);
                     DL_LOG_LAYER_LATENCY_END(this->name, "sub2d");
                 }
diff --git a/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_transpose.hpp b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_transpose.hpp
new file mode 100644
index 00000000000..d89ba8daed5
--- /dev/null
+++ b/tools/sdk/esp32s2/include/esp-face/include/layer/dl_layer_transpose.hpp
@@ -0,0 +1,126 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_tool.hpp"
+#include "dl_layer_base.hpp"
+
+namespace dl
+{
+    namespace layer
+    {
+        /**
+         * @brief 
+         * 
+         * @tparam feature_t 
+         */
+        template <typename feature_t>
+        class Transpose : public Layer
+        {
+        private:
+            int output_exponent;           /*<! exponent of output >*/
+            Tensor<feature_t> *output;     /*<! output ptr of Transpose >*/
+            bool inplace;                  /*<! true: the output will store to input0
+                                                false: the output will store to a separate memory >*/
+            std::vector<int> perm;         /*<! the new arangement of the dims. if perm == {}, the dims arangement will be reversed. >*/
+            std::vector<int> output_shape; /*<! output shape of Transpose >*/
+        public:
+            /**
+             * @brief Construct a new Transpose object
+             * 
+             * @param perm          the new arangement of the dims. if perm == {}, the dims arangement will be reversed.
+             * @param name          name of Transpose layer
+             * @param inplace       true: the output will store to input
+             *                      false: the output will store to a separate memory
+             */
+            Transpose(std::vector<int> perm = {}, const char *name = "Transpose", bool inplace = false) : Layer(name), perm(perm), inplace(inplace), output_shape({})
+            {
+            }
+
+            /**
+             * @brief Destroy the Transpose object
+             * 
+             */
+            ~Transpose()
+            {
+                if ((!this->inplace) && (this->output != NULL))
+                {
+                    delete this->output;
+                }
+            }
+
+            /**
+            * @brief Update output shape and exponent
+            * 
+            * @param input as an input
+            * @param print_shape  whether to print the output shape.
+            */
+            void build(Tensor<feature_t> &input, bool print_shape = false)
+            {
+                this->output_exponent = input.exponent;
+                this->output_shape = input.shape;
+                for (int i = 0; i < this->perm.size(); i++)
+                {
+                    this->output_shape[i] = input.shape[this->perm[i]];
+                }
+                if (!this->inplace)
+                {
+                    if (this->output != NULL)
+                    {
+                        this->output = new Tensor<feature_t>;
+                    }
+                    this->output->set_exponent(this->output_exponent);
+                    this->output->set_shape(this->output_shape);
+                    this->output->free_element();
+                }
+                else
+                {
+                    this->output = &input;
+                    this->output->set_shape(this->output_shape);
+                }
+
+                if (print_shape)
+                {
+                    std::cout << this->name << " | ";
+                    this->output->print_shape();
+                }
+            }
+
+            /**
+             * @brief Get the output
+             * 
+             * @return Tensor<feature_t>& Transpose result
+             */
+            Tensor<feature_t> &get_output()
+            {
+                return *this->output;
+            }
+
+            /**
+             * @brief Call Transpose operation.
+             * 
+             * @param input as an input.
+             * @return Tensor<feature_t>& Transpose result.
+             */
+            Tensor<feature_t> &call(Tensor<feature_t> &input)
+            {
+                DL_LOG_LAYER_LATENCY_INIT();
+
+                if (!this->inplace)
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->set_exponent(input.exponent);
+                    this->output->transpose(input, this->perm);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "transpose");
+                }
+                else
+                {
+                    DL_LOG_LAYER_LATENCY_START();
+                    this->output->transpose(this->perm);
+                    DL_LOG_LAYER_LATENCY_END(this->name, "transpose");
+                }
+                return *this->output;
+            }
+        };
+    } // namespace layer
+} // namespace dl
diff --git a/tools/sdk/esp32s2/include/esp-face/include/model_zoo/color_detector.hpp b/tools/sdk/esp32s2/include/esp-face/include/model_zoo/color_detector.hpp
new file mode 100644
index 00000000000..063ab20b34a
--- /dev/null
+++ b/tools/sdk/esp32s2/include/esp-face/include/model_zoo/color_detector.hpp
@@ -0,0 +1,68 @@
+#pragma once
+
+#include "dl_image.hpp"
+
+typedef struct
+{
+    int area;                /*!< Area of connected domains >*/
+    std::vector<int> center; /*<! centroid of connected domains [x, y] >*/
+    std::vector<int> box;    /*<! [left_up_x, left_up_y, right_down_x, right_down_y] >*/
+} components_stats_t;
+
+class ColorDetector
+{
+private:
+    std::vector<std::vector<components_stats_t>> results; /*!< detection results >*/
+
+public:
+    std::vector<std::vector<uint8_t>> color_thresh; /*!< threshold of colors, The threshold of each color is composed of 6 numbers >*/
+    std::vector<int> area_thresh;                   /*!< the area threshold of each color, 
+                                                                the area that is smaller than the threshold is filtered >*/
+    bool bgr;                                       /*!< true: the input image is in BGR format
+                                                                false: the input image is in RGB format >*/
+
+    /**
+     * @brief get the color threshold of rectangular region in the image
+     * 
+     * @param image the input image
+     * @param box   the coordinates of the rectanglar region : [left_up_x, left_up_y, right_down_x, right_down_y]
+     * @return std::vector<uint8_t> the threshold.
+     */
+    std::vector<uint8_t> cal_color_thresh(dl::Tensor<uint8_t> &image, std::vector<int> box);
+
+    /**
+     * @brief detect the colors based on the color thresholds
+     * 
+     * @param image the input image.
+     * @return std::vector<std::vector<components_stats_t>>&  detection result.
+     */
+    std::vector<std::vector<components_stats_t>> &detect(dl::Tensor<uint8_t> &image);
+
+    /**
+     * @brief Construct a new Color Detector object
+     * 
+     * @param color_thresh  threshold of colors, The threshold of each color is composed of 6 numbers
+     * @param area_thresh   the area threshold of each color,the area that is smaller than the threshold is filtered
+     * @param bgr           true: the input image is in BGR format
+     *                      false: the input image is in RGB format
+     */
+    ColorDetector(std::vector<std::vector<uint8_t>> color_thresh, std::vector<int> area_thresh, bool bgr = false) : color_thresh(color_thresh), area_thresh(area_thresh), bgr(bgr)
+    {
+    }
+
+    /**
+     * @brief Destroy the Color Detector object
+     * 
+     */
+    ~ColorDetector() {}
+
+    /**
+     * @brief Get the results object
+     * 
+     * @return std::vector<std::vector<components_stats_t>>& the detection result.
+     */
+    std::vector<std::vector<components_stats_t>> &get_results()
+    {
+        return this->results;
+    }
+};
\ No newline at end of file
diff --git a/tools/sdk/esp32s2/include/esp-face/include/model_zoo/face_recognition_tool.hpp b/tools/sdk/esp32s2/include/esp-face/include/model_zoo/face_recognition_tool.hpp
index 12fe8a842a2..2226d32daf9 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/model_zoo/face_recognition_tool.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/model_zoo/face_recognition_tool.hpp
@@ -92,7 +92,7 @@ namespace face_recognition_tool
      * @return dl::Tensor<T>* 
      */
     template <typename T>
-    dl::Tensor<T> *transform_mfn_input(dl::Tensor<uint8_t> &image, bool free_input = false, bool do_padding = true);
+    dl::Tensor<T> *transform_mfn_input(dl::Tensor<uint8_t> &image, bool free_input = false);
 
     /**
      * @brief  transform the image to the input of a mfn model 
@@ -106,7 +106,7 @@ namespace face_recognition_tool
      *                             false: do not pad the result
      */
     template <typename T>
-    void transform_mfn_input(dl::Tensor<uint8_t> &image, dl::Tensor<T> &output, bool free_input = false, bool do_padding = true);
+    void transform_mfn_input(dl::Tensor<uint8_t> &image, dl::Tensor<T> &output, bool free_input = false);
 
     /**
      * @brief transform the mfn output embedding to a floating embedding
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn.hpp
index 6dba8016f26..6c737c1d341 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn.hpp
@@ -14,13 +14,13 @@ namespace dl
          * @param filter_shape filter shape with dilation
          * @param stride_y     stride in height
          * @param stride_x     stride in width
-         * @param pad_type     one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET
+         * @param pad_type     one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN
          * @param is_conv2d    one of true or false,
          *                     - true: serve for Conv2D
          *                     - false: serve for other operations
          * @return std::vector<int> 
          */
-        std::vector<int> get_output_shape(const std::vector<int> &input_shape, const std::vector<int> &filter_shape, const int stride_y, const int stride_x, const padding_type_t pad_type, const bool is_conv2d = false);
+        std::vector<int> get_output_shape(const std::vector<int> &input_shape, const std::vector<int> &filter_shape, const int stride_y, const int stride_x, const padding_type_t pad_type, const bool is_conv2d = false, std::vector<int> padding = {});
 
         /**
          * @brief Get the pad size object
@@ -30,7 +30,7 @@ namespace dl
          * @param filter_shape filter shape with dilation
          * @param stride_y     stride in height
          * @param stride_x     stride in width
-         * @param padding_type one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET
+         * @param padding_type one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN
          * @return padding size
          */
         std::vector<int> get_pad_size(const std::vector<int> &output_shape, const std::vector<int> &input_shape, const std::vector<int> &filter_shape, const int stride_y, const int stride_x, const padding_type_t padding_type);
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_add2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_add2d.hpp
index d296be5350b..4d4daaaef05 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_add2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_add2d.hpp
@@ -58,20 +58,20 @@ namespace dl
          */
         template <bool inplace = false, typename feature_t>
         auto add2d(const int output_exponent,
-                    Tensor<feature_t> &input0,
-                    Tensor<feature_t> &input1,
-                    const Activation<feature_t> *activation,
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+                   Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const Activation<feature_t> *activation,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             assert(input0.is_same_shape(input1));
 
             DL_LOG_NN_LATENCY_INIT();
 
             Tensor<feature_t> output;
-            if constexpr(!inplace)
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(output_exponent).set_shape(input0.shape).apply_element();
+                output.set_exponent(output_exponent).set_shape(input0.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_avg_pool2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_avg_pool2d.hpp
index 5b298d98c44..6e7db6ed0e0 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_avg_pool2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_avg_pool2d.hpp
@@ -58,12 +58,12 @@ namespace dl
          * @param filter_shape    filter_shape in [filter_height, filter_width]
          * @param stride_y        stride in height
          * @param stride_x        stride in width
-         * @param padding_type    one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+         * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN,
          *                        - PADDING_VALID: no padding
-         *                        PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+         *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
          *                        such that output has the same height/width dimension as the input,
-         *                        - PADDING_SAME results padding in TensorFlow style
-         *                        - PADDING_SAME_MXNET results padding in MXNET style
+         *                        - PADDING_SAME_END results padding in TensorFlow style
+         *                        - PADDING_SAME_BEGIN results padding in MXNET style
          * @param assign_core     not effective yet
          * @return avg_pool2d result
          */
@@ -81,19 +81,19 @@ namespace dl
             DL_LOG_NN_LATENCY_START();
             std::vector<int> output_shape = get_output_shape(input.shape, filter_shape, stride_y, stride_x, padding_type);
             Tensor<feature_t> output;
-            output.set_exponent(output_exponent).set_shape(output_shape).apply_element();
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
             DL_LOG_NN_LATENCY_END("apply");
+            std::vector<int> padding(4, 0);
 
             DL_LOG_NN_LATENCY_START();
-            if (padding_type == PADDING_SAME || padding_type == PADDING_SAME_MXNET)
+            if (padding_type == PADDING_SAME_END || padding_type == PADDING_SAME_BEGIN)
             {
-                std::vector<int> padding = get_pad_size(output_shape, input.shape, filter_shape, stride_y, stride_x, padding_type);
-                input.set_padding_size(padding);
+                padding = get_pad_size(output_shape, input.shape, filter_shape, stride_y, stride_x, padding_type);
             }
             DL_LOG_NN_LATENCY_END("padding");
 
             DL_LOG_NN_LATENCY_START();
-            avg_pool2d(output, input, input.padding, filter_shape, stride_y, stride_x, assign_core);
+            avg_pool2d(output, input, padding, filter_shape, stride_y, stride_x, assign_core);
             DL_LOG_NN_LATENCY_END("avg_pool2d");
 
             return output;
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_concat.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_concat.hpp
new file mode 100644
index 00000000000..73ed1aae905
--- /dev/null
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_concat.hpp
@@ -0,0 +1,63 @@
+#pragma once
+
+#include <vector>
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        template <typename feature_t>
+        void concat(Tensor<feature_t> &output, std::vector<Tensor<feature_t> *> &inputs, int axis, bool free_inputs = false);
+
+        template <typename feature_t>
+        Tensor<feature_t> concat(std::vector<Tensor<feature_t> *> &inputs, int axis, bool free_inputs = false)
+        {
+            DL_LOG_NN_LATENCY_INIT();
+
+            DL_LOG_NN_LATENCY_START();
+            assert(inputs.size() > 1);
+            int shape_size = inputs[0]->shape.size();
+
+            if (axis < 0)
+            {
+                axis = shape_size + axis;
+            }
+
+            assert((axis < shape_size) && (axis > -1));
+
+            int output_shape_axis = inputs[0]->shape[axis];
+
+            for (int i = 1; i < inputs.size(); i++)
+            {
+                assert(shape_size == inputs[i]->shape.size());
+                assert(inputs[i]->exponent == inputs[i - 1]->exponent);
+                output_shape_axis += inputs[i]->shape[axis];
+
+                for (int j = 0; j < shape_size; j++)
+                {
+                    if (j != axis)
+                    {
+                        assert(inputs[i]->shape[j] == inputs[i - 1]->shape[j]);
+                    }
+                }
+            }
+            DL_LOG_NN_LATENCY_END("assert");
+
+            DL_LOG_NN_LATENCY_START();
+            Tensor<feature_t> output;
+            std::vector<int> output_shape = inputs[0]->shape;
+            output_shape[axis] = output_shape_axis;
+            output.set_shape(output_shape);
+            output.set_exponent(inputs[0]->exponent);
+            output.malloc_element();
+            DL_LOG_NN_LATENCY_END("malloc");
+
+            DL_LOG_NN_LATENCY_START();
+            concat(output, inputs, axis, free_inputs);
+            DL_LOG_NN_LATENCY_END("concat");
+            return output;
+        }
+    } // namespace nn
+} // namespace dl
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_conv2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_conv2d.hpp
index 77e2c617d04..27ba0372730 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_conv2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_conv2d.hpp
@@ -10,7 +10,6 @@ namespace dl
     {
         /**
          * @brief activation(conv2d(input, filter) + bias).
-         * NOTE: When padding_type is SAME, make sure padding is already added in input.
          * 
          * @param output      as an output
          * @param input       as an input
@@ -34,7 +33,6 @@ namespace dl
 
         /**
          * @brief activation(conv2d(input, filter) + bias).
-         * NOTE: When padding_type is SAME, make sure padding is already added in input.
          * 
          * @param output      as an output
          * @param input       as an input
@@ -56,6 +54,29 @@ namespace dl
                     const Activation<int8_t> *const activation = NULL,
                     const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
 
+        /**
+         * @brief activation(conv2d(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param padding     padding size needed in [top, bottom, left, right] of this operation
+         * @param filter      filter of conv2d
+         * @param stride_y    stride in height
+         * @param stride_x    stride in width
+         * @param bias        bias of conv2d, if you don't specify anything, no bias is added
+         * @param activation  activation of conv2d, if you don't specify anything, no activation is applied
+         * @param assign_core not effective yet
+         */
+        void conv2d(Tensor<int8_t> &output,
+                    Tensor<int8_t> &input,
+                    std::vector<int> &padding,
+                    const Filter<int8_t> &filter,
+                    const int stride_y,
+                    const int stride_x,
+                    const Bias<int16_t> *const bias = NULL,
+                    const Activation<int8_t> *const activation = NULL,
+                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
         /**
          * @brief activation(conv2d(input, filter) + bias).
          * 
@@ -67,25 +88,25 @@ namespace dl
          * @param filter          Filter of conv2d
          * @param stride_y        stride in height
          * @param stride_x        stride in width
-         * @param padding_type    one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+         * @param padding_type    one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN,
          *                        - PADDING_VALID: no padding
-         *                        PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+         *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
          *                        such that output has the same height/width dimension as the input,
-         *                        - PADDING_SAME results padding in TensorFlow style
-         *                        - PADDING_SAME_MXNET results padding in MXNET style
+         *                        - PADDING_SAME_END results padding in TensorFlow style
+         *                        - PADDING_SAME_BEGIN results padding in MXNET style
          * @param bias            bias of conv2d, if you don't specify anything, no bias is added
          * @param activation      activation of conv2d, if you don't specify anything, no activation is applied
          * @param assign_core     not effective yet
          * @return conv2d result
          */
-        template <typename feature_t>
+        template <typename feature_t, typename bias_t>
         Tensor<feature_t> conv2d(const int output_exponent,
                                  Tensor<feature_t> &input,
                                  const Filter<feature_t> &filter,
                                  const int stride_y,
                                  const int stride_x,
                                  const padding_type_t padding_type,
-                                 const Bias<feature_t> *bias,
+                                 const Bias<bias_t> *bias,
                                  const Activation<feature_t> *activation,
                                  const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
         {
@@ -94,20 +115,19 @@ namespace dl
             DL_LOG_NN_LATENCY_START();
             std::vector<int> output_shape = get_output_shape(input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type, true);
             Tensor<feature_t> output;
-            output.set_exponent(output_exponent).set_shape(output_shape).apply_element();
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
             DL_LOG_NN_LATENCY_END("apply");
 
+            std::vector<int> padding(4, 0);
             DL_LOG_NN_LATENCY_START();
-            if (padding_type == PADDING_SAME || padding_type == PADDING_SAME_MXNET)
+            if (padding_type == PADDING_SAME_END || padding_type == PADDING_SAME_BEGIN)
             {
-                std::vector<int> padding = get_pad_size(output_shape, input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
-                input.set_padding_size(padding);
-                input.set_padding_value(padding, 0);
+                padding = get_pad_size(output_shape, input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
             }
             DL_LOG_NN_LATENCY_END("padding");
 
             DL_LOG_NN_LATENCY_START();
-            conv2d(output, input, input.padding, filter, stride_y, stride_x, bias, activation, assign_core);
+            conv2d(output, input, padding, filter, stride_y, stride_x, bias, activation, assign_core);
             DL_LOG_NN_LATENCY_END("conv2d");
 
             return output;
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_depthwise_conv2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_depthwise_conv2d.hpp
index 6e972bfd81b..135815a65cb 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_depthwise_conv2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_depthwise_conv2d.hpp
@@ -10,7 +10,6 @@ namespace dl
     {
         /**
          * @brief activate(depthwise_conv2d(input, filter) + bias)
-         * NOTE: When padding_type is SAME, make sure padding is already added in input
          * 
          * @param output      as an output
          * @param input       as an input
@@ -34,7 +33,6 @@ namespace dl
 
         /**
          * @brief activate(depthwise_conv2d(input, filter) + bias)
-         * NOTE: When padding_type is SAME, make sure padding is already added in input
          * 
          * @param output      as an output
          * @param input       as an input
@@ -56,6 +54,29 @@ namespace dl
                               const Activation<int8_t> *activation = NULL,
                               const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
 
+        /**
+         * @brief activate(depthwise_conv2d(input, filter) + bias)
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param padding     padding size needed in [top, bottom, left, right] of this operation
+         * @param filter      Filter of depthwise_conv2d
+         * @param stride_y    stride in height
+         * @param stride_x    stride in width
+         * @param bias        bias of depthwise_conv2d, if you don't specify anything, no bias is added
+         * @param activation  activation of depthwise_conv2d, if you don't specify anything, no activation is applied
+         * @param assign_core not effective yet
+         */
+        void depthwise_conv2d(Tensor<int8_t> &output,
+                              Tensor<int8_t> &input,
+                              std::vector<int> &padding,
+                              const Filter<int8_t> &filter,
+                              const int stride_y,
+                              const int stride_x,
+                              const Bias<int16_t> *bias = NULL,
+                              const Activation<int8_t> *activation = NULL,
+                              const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
         /**
          * @brief activation(depthwise_conv2d(input, filter) + bias)
          * 
@@ -67,25 +88,25 @@ namespace dl
          * @param filter          filter of depthwise_conv2d
          * @param stride_y        stride in height
          * @param stride_x        stride in width
-         * @param pad_type        one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+         * @param pad_type        one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN,
          *                        - PADDING_VALID means no padding
-         *                        PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+         *                        PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
          *                        such that output has the same height/width dimension as the input,
-         *                        - PADDING_SAME results padding in TensorFlow style
-         *                        - PADDING_SAME_MXNET results padding in MXNET style
+         *                        - PADDING_SAME_END results padding in TensorFlow style
+         *                        - PADDING_SAME_BEGIN results padding in MXNET style
          * @param bias            bias of depthwise_conv2d, if you don't specify anything, no bias is added
          * @param activation      activation of depthwise_conv2d, if you don't specify anything, no activation is applied
          * @param assign_core     not effective yet
          * @return depthwise_conv2d result
          */
-        template <typename feature_t>
+        template <typename feature_t, typename bias_t>
         Tensor<feature_t> depthwise_conv2d(const int output_exponent,
                                            Tensor<feature_t> &input,
                                            const Filter<feature_t> &filter,
                                            const int stride_y,
                                            const int stride_x,
                                            const padding_type_t padding_type,
-                                           const Bias<feature_t> *bias,
+                                           const Bias<bias_t> *bias,
                                            const Activation<feature_t> *activation,
                                            const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
         {
@@ -94,20 +115,20 @@ namespace dl
             DL_LOG_NN_LATENCY_START();
             std::vector<int> output_shape = get_output_shape(input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
             Tensor<feature_t> output;
-            output.set_exponent(output_exponent).set_shape(output_shape).apply_element();
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
             DL_LOG_NN_LATENCY_END("apply");
 
+            std::vector<int> padding(4, 0);
+
             DL_LOG_NN_LATENCY_START();
-            if (padding_type == PADDING_SAME || padding_type == PADDING_SAME_MXNET)
+            if (padding_type == PADDING_SAME_END || padding_type == PADDING_SAME_BEGIN)
             {
-                std::vector<int> padding = get_pad_size(output_shape, input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
-                input.set_padding_size(padding);
-                input.set_padding_value(padding, 0);
+                padding = get_pad_size(output_shape, input.shape, filter.shape_with_dilation, stride_y, stride_x, padding_type);
             }
             DL_LOG_NN_LATENCY_END("padding");
 
             DL_LOG_NN_LATENCY_START();
-            depthwise_conv2d(output, input, input.padding, filter, stride_y, stride_x, bias, activation, assign_core);
+            depthwise_conv2d(output, input, padding, filter, stride_y, stride_x, bias, activation, assign_core);
             DL_LOG_NN_LATENCY_END("depthwise_conv2d");
 
             return output;
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_fully_connected.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_fully_connected.hpp
new file mode 100644
index 00000000000..372c84825fd
--- /dev/null
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_fully_connected.hpp
@@ -0,0 +1,126 @@
+#pragma once
+
+#include "dl_constant.hpp"
+#include "dl_variable.hpp"
+#include "dl_nn.hpp"
+
+namespace dl
+{
+    namespace nn
+    {
+        /**
+         * @brief activation(FullyConnected(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param filter      filter of FullyConnected
+         * @param bias        bias of FullyConnected, if you don't specify anything, no bias is added
+         * @param activation  activation of FullyConnected, if you don't specify anything, no activation is applied
+         * @param flatten     true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+         *                    false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+         * @param assign_core not effective yet
+         */
+        void fully_connected(Tensor<int16_t> &output,
+                             Tensor<int16_t> &input,
+                             const Filter<int16_t> &filter,
+                             const Bias<int16_t> *const bias = NULL,
+                             const Activation<int16_t> *const activation = NULL,
+                             const bool flatten = true,
+                             const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(FullyConnected(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param filter      filter of FullyConnected
+         * @param bias        bias of FullyConnected, if you don't specify anything, no bias is added
+         * @param activation  activation of FullyConnected, if you don't specify anything, no activation is applied
+         * @param flatten     true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+         *                    false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+         * @param assign_core not effective yet
+         */
+        void fully_connected(Tensor<int8_t> &output,
+                             Tensor<int8_t> &input,
+                             const Filter<int8_t> &filter,
+                             const Bias<int8_t> *const bias = NULL,
+                             const Activation<int8_t> *const activation = NULL,
+                             const bool flatten = true,
+                             const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(FullyConnected(input, filter) + bias).
+         * 
+         * @param output      as an output
+         * @param input       as an input
+         * @param filter      filter of FullyConnected
+         * @param bias        bias of FullyConnected, if you don't specify anything, no bias is added
+         * @param activation  activation of FullyConnected, if you don't specify anything, no activation is applied
+         * @param flatten     true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+         *                    false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+         * @param assign_core not effective yet
+         */
+        void fully_connected(Tensor<int8_t> &output,
+                             Tensor<int8_t> &input,
+                             const Filter<int8_t> &filter,
+                             const Bias<int16_t> *const bias = NULL,
+                             const Activation<int8_t> *const activation = NULL,
+                             const bool flatten = true,
+                             const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+
+        /**
+         * @brief activation(FullyConnected(input, filter) + bias).
+         * 
+         * @tparam feature_t supports int16_t and int8_t,
+         *         - int16_t: stands for operation in int16_t quantize
+         *         - int8_t: stands for operation in int8_t quantize
+         * @param output_exponent exponent of output
+         * @param input           as an input
+         * @param filter          Filter of FullyConnected
+         * @param bias            bias of FullyConnected, if you don't specify anything, no bias is added
+         * @param activation      activation of FullyConnected, if you don't specify anything, no activation is applied
+         * @param flatten         true: input shape is [x1, x2, ..., xn], filter shape is [1, 1, x1 * x2 * ... * xn, output_dim], output shape is [output_dim]
+         *                        false: input shape is [x1, x2, ..., xn, input_dim], filter shape is [1, 1, input_dim, output_dim], output shape is [x1, x2, ...., xn, output_dim]
+         * @param assign_core     not effective yet
+         * @return FullyConnected result
+         */
+        template <typename feature_t>
+        Tensor<feature_t> fully_connected(const int output_exponent,
+                                          Tensor<feature_t> &input,
+                                          const Filter<feature_t> &filter,
+                                          const Bias<feature_t> *bias,
+                                          const Activation<feature_t> *activation,
+                                          const bool flatten,
+                                          const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE)
+        {
+            DL_LOG_NN_LATENCY_INIT();
+
+            DL_LOG_NN_LATENCY_START();
+            assert(filter.shape.size() == 4);
+            assert(filter.shape[0] == 1);
+            assert(filter.shape[1] == 1);
+
+            std::vector<int> output_shape;
+            if (flatten)
+            {
+                assert(input.get_size() == filter.shape[2]);
+                output_shape = {filter.shape.back()};
+            }
+            else
+            {
+                assert(input.shape.back() == filter->shape[2]);
+                output_shape = input.shape;
+                output_shape[output_shape.size() - 1] = filter.shape.back();
+            }
+            Tensor<feature_t> output;
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
+            DL_LOG_NN_LATENCY_END("apply");
+
+            DL_LOG_NN_LATENCY_START();
+            fully_connected(output, input, filter, bias, activation, flatten, assign_core);
+            DL_LOG_NN_LATENCY_END("fully_connected");
+
+            return output;
+        }
+    } // namespace nn
+} // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_global_avg_pool2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_global_avg_pool2d.hpp
index 723ca9eddc4..724d3ca0952 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_global_avg_pool2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_global_avg_pool2d.hpp
@@ -53,7 +53,7 @@ namespace dl
             std::vector<int> output_shape(input.shape.size(), 1);
             output_shape[2] = input.shape[2];
             Tensor<feature_t> output;
-            output.set_exponent(output_exponent).set_shape(output_shape).apply_element();
+            output.set_exponent(output_exponent).set_shape(output_shape).malloc_element();
             DL_LOG_NN_LATENCY_END("apply");
 
             DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_global_max_pool2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_global_max_pool2d.hpp
index 945645cfd55..f6f15e91bd3 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_global_max_pool2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_global_max_pool2d.hpp
@@ -51,7 +51,7 @@ namespace dl
             std::vector<int> output_shape(input.shape.size(), 1);
             output_shape[2] = input.shape[2];
             Tensor<feature_t> output;
-            output.set_exponent(input.exponent).set_shape(output_shape).apply_element();
+            output.set_exponent(input.exponent).set_shape(output_shape).malloc_element();
             DL_LOG_NN_LATENCY_END("apply");
 
             DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_leakyrelu.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_leakyrelu.hpp
index d3738fc44ca..c41728b1ad5 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_leakyrelu.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_leakyrelu.hpp
@@ -52,17 +52,17 @@ namespace dl
          * @return leakyrelu result or no return(result store to input)
          */
         template <bool inplace = false, typename feature_t>
-        auto leakyrelu(Tensor<feature_t> &input, 
-                        const int activation_alpha, 
-                        const int activation_exponent, 
-                        const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        auto leakyrelu(Tensor<feature_t> &input,
+                       const int activation_alpha,
+                       const int activation_exponent,
+                       const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
-            if constexpr(!inplace)
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(input.exponent).set_shape(input.shape).apply_element();
+                output.set_exponent(input.exponent).set_shape(input.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_max2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_max2d.hpp
index 0a5e8f43221..466089bc386 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_max2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_max2d.hpp
@@ -48,20 +48,20 @@ namespace dl
          * @return max2d result or no return(result store to input0)
          */
         template <bool inplace = false, typename feature_t>
-        auto max2d(Tensor<feature_t> &input0, 
-                    Tensor<feature_t> &input1, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        auto max2d(Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             assert(input0.is_same_shape(input1));
             assert(input0.exponent == input1.exponent);
 
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
-            
-            if constexpr(!inplace)
+
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(input0.exponent).set_shape(input0.shape).apply_element();
+                output.set_exponent(input0.exponent).set_shape(input0.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_max_pool2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_max_pool2d.hpp
index 7f95f3d4278..50d51728ce0 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_max_pool2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_max_pool2d.hpp
@@ -57,12 +57,12 @@ namespace dl
          * @param filter_shape filter shape in [filter_height, filter_width]
          * @param stride_y     stride in height
          * @param stride_x     stride in width
-         * @param padding_type one of PADDING_VALID or PADDING_SAME or PADDING_SAME_MXNET,
+         * @param padding_type one of PADDING_VALID or PADDING_SAME_END or PADDING_SAME_BEGIN,
          *                     - PADDING_VALID: no padding
-         *                     PADDING_SAME and PADDING_SAME_MXNET results in padding with zeros evenly to the left/right or up/down of the input 
+         *                     PADDING_SAME_END and PADDING_SAME_BEGIN results in padding with zeros evenly to the left/right or up/down of the input 
          *                     such that output has the same height/width dimension as the input,
-         *                     - PADDING_SAME results padding in TensorFlow style
-         *                     - PADDING_SAME_MXNET results padding in MXNET style
+         *                     - PADDING_SAME_END results padding in TensorFlow style
+         *                     - PADDING_SAME_BEGIN results padding in MXNET style
          * @param assign_core  not effective yet
          * @return max_pool2d result
          */
@@ -79,20 +79,20 @@ namespace dl
             DL_LOG_NN_LATENCY_START();
             std::vector<int> output_shape = get_output_shape(input.shape, filter_shape, stride_y, stride_x, padding_type);
             Tensor<feature_t> output;
-            output.set_exponent(input.exponent).set_shape(output_shape).apply_element();
+            output.set_exponent(input.exponent).set_shape(output_shape).malloc_element();
             DL_LOG_NN_LATENCY_END("apply");
 
+            std::vector<int> padding(4, 0);
+
             DL_LOG_NN_LATENCY_START();
-            if (padding_type == PADDING_SAME || padding_type == PADDING_SAME_MXNET)
+            if (padding_type == PADDING_SAME_END || padding_type == PADDING_SAME_BEGIN)
             {
-                std::vector<int> padding = get_pad_size(output_shape, input.shape, filter_shape, stride_y, stride_x, padding_type);
-                input.set_padding_size(padding);
-                input.set_padding_value(padding, 0);
+                padding = get_pad_size(output_shape, input.shape, filter_shape, stride_y, stride_x, padding_type);
             }
             DL_LOG_NN_LATENCY_END("padding");
 
             DL_LOG_NN_LATENCY_START();
-            max_pool2d(output, input, input.padding, filter_shape, stride_y, stride_x, assign_core);
+            max_pool2d(output, input, padding, filter_shape, stride_y, stride_x, assign_core);
             DL_LOG_NN_LATENCY_END("max_pool2d");
 
             return output;
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_min2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_min2d.hpp
index 71cb87d50d5..8faddf3c228 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_min2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_min2d.hpp
@@ -47,20 +47,20 @@ namespace dl
          * @return min2d result or no return(result store to input0)
          */
         template <bool inplace = false, typename feature_t>
-        auto min2d(Tensor<feature_t> &input0, 
-                    Tensor<feature_t> &input1, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        auto min2d(Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             assert(input0.is_same_shape(input1));
             assert(input0.exponent == input1.exponent);
 
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
-            
-            if constexpr(!inplace)
+
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(input0.exponent).set_shape(input0.shape).apply_element();
+                output.set_exponent(input0.exponent).set_shape(input0.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_mul2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_mul2d.hpp
index 410528a05a3..909619a8767 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_mul2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_mul2d.hpp
@@ -18,12 +18,12 @@ namespace dl
          * @param assign_core       not effective yet
          * @param output_exponent   exponent of output, only and must specify if inplace operation happens
          */
-        void mul2d(Tensor<int16_t> &output, 
-                    Tensor<int16_t> &input0, 
-                    Tensor<int16_t> &input1, 
-                    const Activation<int16_t> *const activation = NULL, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
-                    const int output_exponent = INT_MIN);
+        void mul2d(Tensor<int16_t> &output,
+                   Tensor<int16_t> &input0,
+                   Tensor<int16_t> &input1,
+                   const Activation<int16_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
 
         /**
          * @brief activation(mul2d(input0, input1)).
@@ -35,12 +35,12 @@ namespace dl
          * @param assign_core       not effective yet
          * @param output_exponent   exponent of output, only and must specify if inplace operation happens
          */
-        void mul2d(Tensor<int8_t> &output, 
-                    Tensor<int8_t> &input0, 
-                    Tensor<int8_t> &input1, 
-                    const Activation<int8_t> *const activation = NULL, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
-                    const int output_exponent = INT_MIN);
+        void mul2d(Tensor<int8_t> &output,
+                   Tensor<int8_t> &input0,
+                   Tensor<int8_t> &input1,
+                   const Activation<int8_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
 
         /**
          * @brief activation(mul2d(input0, input1)).
@@ -57,21 +57,21 @@ namespace dl
          * @return mul2d result or no return(result store to input0)
          */
         template <bool inplace = false, typename feature_t>
-        auto mul2d(const int output_exponent, 
-                    Tensor<feature_t> &input0, 
-                    Tensor<feature_t> &input1, 
-                    const Activation<feature_t> *activation, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        auto mul2d(const int output_exponent,
+                   Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const Activation<feature_t> *activation,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             assert(input0.is_same_shape(input1));
 
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
 
-            if constexpr(!inplace)
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(output_exponent).set_shape(input0.shape).apply_element();
+                output.set_exponent(output_exponent).set_shape(input0.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_prelu.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_prelu.hpp
index fb4315d9fc3..e83e8975604 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_prelu.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_prelu.hpp
@@ -52,17 +52,17 @@ namespace dl
          * @return prelu result or no return(result store to input)
          */
         template <bool inplace = false, typename feature_t>
-        auto prelu(Tensor<feature_t> &input, 
-                    const feature_t *activation_element, 
-                    const int activation_exponent, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        auto prelu(Tensor<feature_t> &input,
+                   const feature_t *activation_element,
+                   const int activation_exponent,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
-            if constexpr(!inplace)
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(input.exponent).set_shape(input.shape).apply_element();
+                output.set_exponent(input.exponent).set_shape(input.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
@@ -76,7 +76,7 @@ namespace dl
                 DL_LOG_NN_LATENCY_START();
                 prelu(input, input, activation_element, activation_exponent, assign_core);
                 DL_LOG_NN_LATENCY_END("prelu");
-            } 
+            }
         }
     } // namespace nn
 } // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_relu.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_relu.hpp
index e4159fdf898..308492dfe4b 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_relu.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_relu.hpp
@@ -15,9 +15,9 @@ namespace dl
          * @param input       as an input
          * @param assign_core not effective yet
          */
-        void relu(Tensor<int16_t> &output, 
-                    Tensor<int16_t> &input, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+        void relu(Tensor<int16_t> &output,
+                  Tensor<int16_t> &input,
+                  const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
 
         /**
          * @brief relu(input).
@@ -26,9 +26,9 @@ namespace dl
          * @param input       as an input
          * @param assign_core not effective yet
          */
-        void relu(Tensor<int8_t> &output, 
-                    Tensor<int8_t> &input, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
+        void relu(Tensor<int8_t> &output,
+                  Tensor<int8_t> &input,
+                  const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE);
 
         /**
          * @brief relu(input)
@@ -46,11 +46,11 @@ namespace dl
         {
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
-            
-            if constexpr(!inplace)
+
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(input.exponent).set_shape(input.shape).apply_element();
+                output.set_exponent(input.exponent).set_shape(input.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_sub2d.hpp b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_sub2d.hpp
index 385188e4ba2..5bbd4940b10 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_sub2d.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/nn/dl_nn_sub2d.hpp
@@ -18,12 +18,12 @@ namespace dl
          * @param assign_core not effective yet
          * @param output_exponent   exponent of output, only and must specify if inplace operation happens 
          */
-        void sub2d(Tensor<int16_t> &output, 
-                    Tensor<int16_t> &input0, 
-                    Tensor<int16_t> &input1, 
-                    const Activation<int16_t> *const activation = NULL, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
-                    const int output_exponent = INT_MIN);
+        void sub2d(Tensor<int16_t> &output,
+                   Tensor<int16_t> &input0,
+                   Tensor<int16_t> &input1,
+                   const Activation<int16_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
 
         /**
          * @brief activation(sub2d(input0, input1)).
@@ -35,12 +35,12 @@ namespace dl
          * @param assign_core not effective yet
          * @param output_exponent   exponent of output, only and must specify if inplace operation happens
          */
-        void sub2d(Tensor<int8_t> &output, 
-                    Tensor<int8_t> &input0, 
-                    Tensor<int8_t> &input1, 
-                    const Activation<int8_t> *const activation = NULL, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
-                    const int output_exponent = INT_MIN);
+        void sub2d(Tensor<int8_t> &output,
+                   Tensor<int8_t> &input0,
+                   Tensor<int8_t> &input1,
+                   const Activation<int8_t> *const activation = NULL,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE,
+                   const int output_exponent = INT_MIN);
 
         /**
          * @brief activation(sub2d(input0, input1)).
@@ -57,20 +57,20 @@ namespace dl
          * @return sub2d result or no return(result store to input0)
          */
         template <bool inplace = false, typename feature_t>
-        auto sub2d(const int output_exponent, 
-                    Tensor<feature_t> &input0, 
-                    Tensor<feature_t> &input1, 
-                    const Activation<feature_t> *activation, 
-                    const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
+        auto sub2d(const int output_exponent,
+                   Tensor<feature_t> &input0,
+                   Tensor<feature_t> &input1,
+                   const Activation<feature_t> *activation,
+                   const std::vector<int> &assign_core = CONFIG_DEFAULT_ASSIGN_CORE) -> typename std::conditional<inplace, void, Tensor<feature_t>>::type
         {
             assert(input0.is_same_shape(input1));
 
             DL_LOG_NN_LATENCY_INIT();
             Tensor<feature_t> output;
-            if constexpr(!inplace)
+            if constexpr (!inplace)
             {
                 DL_LOG_NN_LATENCY_START();
-                output.set_exponent(output_exponent).set_shape(input0.shape).apply_element();
+                output.set_exponent(output_exponent).set_shape(input0.shape).malloc_element();
                 DL_LOG_NN_LATENCY_END("apply");
 
                 DL_LOG_NN_LATENCY_START();
diff --git a/tools/sdk/esp32s2/include/esp-face/include/tool/dl_tool.hpp b/tools/sdk/esp32s2/include/esp-face/include/tool/dl_tool.hpp
index f8e0871a04f..e5490e073d1 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/tool/dl_tool.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/tool/dl_tool.hpp
@@ -67,62 +67,49 @@ namespace dl
         void copy_memory(void *dst, void *src, const int n);
 
         /**
-         * @brief Apply memory without initialized. Must use free_aligned() to free the memory.
+         * @brief Apply memory without initialized. Can use free_aligned() to free the memory.
          * 
          * @param number number of elements
          * @param size   size of element
-         * @param align  number of aligned, e.g., 16 means 16-byte aligned
+         * @param align  number of byte aligned, e.g., 16 means 16-byte aligned
          * @return pointer of allocated memory. NULL for failed
          */
-        inline void *malloc_aligned(int number, int size, int align = 0)
+        inline void *malloc_aligned(int number, int size, int align = 4)
         {
-            int n = number * size;
-            n >>= 4;
-            n += 2;
-            n <<= 4;
-            int total_size = n + align + sizeof(void *) + sizeof(int);
-            void *res = malloc(total_size);
+            assert((align > 0) && (((align & (align-1)) == 0)));
+            int total_size = number * size;
+
+            void *res = heap_caps_aligned_alloc(align, total_size, MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL);
 #if DL_SPIRAM_SUPPORT
             if (NULL == res)
-                res = heap_caps_malloc(total_size, MALLOC_CAP_SPIRAM);
+                res = heap_caps_aligned_alloc(align, total_size, MALLOC_CAP_SPIRAM);
 #endif
             if (NULL == res)
             {
                 printf("Fail to malloc %d bytes from DRAM(%d bytyes) and PSRAM(%d bytes), PSRAM is %s.\n",
                        total_size,
-                       heap_caps_get_free_size(MALLOC_CAP_INTERNAL),
+                       heap_caps_get_free_size(MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL),
                        heap_caps_get_free_size(MALLOC_CAP_SPIRAM),
                        DL_SPIRAM_SUPPORT ? "on" : "off");
                 return NULL;
             }
-            void **data = (void **)res + 2; // 4-byte for pointer, 4-bytes for n
-            void **aligned;
-            if (align)
-                aligned = (void **)(((size_t)data + (align - 1)) & -align);
-            else
-                aligned = data;
-
-            aligned[-1] = res;
-            int *temp = (int *)aligned;
-            temp[-2] = n;
 
-            return (void *)aligned;
+            return (void *)res;
         }
 
         /**
-         * @brief Apply memory with zero-initialized. Must use dl_lib_free() to free the memory.
+         * @brief Apply memory with zero-initialized. Can use free_aligned() to free the memory.
          * 
          * @param number number of elements
          * @param size   size of element
-         * @param align  number of aligned, e.g., 16 means 16-byte aligned
+         * @param align  number of byte aligned, e.g., 16 means 16-byte aligned
          * @return pointer of allocated memory. NULL for failed
          */
-        inline void *calloc_aligned(int number, int size, int align = 0)
+        inline void *calloc_aligned(int number, int size, int align = 4)
         {
 
             void *aligned = malloc_aligned(number, size, align);
-            int n = *((int *)aligned - 2);
-            set_zero(aligned, n);
+            set_zero(aligned, number * size);
 
             return (void *)aligned;
         }
@@ -137,7 +124,70 @@ namespace dl
             if (NULL == address)
                 return;
 
-            free(((void **)address)[-1]);
+            heap_caps_free(address);
+        }
+
+        /**
+         * @brief Apply memory without initialized in preference order: internal aligned, internal, external aligned
+         * 
+         * @param number number of elements
+         * @param size   size of element
+         * @param align  number of byte aligned, e.g., 16 means 16-byte aligned
+         * @return pointer of allocated memory. NULL for failed
+         */
+        inline void *malloc_aligned_prefer(int number, int size, int align = 4)
+        {
+            assert((align > 0) && (((align & (align-1)) == 0)));
+            int total_size = number * size;
+            void *res = heap_caps_aligned_alloc(align, total_size, MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL);
+            if (NULL == res){
+                res = heap_caps_malloc(total_size, MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL);
+            }
+#if DL_SPIRAM_SUPPORT
+            if (NULL == res){
+                res = heap_caps_aligned_alloc(align, total_size, MALLOC_CAP_SPIRAM);
+            }
+#endif
+            if (NULL == res)
+            {
+                printf("Fail to malloc %d bytes from DRAM(%d bytyes) and PSRAM(%d bytes), PSRAM is %s.\n",
+                       total_size,
+                       heap_caps_get_free_size(MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL),
+                       heap_caps_get_free_size(MALLOC_CAP_SPIRAM),
+                       DL_SPIRAM_SUPPORT ? "on" : "off");
+                return NULL;
+            }
+
+            return res;
+        }
+
+        /**
+         * @brief Apply memory with zero-initialized in preference order: internal aligned, internal, external aligned
+         * 
+         * @param number number of elements
+         * @param size   size of element
+         * @param align  number of byte aligned, e.g., 16 means 16-byte aligned
+         * @return pointer of allocated memory. NULL for failed
+         */
+        inline void *calloc_aligned_prefer(int number, int size, int align = 4)
+        {
+            void *res = malloc_aligned_prefer(number, size, align);
+            set_zero(res, number * size);
+
+            return (void *)res;
+        }
+
+        /**
+         * @brief Free the calloc_aligned_prefer() and malloc_aligned_prefer() memory
+         * 
+         * @param address pointer of memory to free
+         */
+        inline void free_aligned_prefer(void *address)
+        {
+            if (NULL == address)
+                return;
+
+            heap_caps_free(address);
         }
 
         /**
diff --git a/tools/sdk/esp32s2/include/esp-face/include/typedef/dl_constant.hpp b/tools/sdk/esp32s2/include/esp-face/include/typedef/dl_constant.hpp
index 73e3b0832e7..07b2dd24ee1 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/typedef/dl_constant.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/typedef/dl_constant.hpp
@@ -57,7 +57,8 @@ namespace dl
          * @param exponent exponent of element
          * @param shape    shape of Filter,
          *                 - 1D: reserved
-         *                 - 2D: [filter_height, filter_width, input_channel, output_channel]
+         *                 - 2D: for convolution is [filter_height, filter_width, input_channel, output_channel],
+         *                       for depthwise convolution is [filter_height, filter_width, input_channel, 1]
          * @param dilation dilation of Filter
          *                 - 1D: reserved
          *                 - 2D: [dilation_in_height, dilation_in_width]
@@ -97,6 +98,9 @@ namespace dl
     {
     public:
         using Constant<T>::Constant;
+        std::vector<int> channel_exponent;    /*<! exponent for per-channel >*/
+
+        Bias(const T *element, const std::vector<int> channel_exponent, const std::vector<int> shape);
     };
 
     /**
diff --git a/tools/sdk/esp32s2/include/esp-face/include/typedef/dl_variable.hpp b/tools/sdk/esp32s2/include/esp-face/include/typedef/dl_variable.hpp
index bf6e8630856..18cbe9707e9 100644
--- a/tools/sdk/esp32s2/include/esp-face/include/typedef/dl_variable.hpp
+++ b/tools/sdk/esp32s2/include/esp-face/include/typedef/dl_variable.hpp
@@ -3,6 +3,7 @@
 #include <stdio.h>
 #include <vector>
 #include <assert.h>
+#include <iostream>
 
 #include "dl_tool.hpp"
 
@@ -17,27 +18,20 @@ namespace dl
     class Tensor
     {
     private:
-        int size;       /*<! size of element including padding */
-        bool auto_free; /*<! free element when object destroy */
+        int size;                     /*<! size of element including padding */
+        bool auto_free;               /*<! free element when object destroy */
+        std::vector<int> axis_offset; /*<! element offset of each axis */
 
     public:
-        T *element;                          /*<! point to element */
-        int exponent;                        /*<! exponent of element */
-        std::vector<int> shape;              /*<! shape of Tensor */
-                                             /*<! 2D: shape is [height, width, channel] */
-                                             /*<! 1D: reserved */
-        std::vector<int> shape_with_padding; /*<! shape with padding of Tensor */
-                                             /*<! 2D: shape_with_padding is [height_with_padding, width_with_padding, channel_with_padding] */
-                                             /*<! 1D: reserved */
-        std::vector<int> padding;            /*<! padding of Tensor */
-                                             /*<!- 2D: padding format is [top, bottom, left, right] */
-                                             /*<! - 1D: reserved */
+        T *element;             /*<! point to element */
+        int exponent;           /*<! exponent of element */
+        std::vector<int> shape; /*<! shape of Tensor */
 
         /**
          * @brief Construct a new Tensor object
          * 
          */
-        Tensor() : size(-1), auto_free(true), element(NULL), exponent(0) {}
+        Tensor() : auto_free(true), element(NULL), exponent(0) { this->set_shape({0}); }
 
         /**
          * @brief Construct a new Tensor object by copying from input.
@@ -49,21 +43,20 @@ namespace dl
          */
         Tensor(Tensor<T> &input, bool deep) : size(input.size),
                                               auto_free(input.auto_free),
-                                              exponent(input.exponent),
-                                              shape(input.shape),
-                                              shape_with_padding(input.shape_with_padding),
-                                              padding(input.padding)
+                                              exponent(input.exponent)
         {
-            if (deep)
+            this->set_shape(input.shape);
+            if (deep && (input.element != NULL))
             {
-                int size_real = input.shape_with_padding.size() ? input.shape_with_padding[0] * input.shape_with_padding[1] * input.shape_with_padding[2] : 0;
-                T *new_element = (T *)tool::calloc_aligned(size_real, sizeof(T), 16);
+                int size_real = input.get_size();
+                T *new_element = (T *)tool::calloc_aligned_prefer(size_real, sizeof(T), 16);
                 tool::copy_memory(new_element, input.element, size_real * sizeof(T));
                 this->element = new_element;
             }
             else
             {
                 this->element = input.element;
+                this->auto_free = false;
             }
         }
 
@@ -77,6 +70,33 @@ namespace dl
                 this->free_element();
         }
 
+        /**
+         * @brief 
+         * 
+         * @param input an input Tensor
+         * @param deep one of true or false
+         *              - true: apply a new memory, copy value from input.element to this new memory
+         *              - false: take over input.element to this->element
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &copy_element(Tensor<T> &input, bool deep)
+        {
+            assert(this->get_size() == input.get_size());
+            assert(input.element != NULL);
+
+            this->malloc_element();
+            if (deep)
+            {
+                tool::copy_memory(this->element, input.element, this->get_size() * sizeof(T));
+            }
+            else
+            {
+                this->element = input.element;
+                this->auto_free = false;
+            }
+            return *this;
+        }
+
         /**
          * @brief Set the auto free object.
          * 
@@ -120,190 +140,144 @@ namespace dl
         }
 
         /**
-         * @brief Set the shape of Tensor. Initial this->padding = {0}. Initial this->size = -1.
+         * @brief Set the shape of Tensor.
          * 
-         * @param shape shape in 
-         *              - 2D: [height, width]
+         * @param shape the target shape 
+         *        
          * @return self
          */
-        Tensor<T> &set_shape(const std::vector<int> shape)
-        {
-            for (int i = 0; i < shape.size(); ++i)
-            {
-                assert(shape[i] > 0);
-            }
-            this->shape = shape;
-            this->shape_with_padding = shape;
-            this->size = -1;
-            this->padding = std::vector<int>(((this->shape.size() - 1) << 1), 0);
-            return *this;
-        }
+        Tensor<T> &set_shape(const std::vector<int> shape);
 
         /**
-         * @brief Set the padding size object.
+         * @brief print the shape of the Tensor
          * 
-         * @param padding padding size in
-         *                - 2D: [top, bottom, left, right]
-         * @return self
          */
-        Tensor &set_padding_size(std::vector<int> &padding)
+        void print_shape()
         {
-            assert(this->shape.size());      // call Tensor.set_shape() first
-            assert(this->shape.size() == 3); // TODO: || this->shape.size() == 2
-
-            if (this->shape.size() == 3)
+            if (this->shape.size())
             {
-                std::vector<int> new_padding = this->padding;
-                bool dont_update = true;
-
-                if (padding[0] > this->padding[0])
+                printf("shape = (");
+                for (int i = 0; i < this->shape.size() - 1; i++)
                 {
-                    new_padding[0] = padding[0];
-                    dont_update = false;
-                }
-
-                if (padding[1] > this->padding[1])
-                {
-                    new_padding[1] = padding[1];
-                    dont_update = false;
-                }
-
-                if (padding[2] > this->padding[2])
-                {
-                    new_padding[2] = padding[2];
-                    dont_update = false;
-                }
-
-                if (padding[3] > this->padding[3])
-                {
-                    new_padding[3] = padding[3];
-                    dont_update = false;
+                    printf("%d, ", this->shape[i]);
                 }
+                printf("%d)\n", this->shape.back());
+            }
+            else
+            {
+                printf("shape = ()\n");
+            }
+        }
 
-                if (dont_update)
-                {
-                    return *this;
-                }
+        /**
+         * @brief flatten the Tensor
+         * 
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &flatten();
 
-                std::vector<int> new_shape_with_padding = this->shape;
+        /**
+         * @brief Change a new shape to the Tensor without changing its data.
+         * 
+         * @param shape  the target shape
+         * @return Tensor<T>&  self
+         */
+        Tensor<T> &reshape(std::vector<int> shape);
 
-                new_shape_with_padding[0] += (new_padding[0] + new_padding[1]);
-                new_shape_with_padding[1] += (new_padding[2] + new_padding[3]);
-                int new_size = new_shape_with_padding[0] * new_shape_with_padding[1] * new_shape_with_padding[2];
+        /**
+         * @brief Remove dims with length==1 from Tensor
+         * 
+         * @param axis the dim to to be remove. make sure the length of the dim is equal to 1.
+         *              if axis == INT32_MAX, all the dims with length==1 will be removed.
+         * @return Tensor<T>& self 
+         */
+        Tensor<T> &squeeze(int axis = INT32_MAX);
 
-                if (this->element) // if this->element != NULL, do padding by copy memory
-                {
-                    T *new_element = (T *)tool::malloc_aligned(new_size, sizeof(T), 16);
-                    T *dst = new_element + ((new_padding[0] * new_shape_with_padding[1]) + new_padding[2]) * new_shape_with_padding[2];
-                    T *src = this->get_element_ptr();
-                    int offset_dst_next_y = new_shape_with_padding[1] * new_shape_with_padding[2];     // width * channel
-                    int src_copy_length = this->shape[1] * this->shape[2];                             // width * channel
-                    int offset_src_next_y = this->shape_with_padding[1] * this->shape_with_padding[2]; // width * channel
-                    for (int y = 0; y < this->shape[0]; y++)
-                    {
-                        tool::copy_memory(dst, src, src_copy_length * sizeof(T));
-                        dst += offset_dst_next_y;
-                        src += offset_src_next_y;
-                    }
+        /**
+         * @brief Insert a new dim that will appear at the axis position in the expanded Tensor shape.
+         * 
+         * @param axis the dim to be inserted
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &expand_dims(int axis);
 
-                    if (this->auto_free)
-                        tool::free_aligned(this->element);
-                    this->element = new_element;
-                    this->auto_free = true;
-                }
-                this->padding = new_padding;
-                this->shape_with_padding = new_shape_with_padding;
-                this->size = new_size;
-            }
-            else if (this->shape.size() == 2)
-            {
-                printf("Tensor.set_padding_size with this->shape.size() == 2 not implement yet.\n");
-            }
+        /**
+         * @brief Insert a new dim that will appear at the axis position in the expanded Tensor shape.
+         * 
+         * @param axis  the dim to be inserted
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &expand_dims(std::vector<int> axis);
 
-            return *this;
-        }
+        /**
+         * @brief Reverse or permute the axes of the Tensor
+         * 
+         * @param perm the new arangement of the dims. if perm == {}, the dims arangement will be reversed. 
+         * @return Tensor<T>& self
+         */
+        Tensor<T> &transpose(std::vector<int> perm = {});
 
         /**
-         * @brief Set the padding value object.
+         * @brief Reverse or permute the axes of the input Tensor
          * 
-         * @param padding padding size in
-         *                - 2D: [top, bottom, left, right]
-         * @param value   value to set
-         * @return self
+         * @param input the input Tensor
+         * @param perm the new arangement of the dims. if perm == {}, the dims arangement will be reversed. 
+         * @return Tensor<T>& self
          */
-        Tensor<T> &set_padding_value(std::vector<int> &padding, T value);
+        Tensor<T> &transpose(Tensor<T> &input, std::vector<int> perm = {});
 
         /**
          * @brief Get the element pointer.
          * 
-         * @param padding padding size in
-         *                - 2D: [top, bottom, left, right]
-         * @return pointer to memory with padding
+         * @return pointer to memory
          */
-        T *get_element_ptr(const std::vector<int> padding = {0, 0, 0, 0})
+        T *get_element_ptr()
         {
-            assert(this->shape.size() == 3); // TODO: || this->shape.size() == 2
-
-            if (this->shape.size() == 3)
-            {
-                return this->element + ((this->padding[0] - padding[0]) * this->shape_with_padding[1] + (this->padding[2] - padding[2])) * this->shape_with_padding[2];
-            }
-            else if (this->shape.size() == 2)
-            {
-                printf("Tensor.get_element_ptr with this->shape.size() == 2 is not implemented.\n");
-            }
-
-            return NULL;
+            return this->element;
         }
 
         /**
          * @brief Get the element value.
          * 
-         * @param index        index in
-         *                     - 2D: [y, x, c]
-         * @param with_padding one of true or false,
-         *                     - true: make padding size in count
-         *                     - false: do not
-         * @return element value
+         * @param index   the index of each dim. 
+         * @return T element value
          */
-        T &get_element_value(const std::vector<int> index, const bool with_padding = false)
+        T get_element_value(const std::vector<int> index)
         {
-            assert(index.size() == this->shape.size());
-            assert(this->shape.size() == 3); // TODO: || this->shape() == 2
-
-            int i = 0;
-            if (this->shape.size() == 3)
-            {
-                int y = index[0];
-                int x = index[1];
-                int c = index[2];
-                i = with_padding ? (y * this->shape_with_padding[1] + x) * this->shape_with_padding[2] + c : ((y + this->padding[0]) * this->shape_with_padding[1] + x + this->padding[2]) * this->shape_with_padding[2] + c;
-            }
-            else if (this->shape.size() == 2)
-            {
-                printf("Tensor.get_element_value with this->shape.size() == 2 is not implemented.\n");
-            }
+            return this->element[this->get_element_index(index)];
+        }
 
-            return this->element[i];
+        /**
+         * @brief Get the element value.
+         * 
+         * @param index  the index of the element.
+         * @return T  element value
+         */
+        T get_element_value(int index)
+        {
+            return this->element[index];
         }
 
         /**
-         * @brief Get the size of element.
+         * @brief Get the size of Tensor.
          * 
-         * @return size of element including padding
+         * @return  the size of Tensor.
          */
         int get_size()
         {
-            if (this->size == -1) // didn't call Tensor.set_padding_size() before
-            {
-                this->size = 1;
-                for (std::vector<int>::iterator d = this->shape.begin(); d != this->shape.end(); d++)
-                    this->size *= *d;
-            }
-
             return this->size;
         }
 
+        /**
+         * @brief Get the axis offset
+         * 
+         * @return std::vector<int> the axis offset
+         */
+        std::vector<int> get_axis_offset()
+        {
+            return this->axis_offset;
+        }
+
         /**
          * @brief Apply memory with zero-initialized only if this->element is NULL.
          * 
@@ -319,7 +293,7 @@ namespace dl
             if (this->element != NULL)
                 return false;
 
-            this->element = (T *)dl::tool::calloc_aligned(this->get_size(), sizeof(T), 16);
+            this->element = (T *)dl::tool::calloc_aligned_prefer(this->get_size(), sizeof(T), 16);
             this->auto_free = auto_free;
 
             return true;
@@ -340,31 +314,7 @@ namespace dl
             if (this->element != NULL)
                 return false;
 
-            this->element = (T *)tool::malloc_aligned(this->get_size(), sizeof(T), 16);
-            this->auto_free = auto_free;
-
-            return true;
-        }
-
-        /**
-         * @brief If this->element != NULL no memory will be applied and no value will be set in padding.
-         * Else apply memory without initialized and set value to padding.
-         * 
-         * @param padding_value value to set in padding
-         * @param auto_free     one of true of false
-         *                      - true: free element when object destroyed
-         *                      - false: do not
-         * @return 
-         *         - true: apply memory and set padding value successfully
-         *         - false: no memory applied and no padding value set
-         */
-        bool apply_element(const T padding_value = 0, const bool auto_free = true)
-        {
-            if (this->element != NULL)
-                return false;
-
-            this->element = (T *)tool::malloc_aligned(this->get_size(), sizeof(T), 16);
-            this->set_padding_value(this->padding, padding_value);
+            this->element = (T *)tool::malloc_aligned_prefer(this->get_size(), sizeof(T), 16);
             this->auto_free = auto_free;
 
             return true;
@@ -379,258 +329,56 @@ namespace dl
         {
             if (this->auto_free && this->element)
             {
-                tool::free_aligned(this->element);
+                tool::free_aligned_prefer(this->element);
                 this->element = NULL;
             }
         }
 
         /**
-         * @brief Print the shape of Tensor in format "shape = ({top_padding} + {height} + {bottom_padding}, {left_padding} + {width} + {right_padding}, {channel}(channel_with_padding))\n".
+         * @brief print the element of the tensor
+         * 
+         * @param axis_index_range  the element range of each dims to be print. if axis_index_range == {}, all the element will be print. 
+         * @param message  to print  
          */
-        void print_shape()
-        {
-            printf("shape = (%d + %d + %d, %d + %d + %d, %d(%d))\n",
-                   this->padding[0], this->shape[0], this->padding[1],
-                   this->padding[2], this->shape[1], this->padding[3],
-                   this->shape[2], this->shape_with_padding[2]);
-        }
+        void print(std::vector<int> axis_index_range = {}, const char *message = "");
 
         /**
-         * @brief Take numpy for example, this function print Tensor[y_start:y_end, x_start:x_end, c_start:c_end].
+         * @brief  print all the element of the Tensor.
          * 
-         * inner box is effective value of Tensor, "0" around is padding.
-         * 
-         * (with padding)
-         *               00000000000000000000000000000000000000000000000000
-         *               00000000000000000000000000000000000000000000000000
-         *               00000000000000000000000000000000000000000000000000
-         *               000000(without padding)                   00000000
-         *               000000                                    00000000
-         *               000000                                    00000000
-         *               000000          effective value           00000000
-         *               000000                                    00000000
-         *               000000                                    00000000
-         *               00000000000000000000000000000000000000000000000000
-         *               00000000000000000000000000000000000000000000000000
-         *               00000000000000000000000000000000000000000000000000
-         * 
-         * @param y_start start index in height
-         * @param y_end   end index in height
-         * @param x_start start index in width
-         * @param x_end   end index in width
-         * @param c_start start index in channel
-         * @param c_end   end index in channel
-         * @param message to print
-         * @param axis    print aligned this axis, effective only if all y_end - y_start, x_end - x_start and c_end - c_start equals to 1
+         * @param message to print  
          * @param with_padding one of true or false,
-         *                     - true: count from (with padding) in upper image
-         *                     - false: count from (without padding) in upper image
+         *                     - true: the padding element will also be ed
+         *                     - false: the padding element will not be ed
          */
-        void print(int y_start, int y_end,
-                   int x_start, int x_end,
-                   int c_start, int c_end,
-                   const char *message, int axis = 0, const bool with_padding = false)
+        void print_all(const char *message = "")
         {
-            assert(y_end > y_start);
-            assert(x_end > x_start);
-            assert(c_end > c_start);
-
-            y_start = DL_MAX(y_start, 0);
-            x_start = DL_MAX(x_start, 0);
-            c_start = DL_MAX(c_start, 0);
-            if (with_padding)
-            {
-                y_end = DL_MIN(y_end, this->shape_with_padding[0]);
-                x_end = DL_MIN(x_end, this->shape_with_padding[1]);
-                c_end = DL_MIN(c_end, this->shape_with_padding[2]);
-            }
-            else
-            {
-                y_end = DL_MIN(y_end, this->shape[0]);
-                x_end = DL_MIN(x_end, this->shape[1]);
-                c_end = DL_MIN(c_end, this->shape[2]);
-            }
-
-            printf("%s[%d:%d, %d:%d, %d:%d] | ", message, y_start, y_end, x_start, x_end, c_start, c_end);
+            std::cout << "\n"
+                      << message << " | ";
             this->print_shape();
 
-            if (y_end - y_start == 1)
-            {
-                if (x_end - x_start == 1)
-                {
-                    for (int c = c_start; c < c_end; c++)
-                        printf("%7d", c);
-                    printf("\n");
-
-                    for (int c = c_start; c < c_end; c++)
-                        printf("%7d", this->get_element_value({y_start, x_start, c}, with_padding));
-                    printf("\n");
-
-                    return;
-                }
-                else
-                {
-                    if (c_end - c_start == 1)
-                    {
-                        for (int x = x_start; x < x_end; x++)
-                            printf("%7d", x);
-                        printf("\n");
-
-                        for (int x = x_start; x < x_end; x++)
-                            printf("%7d", this->get_element_value({y_start, x, c_start}, with_padding));
-                        printf("\n");
-
-                        return;
-                    }
-                }
-            }
-            else
+            for (int i = 0; i < this->get_size(); i++)
             {
-                if (x_end - x_start == 1)
-                {
-                    if (c_end - c_start == 1)
-                    {
-                        for (int y = y_start; y < y_end; y++)
-                            printf("%7d", y);
-                        printf("\n");
-
-                        for (int y = y_start; y < y_end; y++)
-                            printf("%7d", this->get_element_value({y, x_start, c_start}, with_padding));
-                        printf("\n");
-
-                        return;
-                    }
-                }
+                std::cout << this->element[i] << " ";
             }
-
-            if (y_end - y_start == 1)
-                axis = 0;
-
-            if (x_end - x_start == 1)
-                axis = 1;
-
-            if (c_end - c_start == 1)
-                axis = 2;
-
-            if (axis == 0)
-            {
-                // ______c
-                // |
-                // |
-                // x
-                //
-                for (int y = y_start; y < y_end; y++)
-                {
-                    printf("y = %d\n     ", y);
-
-                    for (int c = c_start; c < c_end; c++)
-                        printf("%7d", c);
-                    printf("\n");
-
-                    for (int x = x_start; x < x_end; x++)
-                    {
-                        printf("%5d", x);
-                        for (int c = c_start; c < c_end; c++)
-                            printf("%7d", this->get_element_value({y, x, c}, with_padding));
-                        printf("\n");
-                    }
-                    printf("\n");
-                }
-            }
-            else if (axis == 1)
-            {
-                // ______c
-                // |
-                // |
-                // y
-                //
-                for (int x = x_start; x < x_end; x++)
-                {
-                    printf("x = %d\n     ", x);
-
-                    for (int c = c_start; c < c_end; c++)
-                        printf("%7d", c);
-                    printf("\n");
-
-                    for (int y = y_start; y < y_end; y++)
-                    {
-                        printf("%5d", y);
-                        for (int c = c_start; c < c_end; c++)
-                            printf("%7d", this->get_element_value({y, x, c}, with_padding));
-                        printf("\n");
-                    }
-                    printf("\n");
-                }
-            }
-            else
-            {
-                // ______x
-                // |
-                // |
-                // y
-                //
-                for (int c = c_start; c < c_end; c++)
-                {
-                    printf("c = %d\n     ", c);
-
-                    for (int x = x_start; x < x_end; x++)
-                        printf("%7d", x);
-                    printf("\n");
-
-                    for (int y = y_start; y < y_end; y++)
-                    {
-                        printf("%5d", y);
-                        for (int x = x_start; x < x_end; x++)
-                            printf("%7d", this->get_element_value({y, x, c}, with_padding));
-                        printf("\n");
-                    }
-                    printf("\n");
-                }
-            }
-
+            std::cout << "\n";
             return;
         }
 
         /**
-         * @brief print all the element of the Tensor.
+         * @brief Get the index of each dims
          * 
-         * @param message to print
-         * @param with_padding one of true or false,
-         *                     - true: the padding element will also be printed
-         *                     - false: the padding element will not be printed
+         * @param element_index the index of the element
+         * @return std::vector<int> the index of each dims
          */
-        void print_all(const char *message, const bool with_padding = false)
-        {
-            int y_end;
-            int x_end;
-            int c_end;
-            if (with_padding)
-            {
-                y_end = this->shape_with_padding[0];
-                x_end = this->shape_with_padding[1];
-                c_end = this->shape_with_padding[2];
-            }
-            else
-            {
-                y_end = this->shape[0];
-                x_end = this->shape[1];
-                c_end = this->shape[2];
-            }
+        std::vector<int> get_axis_index(int element_index);
 
-            printf("\n%s | ", message);
-            this->print_shape();
-
-            for (int y = 0; y < y_end; y++)
-            {
-                for (int x = 0; x < x_end; x++)
-                {
-                    for (int c = 0; c < c_end; c++)
-                        printf("%d ", this->get_element_value({y, x, c}, with_padding));
-                }
-            }
-            printf("\n");
-            return;
-        }
+        /**
+         * @brief Get the index of element 
+         * 
+         * @param axis_index the index of each dims
+         * @return int the index of element 
+         */
+        int get_element_index(const std::vector<int> axis_index);
 
         /**
          * @brief Check the element value with input ground-truth.
@@ -638,35 +386,39 @@ namespace dl
          * @param gt_element ground-truth value of element
          * @param bias permissible error
          * @param info one of true or false
-         *             - true: print shape and result
+         *             - true:  shape and result
          *             - false: do not
+         * @param failed_number maximum number of wrong element that will be printed
+         * 
          * @return 
          *         - true: in permissible error
          *         - false: not 
          */
-        bool check_element(T *gt_element, int bias = 2, bool info = true)
+        bool check_element(T *gt_element, int bias = 2, bool info = true, int failed_number = 0)
         {
+            int count = 0;
             if (info)
                 this->print_shape();
-            int i = 0;
-            for (int y = 0; y < this->shape[0]; y++)
+            int size = this->get_size();
+            for (int i = 0; i < size; i++)
             {
-                for (int x = 0; x < this->shape[1]; x++)
+                if (DL_ABS(this->element[i] - gt_element[i]) > bias)
                 {
-                    for (int c = 0; c < this->shape[2]; c++)
+                    std::vector<int> index = get_axis_index(i);
+                    std::cout << "element[";
+                    for (int j = 0; j < index.size() - 1; j++)
                     {
-                        int a = this->get_element_value({y, x, c});
-                        int b = gt_element[i];
-                        int offset = DL_ABS(a - b);
-                        if (offset > bias)
-                        {
-                            printf("element[%d, %d, %d]: %d v.s. %d\n", y, x, c, a, b);
-                            return false;
-                        }
-                        i++;
+                        std::cout << index[j] << ", ";
                     }
+                    std::cout << index.back() << "]: ";
+                    std::cout << +this->element[i] << " v.s. " << +gt_element[i] << "\n";
+                    count++;
+                    if (count > failed_number)
+                        return false;
                 }
             }
+            if (count)
+                return false;
 
             if (info)
                 printf("PASS\n");
@@ -700,35 +452,44 @@ namespace dl
 
         Tensor<T> &operator=(const Tensor<T> &input)
         {
-            this->size = input.size;
             this->auto_free = input.auto_free;
             this->exponent = input.exponent;
-            this->shape = input.shape;
-            this->padding = input.padding;
-            int size_real_tmp = this->shape_with_padding.size() ? this->shape_with_padding[0] * this->shape_with_padding[1] * this->shape_with_padding[2] : 0;
-            int size_input_real = input.shape_with_padding.size() ? input.shape_with_padding[0] * input.shape_with_padding[1] * input.shape_with_padding[2] : 0;
-            this->shape_with_padding = input.shape_with_padding;
-            if (this->element)
+            int size_real_tmp = this->size;
+            int size_input_real = input.size;
+            this->set_shape(input.shape);
+            if (input.element)
             {
-                if (size_real_tmp != size_input_real)
+                if (this->element)
                 {
-                    tool::free_aligned(this->element);
-                    T *new_element = (T *)tool::calloc_aligned(size_input_real, sizeof(T), 16);
-                    tool::copy_memory(new_element, input.element, size_input_real * sizeof(T));
-                    this->element = new_element;
+                    if (size_real_tmp != size_input_real)
+                    {
+                        tool::free_aligned_prefer(this->element);
+                        T *new_element = (T *)tool::malloc_aligned_prefer(size_input_real, sizeof(T), 16);
+                        tool::copy_memory(new_element, input.element, size_input_real * sizeof(T));
+                        this->element = new_element;
+                    }
+                    else
+                    {
+                        tool::copy_memory(this->element, input.element, size_input_real * sizeof(T));
+                    }
                 }
                 else
                 {
-                    tool::copy_memory(this->element, input.element, size_input_real * sizeof(T));
+                    T *new_element = (T *)tool::malloc_aligned_prefer(size_input_real, sizeof(T), 16);
+                    tool::copy_memory(new_element, input.element, size_input_real * sizeof(T));
+                    this->element = new_element;
                 }
+                return *this;
             }
             else
             {
-                T *new_element = (T *)tool::calloc_aligned(size_input_real, sizeof(T), 16);
-                tool::copy_memory(new_element, input.element, size_input_real * sizeof(T));
-                this->element = new_element;
+                if (this->element)
+                {
+                    tool::free_aligned_prefer(this->element);
+                    this->element = NULL;
+                }
+                return *this;
             }
-            return *this;
         }
     };
 } // namespace dl
\ No newline at end of file
diff --git a/tools/sdk/esp32s2/include/esp_https_server/include/esp_https_server.h b/tools/sdk/esp32s2/include/esp_https_server/include/esp_https_server.h
index 00565a963ad..efe726e9e70 100644
--- a/tools/sdk/esp32s2/include/esp_https_server/include/esp_https_server.h
+++ b/tools/sdk/esp32s2/include/esp_https_server/include/esp_https_server.h
@@ -10,6 +10,7 @@
 #include <stdbool.h>
 #include "esp_err.h"
 #include "esp_http_server.h"
+#include "esp_tls.h"
 
 #ifdef __cplusplus
 extern "C" {
@@ -20,6 +21,22 @@ typedef enum {
     HTTPD_SSL_TRANSPORT_INSECURE     // SSL disabled
 } httpd_ssl_transport_mode_t;
 
+/**
+ * @brief Callback data struct, contains the ESP-TLS connection handle
+ */
+typedef struct esp_https_server_user_cb_arg {
+    const esp_tls_t *tls;
+} esp_https_server_user_cb_arg_t;
+
+/**
+ * @brief Callback function prototype
+ * Can be used to get connection or client information (SSL context)
+ * E.g. Client certificate, Socket FD, Connection state, etc.
+ *
+ * @param user_cb Callback data struct
+ */
+typedef void esp_https_server_user_cb(esp_https_server_user_cb_arg_t *user_cb);
+
 /**
  * HTTPS server config struct
  *
@@ -66,6 +83,9 @@ struct httpd_ssl_config {
 
     /** Enable tls session tickets */
     bool session_tickets;
+
+    /** User callback for esp_https_server */
+    esp_https_server_user_cb *user_cb;
 };
 
 typedef struct httpd_ssl_config httpd_ssl_config_t;
@@ -113,6 +133,7 @@ typedef struct httpd_ssl_config httpd_ssl_config_t;
     .port_secure = 443,                           \
     .port_insecure = 80,                          \
     .session_tickets = false,                     \
+    .user_cb = NULL,                              \
 }
 
 /**
diff --git a/tools/sdk/esp32s2/include/esp_hw_support/include/esp_sleep.h b/tools/sdk/esp32s2/include/esp_hw_support/include/esp_sleep.h
index cfdfbc04186..75d2f9726c3 100644
--- a/tools/sdk/esp32s2/include/esp_hw_support/include/esp_sleep.h
+++ b/tools/sdk/esp32s2/include/esp_hw_support/include/esp_sleep.h
@@ -44,6 +44,7 @@ typedef enum {
 #if SOC_PM_SUPPORT_CPU_PD
     ESP_PD_DOMAIN_CPU,             //!< CPU core
 #endif
+    ESP_PD_DOMAIN_RTC8M,           //!< Internal 8M oscillator
     ESP_PD_DOMAIN_VDDSDIO,         //!< VDD_SDIO
     ESP_PD_DOMAIN_MAX              //!< Number of domains
 } esp_sleep_pd_domain_t;
diff --git a/tools/sdk/esp32s2/include/esp_lcd/include/esp_lcd_panel_io.h b/tools/sdk/esp32s2/include/esp_lcd/include/esp_lcd_panel_io.h
index eebcabf42b4..330f4d9a165 100644
--- a/tools/sdk/esp32s2/include/esp_lcd/include/esp_lcd_panel_io.h
+++ b/tools/sdk/esp32s2/include/esp_lcd/include/esp_lcd_panel_io.h
@@ -65,6 +65,22 @@ esp_err_t esp_lcd_panel_io_tx_color(esp_lcd_panel_io_handle_t io, int lcd_cmd, c
  */
 esp_err_t esp_lcd_panel_io_del(esp_lcd_panel_io_handle_t io);
 
+/**
+ * @brief Type of LCD panel IO event data
+ */
+typedef struct {
+} esp_lcd_panel_io_event_data_t;
+
+/**
+ * @brief Declare the prototype of the function that will be invoked when panel IO finishes transferring color data
+ *
+ * @param[in] panel_io LCD panel IO handle, which is created by factory API like `esp_lcd_new_panel_io_spi()`
+ * @param[in] edata Panel IO event data, fed by driver
+ * @param[in] user_ctx User data, passed from `esp_lcd_panel_io_xxx_config_t`
+ * @return Whether a high priority task has been waken up by this function
+ */
+typedef bool (*esp_lcd_panel_io_color_trans_done_cb_t)(esp_lcd_panel_io_handle_t panel_io, esp_lcd_panel_io_event_data_t *edata, void *user_ctx);
+
 /**
  * @brief Panel IO configuration structure, for SPI interface
  */
@@ -74,8 +90,8 @@ typedef struct {
     int spi_mode;    /*!< Traditional SPI mode (0~3) */
     unsigned int pclk_hz;    /*!< Frequency of pixel clock */
     size_t trans_queue_depth; /*!< Size of internal transaction queue */
-    bool (*on_color_trans_done)(esp_lcd_panel_io_handle_t panel_io, void *user_data, void *event_data); /*!< Callback, invoked when color data transfer has finished */
-    void *user_data;    /*!< User private data, passed directly to on_trans_frame_done's user_data */
+    esp_lcd_panel_io_color_trans_done_cb_t on_color_trans_done; /*!< Callback invoked when color data transfer has finished */
+    void *user_ctx;    /*!< User private data, passed directly to on_color_trans_done's user_ctx */
     int lcd_cmd_bits;   /*!< Bit-width of LCD command */
     int lcd_param_bits; /*!< Bit-width of LCD parameter */
     struct {
@@ -100,8 +116,8 @@ esp_err_t esp_lcd_new_panel_io_spi(esp_lcd_spi_bus_handle_t bus, const esp_lcd_p
 
 typedef struct {
     uint32_t dev_addr; /*!< I2C device address */
-    bool (*on_color_trans_done)(esp_lcd_panel_io_handle_t panel_io, void *user_data, void *event_data); /*!< Callback, invoked when color data transfer has finished */
-    void *user_data; /*!< User private data, passed directly to on_trans_frame_done's user_data */
+    esp_lcd_panel_io_color_trans_done_cb_t on_color_trans_done; /*!< Callback invoked when color data transfer has finished */
+    void *user_ctx; /*!< User private data, passed directly to on_color_trans_done's user_ctx */
     size_t control_phase_bytes; /*!< I2C LCD panel will encode control information (e.g. D/C seclection) into control phase, in several bytes */
     unsigned int dc_bit_offset; /*!< Offset of the D/C selection bit in control phase */
     int lcd_cmd_bits;           /*!< Bit-width of LCD command */
@@ -168,8 +184,8 @@ typedef struct {
     int cs_gpio_num;         /*!< GPIO used for CS line, set to -1 will declaim exclusively use of I80 bus */
     unsigned int pclk_hz;    /*!< Frequency of pixel clock */
     size_t trans_queue_depth; /*!< Transaction queue size, larger queue, higher throughput */
-    bool (*on_color_trans_done)(esp_lcd_panel_io_handle_t panel_io, void *user_data, void *event_data); /*!< Callback, invoked when color data was tranferred done */
-    void *user_data;    /*!< User private data, passed directly to on_trans_done's user_data */
+    esp_lcd_panel_io_color_trans_done_cb_t on_color_trans_done; /*!< Callback invoked when color data was tranferred done */
+    void *user_ctx;    /*!< User private data, passed directly to on_color_trans_done's user_ctx */
     int lcd_cmd_bits;   /*!< Bit-width of LCD command */
     int lcd_param_bits; /*!< Bit-width of LCD parameter */
     struct {
diff --git a/tools/sdk/esp32s2/include/esp_lcd/include/esp_lcd_panel_rgb.h b/tools/sdk/esp32s2/include/esp_lcd/include/esp_lcd_panel_rgb.h
index 2ddd2b6b9a6..1368bb787f6 100644
--- a/tools/sdk/esp32s2/include/esp_lcd/include/esp_lcd_panel_rgb.h
+++ b/tools/sdk/esp32s2/include/esp_lcd/include/esp_lcd_panel_rgb.h
@@ -18,6 +18,37 @@ extern "C" {
 #if SOC_LCD_RGB_SUPPORTED
 /**
  * @brief LCD RGB timing structure
+ *
+ *                                                 Total Width
+ *                             <--------------------------------------------------->
+ *                       Hsync width HBP             Active Width                HFP
+ *                             <---><--><--------------------------------------><--->
+ *                         ____    ____|_______________________________________|____|
+ *                             |___|   |                                       |    |
+ *                                     |                                       |    |
+ *                         __|         |                                       |    |
+ *            /|\    /|\  |            |                                       |    |
+ *             | VSYNC|   |            |                                       |    |
+ *             |Width\|/  |__          |                                       |    |
+ *             |     /|\     |         |                                       |    |
+ *             |  VBP |      |         |                                       |    |
+ *             |     \|/_____|_________|_______________________________________|    |
+ *             |     /|\     |         | / / / / / / / / / / / / / / / / / / / |    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *    Total    |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *    Heigh    |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |Active|      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |Heigh |      |         |/ / / / / / Active Display Area / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |      |      |         |/ / / / / / / / / / / / / / / / / / / /|    |
+ *             |     \|/_____|_________|_______________________________________|    |
+ *             |     /|\     |                                                      |
+ *             |  VFP |      |                                                      |
+ *            \|/    \|/_____|______________________________________________________|
+ *
  */
 typedef struct {
     unsigned int pclk_hz;           /*!< Frequency of pixel clock */
@@ -38,6 +69,22 @@ typedef struct {
     } flags;
 } esp_lcd_rgb_timing_t;
 
+/**
+ * @brief Type of RGB LCD panel event data
+ */
+typedef struct {
+} esp_lcd_rgb_panel_event_data_t;
+
+/**
+ * @brief Declare the prototype of the function that will be invoked when panel IO finishes transferring color data
+ *
+ * @param[in] panel LCD panel handle, returned from `esp_lcd_new_rgb_panel`
+ * @param[in] edata Panel event data, fed by driver
+ * @param[in] user_ctx User data, passed from `esp_lcd_rgb_panel_config_t`
+ * @return Whether a high priority task has been waken up by this function
+ */
+typedef bool (*esp_lcd_rgb_panel_frame_trans_done_cb_t)(esp_lcd_panel_handle_t panel, esp_lcd_rgb_panel_event_data_t *edata, void *user_ctx);
+
 /**
  * @brief LCD RGB panel configuration structure
  */
@@ -51,8 +98,8 @@ typedef struct {
     int pclk_gpio_num;            /*!< GPIO used for PCLK signal */
     int data_gpio_nums[SOC_LCD_RGB_DATA_WIDTH]; /*!< GPIOs used for data lines */
     int disp_gpio_num; /*!< GPIO used for display control signal, set to -1 if it's not used */
-    bool (*on_frame_trans_done)(esp_lcd_panel_handle_t panel, void *user_data); /*!< Callback, invoked when one frame buffer has transferred done */
-    void *user_data; /*!< User data which would be passed to on_frame_trans_done's user_data */
+    esp_lcd_rgb_panel_frame_trans_done_cb_t on_frame_trans_done; /*!< Callback invoked when one frame buffer has transferred done */
+    void *user_ctx; /*!< User data which would be passed to on_frame_trans_done's user_ctx */
     struct {
         unsigned int disp_active_low: 1; /*!< If this flag is enabled, a low level of display control signal can turn the screen on; vice versa */
         unsigned int relax_on_idle: 1;   /*!< If this flag is enabled, the host won't refresh the LCD if nothing changed in host's frame buffer (this is usefull for LCD with built-in GRAM) */
diff --git a/tools/sdk/esp32s2/include/esp_littlefs/include/esp_littlefs.h b/tools/sdk/esp32s2/include/esp_littlefs/include/esp_littlefs.h
index dbb1028790d..7b5abe9248f 100644
--- a/tools/sdk/esp32s2/include/esp_littlefs/include/esp_littlefs.h
+++ b/tools/sdk/esp32s2/include/esp_littlefs/include/esp_littlefs.h
@@ -14,7 +14,7 @@
 #include <sys/stat.h>
 #include <sys/time.h>
 #include <sys/poll.h>
-#include <dirent.h>
+#include <sys/dirent.h>
 #include <string.h>
 #include "sdkconfig.h"
 
diff --git a/tools/sdk/esp32s2/include/esp_wifi/include/esp_coexist.h b/tools/sdk/esp32s2/include/esp_wifi/include/esp_coexist.h
index 1ff624d9b55..14b7c9aa506 100644
--- a/tools/sdk/esp32s2/include/esp_wifi/include/esp_coexist.h
+++ b/tools/sdk/esp32s2/include/esp_wifi/include/esp_coexist.h
@@ -1,16 +1,8 @@
-// Copyright 2015-2018 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #ifndef __ESP_COEXIST_H__
 #define __ESP_COEXIST_H__
@@ -32,6 +24,13 @@ typedef enum {
     ESP_COEX_PREFER_NUM,            /*!< Prefer value numbers */
 } esp_coex_prefer_t;
 
+typedef enum {
+    EXTERN_COEX_WIRE_1 = 0,
+    EXTERN_COEX_WIRE_2,
+    EXTERN_COEX_WIRE_3,
+    EXTERN_COEX_WIRE_NUM,
+} external_coex_wire_t;
+
 /**
  * @brief coex status type
  */
@@ -41,6 +40,36 @@ typedef enum {
     ESP_COEX_ST_TYPE_BT,
 } esp_coex_status_type_t;
 
+/**
+ * @brief external coex gpio pti
+ */
+typedef struct {
+    int32_t in_pin0;
+    int32_t in_pin1;
+    int32_t out_pin0;
+} esp_external_coex_gpio_set_t;
+
+/**
+ * @brief external coex pti level
+ */
+typedef enum {
+    EXTERN_COEX_PTI_MID = 0,
+    EXTERN_COEX_PTI_HIGH,
+    EXTERN_COEX_PTI_NUM,
+} esp_coex_pti_level_t;
+
+/**
+ * @brief external coex pti
+ */
+typedef struct {
+    uint32_t in_pti1;
+    uint32_t in_pti2;
+    uint32_t in_pti3;
+    uint32_t out_pti1;
+    uint32_t out_pti2;
+    uint32_t out_pti3;
+} esp_external_coex_pti_set_t;
+
 #define ESP_COEX_BLE_ST_MESH_CONFIG        0x08
 #define ESP_COEX_BLE_ST_MESH_TRAFFIC       0x10
 #define ESP_COEX_BLE_ST_MESH_STANDBY       0x20
@@ -84,6 +113,18 @@ esp_err_t esp_coex_status_bit_set(esp_coex_status_type_t type, uint32_t status);
  */
 esp_err_t esp_coex_status_bit_clear(esp_coex_status_type_t type, uint32_t status);
 
+#if CONFIG_EXTERNAL_COEX_ENABLE
+/**
+ * @brief Setup gpio pin and corresponding pti level, start external coex.
+ * @param wire_type : to select the whole external coex gpio number.
+ * @param gpio_pin : gpio pin number to choose.
+ * @return : ESP_OK - success, other - failed
+ */
+esp_err_t esp_enable_extern_coex_gpio_pin(external_coex_wire_t wire_type,
+                    esp_external_coex_gpio_set_t gpio_pin);
+
+esp_err_t esp_disable_extern_coex_gpio_pin();
+#endif
 
 #ifdef __cplusplus
 }
diff --git a/tools/sdk/esp32s2/include/esp_wifi/include/esp_coexist_internal.h b/tools/sdk/esp32s2/include/esp_wifi/include/esp_coexist_internal.h
index 3501f0da6fb..7ba06d4c690 100644
--- a/tools/sdk/esp32s2/include/esp_wifi/include/esp_coexist_internal.h
+++ b/tools/sdk/esp32s2/include/esp_wifi/include/esp_coexist_internal.h
@@ -1,21 +1,14 @@
-// Copyright 2018-2018 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2018-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #ifndef __ESP_COEXIST_INTERNAL_H__
 #define __ESP_COEXIST_INTERNAL_H__
 
 #include <stdbool.h>
+#include "esp_coexist.h"
 #include "esp_coexist_adapter.h"
 
 #ifdef __cplusplus
@@ -210,6 +203,29 @@ int coex_schm_curr_phase_idx_get(void);
  */
 esp_err_t esp_coex_adapter_register(coex_adapter_funcs_t *funcs);
 
+#if CONFIG_EXTERNAL_COEX_ENABLE
+/**
+  * @brief     Set external coexistence pti level and enable it.
+  *
+  * @param     level1    external coex low pti
+  * @param     level2    external coex mid pti
+  * @param     level3    external coex high pti
+  *
+  * @return
+  *    - ESP_OK: succeed
+  */
+esp_err_t esp_coex_external_set(esp_coex_pti_level_t level1,
+         esp_coex_pti_level_t level2, esp_coex_pti_level_t level3);
+
+/**
+  * @brief     Disable external coexist
+  *
+  * @return
+  *    - ESP_OK: succeed
+  */
+void esp_coex_external_stop(void);
+#endif    /*External Coex*/
+
 /**
   * @brief     Check the MD5 values of the coexistence adapter header files in IDF and WiFi library
   *
diff --git a/tools/sdk/esp32s2/include/esp_wifi/include/esp_wifi_types.h b/tools/sdk/esp32s2/include/esp_wifi/include/esp_wifi_types.h
index 503e8d7bb05..cee23f3fafc 100644
--- a/tools/sdk/esp32s2/include/esp_wifi/include/esp_wifi_types.h
+++ b/tools/sdk/esp32s2/include/esp_wifi/include/esp_wifi_types.h
@@ -1,16 +1,8 @@
-// Copyright 2015-2016 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 
 #ifndef __ESP_WIFI_TYPES_H__
@@ -80,6 +72,7 @@ typedef enum {
     WIFI_REASON_ASSOC_NOT_AUTHED         = 9,
     WIFI_REASON_DISASSOC_PWRCAP_BAD      = 10,
     WIFI_REASON_DISASSOC_SUPCHAN_BAD     = 11,
+    WIFI_REASON_BSS_TRANSITION_DISASSOC  = 12,
     WIFI_REASON_IE_INVALID               = 13,
     WIFI_REASON_MIC_FAILURE              = 14,
     WIFI_REASON_4WAY_HANDSHAKE_TIMEOUT   = 15,
@@ -250,7 +243,8 @@ typedef struct {
     wifi_pmf_config_t pmf_cfg;    /**< Configuration for Protected Management Frame. Will be advertized in RSN Capabilities in RSN IE. */
     uint32_t rm_enabled:1;        /**< Whether Radio Measurements are enabled for the connection */
     uint32_t btm_enabled:1;       /**< Whether BSS Transition Management is enabled for the connection */
-    uint32_t reserved:30;         /**< Reserved for future feature set */
+    uint32_t mbo_enabled:1;       /**< Whether MBO is enabled for the connection */
+    uint32_t reserved:29;         /**< Reserved for future feature set */
 } wifi_sta_config_t;
 
 /** @brief Configuration data for ESP32 AP or STA.
diff --git a/tools/sdk/esp32s2/include/freertos/include/esp_additions/freertos/FreeRTOSConfig.h b/tools/sdk/esp32s2/include/freertos/include/esp_additions/freertos/FreeRTOSConfig.h
index a01a56e9fd4..675af141ebc 100644
--- a/tools/sdk/esp32s2/include/freertos/include/esp_additions/freertos/FreeRTOSConfig.h
+++ b/tools/sdk/esp32s2/include/freertos/include/esp_additions/freertos/FreeRTOSConfig.h
@@ -90,7 +90,6 @@
 #define portNUM_PROCESSORS                              1
 #endif
 
-#define configASSERT_2                                  0
 #define portUSING_MPU_WRAPPERS                          0
 #define configUSE_MUTEX                                 1
 
@@ -206,7 +205,6 @@
 #define configGENERATE_RUN_TIME_STATS                   1       /* Used by vTaskGetRunTimeStats() */
 #endif
 
-#define configUSE_TRACE_FACILITY_2                      0
 #define configBENCHMARK                                 0
 #define configUSE_16_BIT_TICKS                          0
 #define configIDLE_SHOULD_YIELD                         0
@@ -306,4 +304,9 @@ extern void vPortCleanUpTCB ( void *pxTCB );
 
 #define configTASK_NOTIFICATION_ARRAY_ENTRIES           1
 
+// backward compatibility for 4.4
+#define xTaskRemoveFromUnorderedEventList vTaskRemoveFromUnorderedEventList
+
+#define configNUM_CORES                                 portNUM_PROCESSORS
+
 #endif /* FREERTOS_CONFIG_H */
diff --git a/tools/sdk/esp32s2/include/freertos/include/freertos/event_groups.h b/tools/sdk/esp32s2/include/freertos/include/freertos/event_groups.h
index 84505ddaaa0..9792296e566 100644
--- a/tools/sdk/esp32s2/include/freertos/include/freertos/event_groups.h
+++ b/tools/sdk/esp32s2/include/freertos/include/freertos/event_groups.h
@@ -64,7 +64,7 @@
  * used to create a synchronisation point between multiple tasks (a
  * 'rendezvous').
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup EventGroup EventGroup
  * @endcond
  */
@@ -78,7 +78,7 @@
  * xEventGroupCreate() returns an EventGroupHandle_t variable that can then
  * be used as a parameter to other event group functions.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup EventGroupHandle_t EventGroupHandle_t
  * @endcond
  * \ingroup EventGroup
@@ -94,7 +94,7 @@ typedef struct EventGroupDef_t * EventGroupHandle_t;
  * number of bits it holds is set by configUSE_16_BIT_TICKS (16 bits if set to 1,
  * 32 bits if set to 0.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup EventBits_t EventBits_t
  * @endcond
  * \ingroup EventGroup
@@ -102,7 +102,7 @@ typedef struct EventGroupDef_t * EventGroupHandle_t;
 typedef TickType_t               EventBits_t;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  * EventGroupHandle_t xEventGroupCreate( void );
@@ -152,7 +152,7 @@ typedef TickType_t               EventBits_t;
  *      // The event group was created.
  *  }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupCreate xEventGroupCreate
  * @endcond
  * \ingroup EventGroup
@@ -162,7 +162,7 @@ typedef TickType_t               EventBits_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  * EventGroupHandle_t xEventGroupCreateStatic( EventGroupHandle_t * pxEventGroupBuffer );
@@ -217,7 +217,7 @@ typedef TickType_t               EventBits_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupWaitBits(    EventGroupHandle_t xEventGroup,
@@ -307,7 +307,7 @@ typedef TickType_t               EventBits_t;
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupWaitBits xEventGroupWaitBits
  * @endcond
  * \ingroup EventGroup
@@ -319,7 +319,7 @@ EventBits_t xEventGroupWaitBits( EventGroupHandle_t xEventGroup,
                                  TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup, const EventBits_t uxBitsToClear );
@@ -372,7 +372,7 @@ EventBits_t xEventGroupWaitBits( EventGroupHandle_t xEventGroup,
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupClearBits xEventGroupClearBits
  * @endcond
  * \ingroup EventGroup
@@ -381,7 +381,7 @@ EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup,
                                   const EventBits_t uxBitsToClear ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  BaseType_t xEventGroupClearBitsFromISR( EventGroupHandle_t xEventGroup, const EventBits_t uxBitsToSet );
@@ -432,7 +432,7 @@ EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup,
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupClearBitsFromISR xEventGroupClearBitsFromISR
  * @endcond
  * \ingroup EventGroup
@@ -446,7 +446,7 @@ EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup,
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup, const EventBits_t uxBitsToSet );
@@ -516,7 +516,7 @@ EventBits_t xEventGroupClearBits( EventGroupHandle_t xEventGroup,
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupSetBits xEventGroupSetBits
  * @endcond
  * \ingroup EventGroup
@@ -525,7 +525,7 @@ EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup,
                                 const EventBits_t uxBitsToSet ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  BaseType_t xEventGroupSetBitsFromISR( EventGroupHandle_t xEventGroup, const EventBits_t uxBitsToSet, BaseType_t *pxHigherPriorityTaskWoken );
@@ -595,7 +595,7 @@ EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup,
  *      }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupSetBitsFromISR xEventGroupSetBitsFromISR
  * @endcond
  * \ingroup EventGroup
@@ -610,7 +610,7 @@ EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup,
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupSync(    EventGroupHandle_t xEventGroup,
@@ -732,7 +732,7 @@ EventBits_t xEventGroupSetBits( EventGroupHandle_t xEventGroup,
  * }
  *
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupSync xEventGroupSync
  * @endcond
  * \ingroup EventGroup
@@ -744,7 +744,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupGetBits( EventGroupHandle_t xEventGroup );
@@ -758,7 +758,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
  *
  * @return The event group bits at the time xEventGroupGetBits() was called.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupGetBits xEventGroupGetBits
  * @endcond
  * \ingroup EventGroup
@@ -766,7 +766,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
 #define xEventGroupGetBits( xEventGroup )    xEventGroupClearBits( xEventGroup, 0 )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  EventBits_t xEventGroupGetBitsFromISR( EventGroupHandle_t xEventGroup );
@@ -779,7 +779,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
  *
  * @return The event group bits at the time xEventGroupGetBitsFromISR() was called.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xEventGroupGetBitsFromISR xEventGroupGetBitsFromISR
  * @endcond
  * \ingroup EventGroup
@@ -787,7 +787,7 @@ EventBits_t xEventGroupSync( EventGroupHandle_t xEventGroup,
 EventBits_t xEventGroupGetBitsFromISR( EventGroupHandle_t xEventGroup ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * event_groups.h
  * @code{c}
  *  void xEventGroupDelete( EventGroupHandle_t xEventGroup );
@@ -802,7 +802,7 @@ EventBits_t xEventGroupGetBitsFromISR( EventGroupHandle_t xEventGroup ) PRIVILEG
  */
 void vEventGroupDelete( EventGroupHandle_t xEventGroup ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 
 /* For internal use only. */
 void vEventGroupSetBitsCallback( void * pvEventGroup,
diff --git a/tools/sdk/esp32s2/include/freertos/include/freertos/message_buffer.h b/tools/sdk/esp32s2/include/freertos/include/freertos/message_buffer.h
index e57c589fbac..af5c3290b7c 100644
--- a/tools/sdk/esp32s2/include/freertos/include/freertos/message_buffer.h
+++ b/tools/sdk/esp32s2/include/freertos/include/freertos/message_buffer.h
@@ -85,7 +85,7 @@ typedef void * MessageBufferHandle_t;
 /*-----------------------------------------------------------*/
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -139,7 +139,7 @@ typedef void * MessageBufferHandle_t;
  *  }
  *
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferCreate xMessageBufferCreate
  * @endcond
  * \ingroup MessageBufferManagement
@@ -148,7 +148,7 @@ typedef void * MessageBufferHandle_t;
     ( MessageBufferHandle_t ) xStreamBufferGenericCreate( xBufferSizeBytes, ( size_t ) 0, pdTRUE )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -210,7 +210,7 @@ typedef void * MessageBufferHandle_t;
  * }
  *
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferCreateStatic xMessageBufferCreateStatic
  * @endcond
  * \ingroup MessageBufferManagement
@@ -219,7 +219,7 @@ typedef void * MessageBufferHandle_t;
     ( MessageBufferHandle_t ) xStreamBufferGenericCreateStatic( xBufferSizeBytes, 0, pdTRUE, pucMessageBufferStorageArea, pxStaticMessageBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -314,7 +314,7 @@ typedef void * MessageBufferHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferSend xMessageBufferSend
  * @endcond
  * \ingroup MessageBufferManagement
@@ -323,7 +323,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferSend( ( StreamBufferHandle_t ) xMessageBuffer, pvTxData, xDataLengthBytes, xTicksToWait )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -423,7 +423,7 @@ typedef void * MessageBufferHandle_t;
  *  portYIELD_FROM_ISR( xHigherPriorityTaskWoken );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferSendFromISR xMessageBufferSendFromISR
  * @endcond
  * \ingroup MessageBufferManagement
@@ -432,7 +432,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferSendFromISR( ( StreamBufferHandle_t ) xMessageBuffer, pvTxData, xDataLengthBytes, pxHigherPriorityTaskWoken )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -516,7 +516,7 @@ typedef void * MessageBufferHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferReceive xMessageBufferReceive
  * @endcond
  * \ingroup MessageBufferManagement
@@ -526,7 +526,7 @@ typedef void * MessageBufferHandle_t;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -622,7 +622,7 @@ typedef void * MessageBufferHandle_t;
  *  portYIELD_FROM_ISR( xHigherPriorityTaskWoken );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferReceiveFromISR xMessageBufferReceiveFromISR
  * @endcond
  * \ingroup MessageBufferManagement
@@ -631,7 +631,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferReceiveFromISR( ( StreamBufferHandle_t ) xMessageBuffer, pvRxData, xBufferLengthBytes, pxHigherPriorityTaskWoken )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -654,7 +654,7 @@ typedef void * MessageBufferHandle_t;
     vStreamBufferDelete( ( StreamBufferHandle_t ) xMessageBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * BaseType_t xMessageBufferIsFull( MessageBufferHandle_t xMessageBuffer ) );
@@ -674,7 +674,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferIsFull( ( StreamBufferHandle_t ) xMessageBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * BaseType_t xMessageBufferIsEmpty( MessageBufferHandle_t xMessageBuffer ) );
@@ -693,7 +693,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferIsEmpty( ( StreamBufferHandle_t ) xMessageBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * BaseType_t xMessageBufferReset( MessageBufferHandle_t xMessageBuffer );
@@ -712,7 +712,7 @@ typedef void * MessageBufferHandle_t;
  * the message queue to wait for space to become available, or to wait for a
  * a message to be available, then pdFAIL is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferReset xMessageBufferReset
  * @endcond
  * \ingroup MessageBufferManagement
@@ -722,7 +722,7 @@ typedef void * MessageBufferHandle_t;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * size_t xMessageBufferSpaceAvailable( MessageBufferHandle_t xMessageBuffer ) );
@@ -740,7 +740,7 @@ typedef void * MessageBufferHandle_t;
  * architecture, so if xMessageBufferSpacesAvailable() returns 10, then the size
  * of the largest message that can be written to the message buffer is 6 bytes.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferSpaceAvailable xMessageBufferSpaceAvailable
  * @endcond
  * \ingroup MessageBufferManagement
@@ -751,7 +751,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferSpacesAvailable( ( StreamBufferHandle_t ) xMessageBuffer ) /* Corrects typo in original macro name. */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  * @code{c}
  * size_t xMessageBufferNextLengthBytes( MessageBufferHandle_t xMessageBuffer ) );
@@ -767,7 +767,7 @@ typedef void * MessageBufferHandle_t;
  * @return The length (in bytes) of the next message in the message buffer, or 0
  * if the message buffer is empty.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferNextLengthBytes xMessageBufferNextLengthBytes
  * @endcond
  * \ingroup MessageBufferManagement
@@ -776,7 +776,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferNextMessageLengthBytes( ( StreamBufferHandle_t ) xMessageBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -811,7 +811,7 @@ typedef void * MessageBufferHandle_t;
  * @return If a task was removed from the Blocked state then pdTRUE is returned.
  * Otherwise pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferSendCompletedFromISR xMessageBufferSendCompletedFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -820,7 +820,7 @@ typedef void * MessageBufferHandle_t;
     xStreamBufferSendCompletedFromISR( ( StreamBufferHandle_t ) xMessageBuffer, pxHigherPriorityTaskWoken )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -856,7 +856,7 @@ typedef void * MessageBufferHandle_t;
  * @return If a task was removed from the Blocked state then pdTRUE is returned.
  * Otherwise pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xMessageBufferReceiveCompletedFromISR xMessageBufferReceiveCompletedFromISR
  * @endcond
  * \ingroup StreamBufferManagement
diff --git a/tools/sdk/esp32s2/include/freertos/include/freertos/queue.h b/tools/sdk/esp32s2/include/freertos/include/freertos/queue.h
index 81cccc05df3..05ca7de4546 100644
--- a/tools/sdk/esp32s2/include/freertos/include/freertos/queue.h
+++ b/tools/sdk/esp32s2/include/freertos/include/freertos/queue.h
@@ -62,7 +62,7 @@ typedef struct QueueDefinition   * QueueSetHandle_t;
  */
 typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 
 /* For internal use only. */
 #define queueSEND_TO_BACK                     ( ( BaseType_t ) 0 )
@@ -80,7 +80,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 /** @endcond */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * QueueHandle_t xQueueCreate(
@@ -146,7 +146,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueCreate xQueueCreate
  * @endcond
  * \ingroup QueueManagement
@@ -156,7 +156,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * QueueHandle_t xQueueCreateStatic(
@@ -235,7 +235,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueCreateStatic xQueueCreateStatic
  * @endcond
  * \ingroup QueueManagement
@@ -245,7 +245,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendToToFront(
@@ -321,7 +321,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSend xQueueSend
  * @endcond
  * \ingroup QueueManagement
@@ -330,7 +330,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
     xQueueGenericSend( ( xQueue ), ( pvItemToQueue ), ( xTicksToWait ), queueSEND_TO_FRONT )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendToBack(
@@ -408,7 +408,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSend xQueueSend
  * @endcond
  * \ingroup QueueManagement
@@ -417,7 +417,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
     xQueueGenericSend( ( xQueue ), ( pvItemToQueue ), ( xTicksToWait ), queueSEND_TO_BACK )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSend(
@@ -497,7 +497,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSend xQueueSend
  * @endcond
  * \ingroup QueueManagement
@@ -506,7 +506,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
     xQueueGenericSend( ( xQueue ), ( pvItemToQueue ), ( xTicksToWait ), queueSEND_TO_BACK )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueOverwrite(
@@ -585,7 +585,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ...
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueOverwrite xQueueOverwrite
  * @endcond
  * \ingroup QueueManagement
@@ -595,7 +595,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueGenericSend(
@@ -678,7 +678,7 @@ typedef struct QueueDefinition   * QueueSetMemberHandle_t;
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSend xQueueSend
  * @endcond
  * \ingroup QueueManagement
@@ -689,7 +689,7 @@ BaseType_t xQueueGenericSend( QueueHandle_t xQueue,
                               const BaseType_t xCopyPosition ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueuePeek(
@@ -780,7 +780,7 @@ BaseType_t xQueueGenericSend( QueueHandle_t xQueue,
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueuePeek xQueuePeek
  * @endcond
  * \ingroup QueueManagement
@@ -790,7 +790,7 @@ BaseType_t xQueuePeek( QueueHandle_t xQueue,
                        TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueuePeekFromISR(
@@ -820,7 +820,7 @@ BaseType_t xQueuePeek( QueueHandle_t xQueue,
  * @return pdTRUE if an item was successfully received from the queue,
  * otherwise pdFALSE.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueuePeekFromISR xQueuePeekFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -829,7 +829,7 @@ BaseType_t xQueuePeekFromISR( QueueHandle_t xQueue,
                               void * const pvBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueReceive(
@@ -917,7 +917,7 @@ BaseType_t xQueuePeekFromISR( QueueHandle_t xQueue,
  *  // ... Rest of task code.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueReceive xQueueReceive
  * @endcond
  * \ingroup QueueManagement
@@ -927,7 +927,7 @@ BaseType_t xQueueReceive( QueueHandle_t xQueue,
                           TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * UBaseType_t uxQueueMessagesWaiting( const QueueHandle_t xQueue );
@@ -940,7 +940,7 @@ BaseType_t xQueueReceive( QueueHandle_t xQueue,
  *
  * @return The number of messages available in the queue.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup uxQueueMessagesWaiting uxQueueMessagesWaiting
  * @endcond
  * \ingroup QueueManagement
@@ -948,7 +948,7 @@ BaseType_t xQueueReceive( QueueHandle_t xQueue,
 UBaseType_t uxQueueMessagesWaiting( const QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * UBaseType_t uxQueueSpacesAvailable( const QueueHandle_t xQueue );
@@ -963,7 +963,7 @@ UBaseType_t uxQueueMessagesWaiting( const QueueHandle_t xQueue ) PRIVILEGED_FUNC
  *
  * @return The number of spaces available in the queue.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup uxQueueMessagesWaiting uxQueueMessagesWaiting
  * @endcond
  * \ingroup QueueManagement
@@ -971,7 +971,7 @@ UBaseType_t uxQueueMessagesWaiting( const QueueHandle_t xQueue ) PRIVILEGED_FUNC
 UBaseType_t uxQueueSpacesAvailable( const QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * void vQueueDelete( QueueHandle_t xQueue );
@@ -983,7 +983,7 @@ UBaseType_t uxQueueSpacesAvailable( const QueueHandle_t xQueue ) PRIVILEGED_FUNC
  *
  * @param xQueue A handle to the queue to be deleted.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vQueueDelete vQueueDelete
  * @endcond
  * \ingroup QueueManagement
@@ -991,7 +991,7 @@ UBaseType_t uxQueueSpacesAvailable( const QueueHandle_t xQueue ) PRIVILEGED_FUNC
 void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendToFrontFromISR(
@@ -1057,7 +1057,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSendFromISR xQueueSendFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1067,7 +1067,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendToBackFromISR(
@@ -1133,7 +1133,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSendFromISR xQueueSendFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1142,7 +1142,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
     xQueueGenericSendFromISR( ( xQueue ), ( pvItemToQueue ), ( pxHigherPriorityTaskWoken ), queueSEND_TO_BACK )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueOverwriteFromISR(
@@ -1225,7 +1225,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueOverwriteFromISR xQueueOverwriteFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1234,7 +1234,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
     xQueueGenericSendFromISR( ( xQueue ), ( pvItemToQueue ), ( pxHigherPriorityTaskWoken ), queueOVERWRITE )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueSendFromISR(
@@ -1304,7 +1304,7 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueSendFromISR xQueueSendFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1312,10 +1312,10 @@ void vQueueDelete( QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 #define xQueueSendFromISR( xQueue, pvItemToQueue, pxHigherPriorityTaskWoken ) \
     xQueueGenericSendFromISR( ( xQueue ), ( pvItemToQueue ), ( pxHigherPriorityTaskWoken ), queueSEND_TO_BACK )
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /**@{*/
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueGenericSendFromISR(
@@ -1402,7 +1402,7 @@ BaseType_t xQueueGiveFromISR( QueueHandle_t xQueue,
 /** @endcond */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * queue. h
  * @code{c}
  * BaseType_t xQueueReceiveFromISR(
@@ -1487,7 +1487,7 @@ BaseType_t xQueueGiveFromISR( QueueHandle_t xQueue,
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xQueueReceiveFromISR xQueueReceiveFromISR
  * @endcond
  * \ingroup QueueManagement
@@ -1504,7 +1504,7 @@ BaseType_t xQueueIsQueueEmptyFromISR( const QueueHandle_t xQueue ) PRIVILEGED_FU
 BaseType_t xQueueIsQueueFullFromISR( const QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 UBaseType_t uxQueueMessagesWaitingFromISR( const QueueHandle_t xQueue ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /*
  * The functions defined above are for passing data to and from tasks.  The
  * functions below are the equivalents for passing data to and from
@@ -1778,7 +1778,7 @@ QueueSetMemberHandle_t xQueueSelectFromSet( QueueSetHandle_t xQueueSet,
  */
 QueueSetMemberHandle_t xQueueSelectFromSetFromISR( QueueSetHandle_t xQueueSet ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 
 /* Not public API functions. */
 void vQueueWaitForMessageRestricted( QueueHandle_t xQueue,
diff --git a/tools/sdk/esp32s2/include/freertos/include/freertos/semphr.h b/tools/sdk/esp32s2/include/freertos/include/freertos/semphr.h
index 7e99c0b396c..2041641b91d 100644
--- a/tools/sdk/esp32s2/include/freertos/include/freertos/semphr.h
+++ b/tools/sdk/esp32s2/include/freertos/include/freertos/semphr.h
@@ -39,7 +39,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define semSEMAPHORE_QUEUE_ITEM_LENGTH      ( ( uint8_t ) 0U )
 #define semGIVE_BLOCK_TIME                  ( ( TickType_t ) 0U )
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /**
  * semphr. h
  * @code{c}
@@ -88,7 +88,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vSemaphoreCreateBinary vSemaphoreCreateBinary
  * @endcond
  * \ingroup Semaphores
@@ -106,7 +106,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 /** @endcond */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateBinary( void );
@@ -163,7 +163,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateBinary xSemaphoreCreateBinary
  * @endcond
  * \ingroup Semaphores
@@ -173,7 +173,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateBinaryStatic( StaticSemaphore_t *pxSemaphoreBuffer );
@@ -229,7 +229,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  // Rest of task code goes here.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateBinaryStatic xSemaphoreCreateBinaryStatic
  * @endcond
  * \ingroup Semaphores
@@ -239,7 +239,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * xSemaphoreTake(
@@ -304,7 +304,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreTake xSemaphoreTake
  * @endcond
  * \ingroup Semaphores
@@ -312,7 +312,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreTake( xSemaphore, xBlockTime )    xQueueSemaphoreTake( ( xSemaphore ), ( xBlockTime ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * xSemaphoreTakeRecursive(
@@ -403,7 +403,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreTakeRecursive xSemaphoreTakeRecursive
  * @endcond
  * \ingroup Semaphores
@@ -465,7 +465,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreGive xSemaphoreGive
  * @endcond
  * \ingroup Semaphores
@@ -473,7 +473,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreGive( xSemaphore )    xQueueGenericSend( ( QueueHandle_t ) ( xSemaphore ), NULL, semGIVE_BLOCK_TIME, queueSEND_TO_BACK )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * xSemaphoreGiveRecursive( SemaphoreHandle_t xMutex );
@@ -555,7 +555,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreGiveRecursive xSemaphoreGiveRecursive
  * @endcond
  * \ingroup Semaphores
@@ -641,7 +641,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreGiveFromISR xSemaphoreGiveFromISR
  * @endcond
  * \ingroup Semaphores
@@ -649,7 +649,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreGiveFromISR( xSemaphore, pxHigherPriorityTaskWoken )    xQueueGiveFromISR( ( QueueHandle_t ) ( xSemaphore ), ( pxHigherPriorityTaskWoken ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * xSemaphoreTakeFromISR(
@@ -686,7 +686,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreTakeFromISR( xSemaphore, pxHigherPriorityTaskWoken )    xQueueReceiveFromISR( ( QueueHandle_t ) ( xSemaphore ), NULL, ( pxHigherPriorityTaskWoken ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateMutex( void );
@@ -741,7 +741,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateMutex xSemaphoreCreateMutex
  * @endcond
  * \ingroup Semaphores
@@ -751,7 +751,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateMutexStatic( StaticSemaphore_t *pxMutexBuffer );
@@ -808,7 +808,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  // so there is no need to check it.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateMutexStatic xSemaphoreCreateMutexStatic
  * @endcond
  * \ingroup Semaphores
@@ -951,7 +951,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateCounting( UBaseType_t uxMaxCount, UBaseType_t uxInitialCount );
@@ -1027,7 +1027,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateCounting xSemaphoreCreateCounting
  * @endcond
  * \ingroup Semaphores
@@ -1037,7 +1037,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * SemaphoreHandle_t xSemaphoreCreateCountingStatic( UBaseType_t uxMaxCount, UBaseType_t uxInitialCount, StaticSemaphore_t *pxSemaphoreBuffer );
@@ -1118,7 +1118,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *  // is no need to check its value.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xSemaphoreCreateCountingStatic xSemaphoreCreateCountingStatic
  * @endcond
  * \ingroup Semaphores
@@ -1128,7 +1128,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr. h
  * @code{c}
  * void vSemaphoreDelete( SemaphoreHandle_t xSemaphore );
@@ -1140,7 +1140,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
  *
  * @param xSemaphore A handle to the semaphore to be deleted.
  *
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * \defgroup vSemaphoreDelete vSemaphoreDelete
  * @endcond
  * \ingroup Semaphores
@@ -1148,7 +1148,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define vSemaphoreDelete( xSemaphore )                   vQueueDelete( ( QueueHandle_t ) ( xSemaphore ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr.h
  * @code{c}
  * TaskHandle_t xSemaphoreGetMutexHolder( SemaphoreHandle_t xMutex );
@@ -1167,7 +1167,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreGetMutexHolder( xSemaphore )           xQueueGetMutexHolder( ( xSemaphore ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr.h
  * @code{c}
  * TaskHandle_t xSemaphoreGetMutexHolderFromISR( SemaphoreHandle_t xMutex );
@@ -1182,7 +1182,7 @@ typedef QueueHandle_t SemaphoreHandle_t;
 #define xSemaphoreGetMutexHolderFromISR( xSemaphore )    xQueueGetMutexHolderFromISR( ( xSemaphore ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * semphr.h
  * @code{c}
  * UBaseType_t uxSemaphoreGetCount( SemaphoreHandle_t xSemaphore );
diff --git a/tools/sdk/esp32s2/include/freertos/include/freertos/stream_buffer.h b/tools/sdk/esp32s2/include/freertos/include/freertos/stream_buffer.h
index 9e58cff120c..a20dcf0375e 100644
--- a/tools/sdk/esp32s2/include/freertos/include/freertos/stream_buffer.h
+++ b/tools/sdk/esp32s2/include/freertos/include/freertos/stream_buffer.h
@@ -71,7 +71,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * message_buffer.h
  *
  * @code{c}
@@ -134,7 +134,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferCreate xStreamBufferCreate
  * @endcond
  * \ingroup StreamBufferManagement
@@ -142,7 +142,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
 #define xStreamBufferCreate( xBufferSizeBytes, xTriggerLevelBytes )    xStreamBufferGenericCreate( xBufferSizeBytes, xTriggerLevelBytes, pdFALSE )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -220,7 +220,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
  * }
  *
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferCreateStatic xStreamBufferCreateStatic
  * @endcond
  * \ingroup StreamBufferManagement
@@ -229,7 +229,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
     xStreamBufferGenericCreateStatic( xBufferSizeBytes, xTriggerLevelBytes, pdFALSE, pucStreamBufferStorageArea, pxStaticStreamBuffer )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -319,7 +319,7 @@ typedef struct StreamBufferDef_t * StreamBufferHandle_t;
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSend xStreamBufferSend
  * @endcond
  * \ingroup StreamBufferManagement
@@ -330,7 +330,7 @@ size_t xStreamBufferSend( StreamBufferHandle_t xStreamBuffer,
                           TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -424,7 +424,7 @@ size_t xStreamBufferSend( StreamBufferHandle_t xStreamBuffer,
  *  taskYIELD_FROM_ISR( xHigherPriorityTaskWoken );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSendFromISR xStreamBufferSendFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -435,7 +435,7 @@ size_t xStreamBufferSendFromISR( StreamBufferHandle_t xStreamBuffer,
                                  BaseType_t * const pxHigherPriorityTaskWoken ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -517,7 +517,7 @@ size_t xStreamBufferSendFromISR( StreamBufferHandle_t xStreamBuffer,
  *  }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferReceive xStreamBufferReceive
  * @endcond
  * \ingroup StreamBufferManagement
@@ -528,7 +528,7 @@ size_t xStreamBufferReceive( StreamBufferHandle_t xStreamBuffer,
                              TickType_t xTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -607,7 +607,7 @@ size_t xStreamBufferReceive( StreamBufferHandle_t xStreamBuffer,
  *  taskYIELD_FROM_ISR( xHigherPriorityTaskWoken );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferReceiveFromISR xStreamBufferReceiveFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -618,7 +618,7 @@ size_t xStreamBufferReceiveFromISR( StreamBufferHandle_t xStreamBuffer,
                                     BaseType_t * const pxHigherPriorityTaskWoken ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -636,7 +636,7 @@ size_t xStreamBufferReceiveFromISR( StreamBufferHandle_t xStreamBuffer,
  *
  * @param xStreamBuffer The handle of the stream buffer to be deleted.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vStreamBufferDelete vStreamBufferDelete
  * @endcond
  * \ingroup StreamBufferManagement
@@ -644,7 +644,7 @@ size_t xStreamBufferReceiveFromISR( StreamBufferHandle_t xStreamBuffer,
 void vStreamBufferDelete( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -660,7 +660,7 @@ void vStreamBufferDelete( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTI
  * @return If the stream buffer is full then pdTRUE is returned.  Otherwise
  * pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferIsFull xStreamBufferIsFull
  * @endcond
  * \ingroup StreamBufferManagement
@@ -668,7 +668,7 @@ void vStreamBufferDelete( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTI
 BaseType_t xStreamBufferIsFull( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -684,7 +684,7 @@ BaseType_t xStreamBufferIsFull( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_
  * @return If the stream buffer is empty then pdTRUE is returned.  Otherwise
  * pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferIsEmpty xStreamBufferIsEmpty
  * @endcond
  * \ingroup StreamBufferManagement
@@ -692,7 +692,7 @@ BaseType_t xStreamBufferIsFull( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_
 BaseType_t xStreamBufferIsEmpty( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -711,7 +711,7 @@ BaseType_t xStreamBufferIsEmpty( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED
  * a task blocked waiting to send to or read from the stream buffer then the
  * stream buffer is not reset and pdFAIL is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferReset xStreamBufferReset
  * @endcond
  * \ingroup StreamBufferManagement
@@ -719,7 +719,7 @@ BaseType_t xStreamBufferIsEmpty( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED
 BaseType_t xStreamBufferReset( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -736,7 +736,7 @@ BaseType_t xStreamBufferReset( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_F
  * @return The number of bytes that can be written to the stream buffer before
  * the stream buffer would be full.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSpacesAvailable xStreamBufferSpacesAvailable
  * @endcond
  * \ingroup StreamBufferManagement
@@ -744,7 +744,7 @@ BaseType_t xStreamBufferReset( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_F
 size_t xStreamBufferSpacesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -761,7 +761,7 @@ size_t xStreamBufferSpacesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVIL
  * @return The number of bytes that can be read from the stream buffer before
  * the stream buffer would be empty.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferBytesAvailable xStreamBufferBytesAvailable
  * @endcond
  * \ingroup StreamBufferManagement
@@ -769,7 +769,7 @@ size_t xStreamBufferSpacesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVIL
 size_t xStreamBufferBytesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -802,7 +802,7 @@ size_t xStreamBufferBytesAvailable( StreamBufferHandle_t xStreamBuffer ) PRIVILE
  * then the trigger level will be updated and pdTRUE is returned.  Otherwise
  * pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSetTriggerLevel xStreamBufferSetTriggerLevel
  * @endcond
  * \ingroup StreamBufferManagement
@@ -811,7 +811,7 @@ BaseType_t xStreamBufferSetTriggerLevel( StreamBufferHandle_t xStreamBuffer,
                                          size_t xTriggerLevel ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -846,7 +846,7 @@ BaseType_t xStreamBufferSetTriggerLevel( StreamBufferHandle_t xStreamBuffer,
  * @return If a task was removed from the Blocked state then pdTRUE is returned.
  * Otherwise pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferSendCompletedFromISR xStreamBufferSendCompletedFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -855,7 +855,7 @@ BaseType_t xStreamBufferSendCompletedFromISR( StreamBufferHandle_t xStreamBuffer
                                               BaseType_t * pxHigherPriorityTaskWoken ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * stream_buffer.h
  *
  * @code{c}
@@ -891,7 +891,7 @@ BaseType_t xStreamBufferSendCompletedFromISR( StreamBufferHandle_t xStreamBuffer
  * @return If a task was removed from the Blocked state then pdTRUE is returned.
  * Otherwise pdFALSE is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xStreamBufferReceiveCompletedFromISR xStreamBufferReceiveCompletedFromISR
  * @endcond
  * \ingroup StreamBufferManagement
@@ -899,7 +899,7 @@ BaseType_t xStreamBufferSendCompletedFromISR( StreamBufferHandle_t xStreamBuffer
 BaseType_t xStreamBufferReceiveCompletedFromISR( StreamBufferHandle_t xStreamBuffer,
                                                  BaseType_t * pxHigherPriorityTaskWoken ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /* Functions below here are not part of the public API. */
 StreamBufferHandle_t xStreamBufferGenericCreate( size_t xBufferSizeBytes,
                                                  size_t xTriggerLevelBytes,
diff --git a/tools/sdk/esp32s2/include/freertos/include/freertos/task.h b/tools/sdk/esp32s2/include/freertos/include/freertos/task.h
index 9135b76f014..125a924d061 100644
--- a/tools/sdk/esp32s2/include/freertos/include/freertos/task.h
+++ b/tools/sdk/esp32s2/include/freertos/include/freertos/task.h
@@ -76,7 +76,7 @@
  * returns (via a pointer parameter) an TaskHandle_t variable that can then
  * be used as a parameter to vTaskDelete to delete the task.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup TaskHandle_t TaskHandle_t
  * @endcond
  * \ingroup Tasks
@@ -114,7 +114,7 @@ typedef enum
     eSetValueWithoutOverwrite     /* Set the task's notification value if the previous value has been read by the task. */
 } eNotifyAction;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /**
  * Used internally only.
  */
@@ -189,11 +189,13 @@ typedef enum
 #define tskIDLE_PRIORITY    ( ( UBaseType_t ) 0U )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro for forcing a context switch.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskYIELD taskYIELD
  * @endcond
  * \ingroup SchedulerControl
@@ -201,7 +203,9 @@ typedef enum
 #define taskYIELD()                        portYIELD()
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro to mark the start of a critical code region.  Preemptive context
  * switches cannot occur when in a critical region.
@@ -209,7 +213,7 @@ typedef enum
  * @note This may alter the stack (depending on the portable implementation)
  * so must be used with care!
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskENTER_CRITICAL taskENTER_CRITICAL
  * @endcond
  * \ingroup SchedulerControl
@@ -228,7 +232,9 @@ typedef enum
 #endif //  ESP_PLATFORM
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro to mark the end of a critical code region.  Preemptive context
  * switches cannot occur when in a critical region.
@@ -236,7 +242,7 @@ typedef enum
  * @note This may alter the stack (depending on the portable implementation)
  * so must be used with care!
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskEXIT_CRITICAL taskEXIT_CRITICAL
  * @endcond
  * \ingroup SchedulerControl
@@ -255,11 +261,13 @@ typedef enum
 #define taskEXIT_CRITICAL_ISR( )        portEXIT_CRITICAL_ISR( )
 #endif // ESP_PLATFORM
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro to disable all maskable interrupts.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskDISABLE_INTERRUPTS taskDISABLE_INTERRUPTS
  * @endcond
  * \ingroup SchedulerControl
@@ -267,11 +275,13 @@ typedef enum
 #define taskDISABLE_INTERRUPTS()           portDISABLE_INTERRUPTS()
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
+ * @endcond
  *
  * Macro to enable microcontroller interrupts.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup taskENABLE_INTERRUPTS taskENABLE_INTERRUPTS
  * @endcond
  * \ingroup SchedulerControl
@@ -422,7 +432,7 @@ typedef enum
  *   }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCreate xTaskCreate
  * @endcond
  * \ingroup Tasks
@@ -430,14 +440,14 @@ typedef enum
 #if ( configSUPPORT_DYNAMIC_ALLOCATION == 1 )
 
     static inline IRAM_ATTR BaseType_t xTaskCreate(
-            TaskFunction_t pvTaskCode,
-            const char * const pcName,
-            const uint32_t usStackDepth,
-            void * const pvParameters,
-            UBaseType_t uxPriority,
-            TaskHandle_t * const pvCreatedTask)
+                            TaskFunction_t pvTaskCode,
+                            const char * const pcName,     /*lint !e971 Unqualified char types are allowed for strings and single characters only. */
+                            const uint32_t usStackDepth,
+                            void * const pvParameters,
+                            UBaseType_t uxPriority,
+                            TaskHandle_t * const pxCreatedTask) PRIVILEGED_FUNCTION
     {
-        return xTaskCreatePinnedToCore( pvTaskCode, pcName, usStackDepth, pvParameters, uxPriority, pvCreatedTask, tskNO_AFFINITY );
+        return xTaskCreatePinnedToCore( pvTaskCode, pcName, usStackDepth, pvParameters, uxPriority, pxCreatedTask, tskNO_AFFINITY );
     }
 
 #endif
@@ -599,20 +609,20 @@ typedef enum
 
 #if( configSUPPORT_STATIC_ALLOCATION == 1 )
     static inline IRAM_ATTR TaskHandle_t xTaskCreateStatic(
-            TaskFunction_t pvTaskCode,
-            const char * const pcName,
-            const uint32_t ulStackDepth,
-            void * const pvParameters,
-            UBaseType_t uxPriority,
-            StackType_t * const pxStackBuffer,
-            StaticTask_t * const pxTaskBuffer)
+                                    TaskFunction_t pvTaskCode,
+                                    const char * const pcName,     /*lint !e971 Unqualified char types are allowed for strings and single characters only. */
+                                    const uint32_t ulStackDepth,
+                                    void * const pvParameters,
+                                    UBaseType_t uxPriority,
+                                    StackType_t * const puxStackBuffer,
+                                    StaticTask_t * const pxTaskBuffer) PRIVILEGED_FUNCTION
     {
-        return xTaskCreateStaticPinnedToCore( pvTaskCode, pcName, ulStackDepth, pvParameters, uxPriority, pxStackBuffer, pxTaskBuffer, tskNO_AFFINITY );
+        return xTaskCreateStaticPinnedToCore( pvTaskCode, pcName, ulStackDepth, pvParameters, uxPriority, puxStackBuffer, pxTaskBuffer, tskNO_AFFINITY );
     }
 #endif /* configSUPPORT_STATIC_ALLOCATION */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskCreateRestricted( TaskParameters_t *pxTaskDefinition, TaskHandle_t *pxCreatedTask );
@@ -683,18 +693,18 @@ typedef enum
  *  for( ;; );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCreateRestricted xTaskCreateRestricted
  * @endcond
  * \ingroup Tasks
  */
 #if ( portUSING_MPU_WRAPPERS == 1 )
     BaseType_t xTaskCreateRestricted( const TaskParameters_t * const pxTaskDefinition,
-                                      TaskHandle_t * pxCreatedTask );
+                                      TaskHandle_t * pxCreatedTask ) PRIVILEGED_FUNCTION;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskCreateRestrictedStatic( TaskParameters_t *pxTaskDefinition, TaskHandle_t *pxCreatedTask );
@@ -777,7 +787,7 @@ typedef enum
  *  for( ;; );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCreateRestrictedStatic xTaskCreateRestrictedStatic
  * @endcond
  * \ingroup Tasks
@@ -788,7 +798,7 @@ typedef enum
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskAllocateMPURegions( TaskHandle_t xTask, const MemoryRegion_t * const pxRegions );
@@ -833,7 +843,7 @@ typedef enum
  *  // defined or shared regions have been declared elsewhere).
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCreateRestricted xTaskCreateRestricted
  * @endcond
  * \ingroup Tasks
@@ -842,7 +852,7 @@ void vTaskAllocateMPURegions( TaskHandle_t xTask,
                               const MemoryRegion_t * const pxRegions ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskDelete( TaskHandle_t xTask );
@@ -881,7 +891,7 @@ void vTaskAllocateMPURegions( TaskHandle_t xTask,
  *   vTaskDelete( xHandle );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskDelete vTaskDelete
  * @endcond
  * \ingroup Tasks
@@ -893,10 +903,12 @@ void vTaskDelete( TaskHandle_t xTaskToDelete ) PRIVILEGED_FUNCTION;
  *----------------------------------------------------------*/
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskDelay( const TickType_t xTicksToDelay );
  * @endcode
+ * @endcond
  *
  * Delay a task for a given number of ticks.  The actual time that the
  * task remains blocked depends on the tick rate.  The constant
@@ -938,7 +950,7 @@ void vTaskDelete( TaskHandle_t xTaskToDelete ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskDelay vTaskDelay
  * @endcond
  * \ingroup TaskCtrl
@@ -946,10 +958,12 @@ void vTaskDelete( TaskHandle_t xTaskToDelete ) PRIVILEGED_FUNCTION;
 void vTaskDelay( const TickType_t xTicksToDelay ) PRIVILEGED_FUNCTION;
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskDelayUntil( TickType_t *pxPreviousWakeTime, const TickType_t xTimeIncrement );
  * @endcode
+ * @endcond
  *
  * INCLUDE_xTaskDelayUntil must be defined as 1 for this function to be available.
  * See the configuration section for more information.
@@ -1007,7 +1021,7 @@ void vTaskDelay( const TickType_t xTicksToDelay ) PRIVILEGED_FUNCTION;
  *     }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskDelayUntil xTaskDelayUntil
  * @endcond
  * \ingroup TaskCtrl
@@ -1026,7 +1040,7 @@ BaseType_t xTaskDelayUntil( TickType_t * const pxPreviousWakeTime,
 
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskAbortDelay( TaskHandle_t xTask );
@@ -1054,7 +1068,7 @@ BaseType_t xTaskDelayUntil( TickType_t * const pxPreviousWakeTime,
  * @return If the task referenced by xTask was not in the Blocked state then
  * pdFAIL is returned.  Otherwise pdPASS is returned.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskAbortDelay xTaskAbortDelay
  * @endcond
  * \ingroup TaskCtrl
@@ -1062,7 +1076,7 @@ BaseType_t xTaskDelayUntil( TickType_t * const pxPreviousWakeTime,
 BaseType_t xTaskAbortDelay( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * UBaseType_t uxTaskPriorityGet( const TaskHandle_t xTask );
@@ -1107,7 +1121,7 @@ BaseType_t xTaskAbortDelay( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
  *   }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup uxTaskPriorityGet uxTaskPriorityGet
  * @endcond
  * \ingroup TaskCtrl
@@ -1115,7 +1129,7 @@ BaseType_t xTaskAbortDelay( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 UBaseType_t uxTaskPriorityGet( const TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * UBaseType_t uxTaskPriorityGetFromISR( const TaskHandle_t xTask );
@@ -1127,7 +1141,7 @@ UBaseType_t uxTaskPriorityGet( const TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 UBaseType_t uxTaskPriorityGetFromISR( const TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * eTaskState eTaskGetState( TaskHandle_t xTask );
@@ -1149,7 +1163,7 @@ UBaseType_t uxTaskPriorityGetFromISR( const TaskHandle_t xTask ) PRIVILEGED_FUNC
 eTaskState eTaskGetState( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskGetInfo( TaskHandle_t xTask, TaskStatus_t *pxTaskStatus, BaseType_t xGetFreeStackSpace, eTaskState eState );
@@ -1203,7 +1217,7 @@ eTaskState eTaskGetState( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
  *                eInvalid ); // Include the task state in xTaskDetails.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskGetInfo vTaskGetInfo
  * @endcond
  * \ingroup TaskCtrl
@@ -1214,7 +1228,7 @@ void vTaskGetInfo( TaskHandle_t xTask,
                    eTaskState eState ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskPrioritySet( TaskHandle_t xTask, UBaseType_t uxNewPriority );
@@ -1254,7 +1268,7 @@ void vTaskGetInfo( TaskHandle_t xTask,
  *   vTaskPrioritySet( NULL, tskIDLE_PRIORITY + 1 );
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskPrioritySet vTaskPrioritySet
  * @endcond
  * \ingroup TaskCtrl
@@ -1263,7 +1277,7 @@ void vTaskPrioritySet( TaskHandle_t xTask,
                        UBaseType_t uxNewPriority ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskSuspend( TaskHandle_t xTaskToSuspend );
@@ -1312,7 +1326,7 @@ void vTaskPrioritySet( TaskHandle_t xTask,
  *   // with our handle as the parameter.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskSuspend vTaskSuspend
  * @endcond
  * \ingroup TaskCtrl
@@ -1320,7 +1334,7 @@ void vTaskPrioritySet( TaskHandle_t xTask,
 void vTaskSuspend( TaskHandle_t xTaskToSuspend ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskResume( TaskHandle_t xTaskToResume );
@@ -1367,7 +1381,7 @@ void vTaskSuspend( TaskHandle_t xTaskToSuspend ) PRIVILEGED_FUNCTION;
  *   // time in accordance with its priority within the system.
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskResume vTaskResume
  * @endcond
  * \ingroup TaskCtrl
@@ -1375,7 +1389,7 @@ void vTaskSuspend( TaskHandle_t xTaskToSuspend ) PRIVILEGED_FUNCTION;
 void vTaskResume( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void xTaskResumeFromISR( TaskHandle_t xTaskToResume );
@@ -1402,7 +1416,7 @@ void vTaskResume( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
  * otherwise pdFALSE. This is used by the ISR to determine if a context switch
  * may be required following the ISR.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskResumeFromISR vTaskResumeFromISR
  * @endcond
  * \ingroup TaskCtrl
@@ -1412,9 +1426,9 @@ BaseType_t xTaskResumeFromISR( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
 /*-----------------------------------------------------------
  * SCHEDULER CONTROL
  *----------------------------------------------------------*/
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskStartScheduler( void );
@@ -1445,7 +1459,7 @@ BaseType_t xTaskResumeFromISR( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskStartScheduler vTaskStartScheduler
  * @endcond
  * \ingroup SchedulerControl
@@ -1453,7 +1467,7 @@ BaseType_t xTaskResumeFromISR( TaskHandle_t xTaskToResume ) PRIVILEGED_FUNCTION;
 void vTaskStartScheduler( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskEndScheduler( void );
@@ -1507,7 +1521,7 @@ void vTaskStartScheduler( void ) PRIVILEGED_FUNCTION;
  * }
  * @endcode
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskEndScheduler vTaskEndScheduler
  * @endcond
  * \ingroup SchedulerControl
@@ -1517,7 +1531,7 @@ void vTaskEndScheduler( void ) PRIVILEGED_FUNCTION;
 /** @endcond */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskSuspendAll( void );
@@ -1566,7 +1580,7 @@ void vTaskEndScheduler( void ) PRIVILEGED_FUNCTION;
  *   }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskSuspendAll vTaskSuspendAll
  * @endcond
  * \ingroup SchedulerControl
@@ -1574,7 +1588,7 @@ void vTaskEndScheduler( void ) PRIVILEGED_FUNCTION;
 void vTaskSuspendAll( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskResumeAll( void );
@@ -1626,7 +1640,7 @@ void vTaskSuspendAll( void ) PRIVILEGED_FUNCTION;
  *   }
  * }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskResumeAll xTaskResumeAll
  * @endcond
  * \ingroup SchedulerControl
@@ -1638,7 +1652,7 @@ BaseType_t xTaskResumeAll( void ) PRIVILEGED_FUNCTION;
  *----------------------------------------------------------*/
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * TickType_t xTaskGetTickCount( void );
@@ -1647,7 +1661,7 @@ BaseType_t xTaskResumeAll( void ) PRIVILEGED_FUNCTION;
  *
  * @return The count of ticks since vTaskStartScheduler was called.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskGetTickCount xTaskGetTickCount
  * @endcond
  * \ingroup TaskUtils
@@ -1655,7 +1669,7 @@ BaseType_t xTaskResumeAll( void ) PRIVILEGED_FUNCTION;
 TickType_t xTaskGetTickCount( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * TickType_t xTaskGetTickCountFromISR( void );
@@ -1669,7 +1683,7 @@ TickType_t xTaskGetTickCount( void ) PRIVILEGED_FUNCTION;
  * microcontroller being used or interrupt nesting is either not supported or
  * not being used.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskGetTickCountFromISR xTaskGetTickCountFromISR
  * @endcond
  * \ingroup TaskUtils
@@ -1677,7 +1691,7 @@ TickType_t xTaskGetTickCount( void ) PRIVILEGED_FUNCTION;
 TickType_t xTaskGetTickCountFromISR( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * uint16_t uxTaskGetNumberOfTasks( void );
@@ -1689,7 +1703,7 @@ TickType_t xTaskGetTickCountFromISR( void ) PRIVILEGED_FUNCTION;
  * has been deleted but not yet freed by the idle task will also be
  * included in the count.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup uxTaskGetNumberOfTasks uxTaskGetNumberOfTasks
  * @endcond
  * \ingroup TaskUtils
@@ -1697,7 +1711,7 @@ TickType_t xTaskGetTickCountFromISR( void ) PRIVILEGED_FUNCTION;
 UBaseType_t uxTaskGetNumberOfTasks( void ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * char *pcTaskGetName( TaskHandle_t xTaskToQuery );
@@ -1708,7 +1722,7 @@ UBaseType_t uxTaskGetNumberOfTasks( void ) PRIVILEGED_FUNCTION;
  * xTaskToQuery.  A task can query its own name by either passing in its own
  * handle, or by setting xTaskToQuery to NULL.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup pcTaskGetName pcTaskGetName
  * @endcond
  * \ingroup TaskUtils
@@ -1716,7 +1730,7 @@ UBaseType_t uxTaskGetNumberOfTasks( void ) PRIVILEGED_FUNCTION;
 char * pcTaskGetName( TaskHandle_t xTaskToQuery ) PRIVILEGED_FUNCTION;     /*lint !e971 Unqualified char types are allowed for strings and single characters only. */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * TaskHandle_t xTaskGetHandle( const char *pcNameToQuery );
@@ -1730,7 +1744,7 @@ char * pcTaskGetName( TaskHandle_t xTaskToQuery ) PRIVILEGED_FUNCTION;     /*lin
  * NULL is returned if no matching name is found.  INCLUDE_xTaskGetHandle
  * must be set to 1 in FreeRTOSConfig.h for pcTaskGetHandle() to be available.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup pcTaskGetHandle pcTaskGetHandle
  * @endcond
  * \ingroup TaskUtils
@@ -1813,7 +1827,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 #ifdef configUSE_APPLICATION_TASK_TAG
     #if configUSE_APPLICATION_TASK_TAG == 1
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * void vTaskSetApplicationTaskTag( TaskHandle_t xTask, TaskHookFunction_t pxHookFunction );
@@ -1830,7 +1844,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
                                          TaskHookFunction_t pxHookFunction ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * void xTaskGetApplicationTaskTag( TaskHandle_t xTask );
@@ -1844,7 +1858,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
         TaskHookFunction_t xTaskGetApplicationTaskTag( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * void xTaskGetApplicationTaskTagFromISR( TaskHandle_t xTask );
@@ -1932,7 +1946,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 #if ( configCHECK_FOR_STACK_OVERFLOW > 0 )
 
      /**
-      * @cond
+      * @cond !DOC_EXCLUDE_HEADER_SECTION
       * task.h
       * @code{c}
       * void vApplicationStackOverflowHook( TaskHandle_t xTask char *pcTaskName);
@@ -1952,7 +1966,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 
 #if  (  configUSE_TICK_HOOK > 0 )
     /**
-     * @cond
+     * @cond !DOC_EXCLUDE_HEADER_SECTION
      *  task.h
      * @code{c}
      * void vApplicationTickHook( void );
@@ -1967,7 +1981,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 
 #if ( configSUPPORT_STATIC_ALLOCATION == 1 )
     /**
-     * @cond
+     * @cond !DOC_EXCLUDE_HEADER_SECTION
      * task.h
      * @code{c}
      * void vApplicationGetIdleTaskMemory( StaticTask_t ** ppxIdleTaskTCBBuffer, StackType_t ** ppxIdleTaskStackBuffer, uint32_t *pulIdleTaskStackSize )
@@ -1986,7 +2000,7 @@ uint8_t* pxTaskGetStackStart( TaskHandle_t xTask) PRIVILEGED_FUNCTION;
 #endif
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * BaseType_t xTaskCallApplicationTaskHook( TaskHandle_t xTask, void *pvParameter );
@@ -2155,7 +2169,7 @@ UBaseType_t uxTaskGetSystemState( TaskStatus_t * const pxTaskStatusArray,
  * enough to contain the generated report.  Approximately 40 bytes per
  * task should be sufficient.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskList vTaskList
  * @endcond
  * \ingroup TaskUtils
@@ -2210,7 +2224,7 @@ void vTaskList( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lint !e971 Unq
  * contain the generated report.  Approximately 40 bytes per task should
  * be sufficient.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskGetRunTimeStats vTaskGetRunTimeStats
  * @endcond
  * \ingroup TaskUtils
@@ -2218,7 +2232,7 @@ void vTaskList( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lint !e971 Unq
 void vTaskGetRunTimeStats( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lint !e971 Unqualified char types are allowed for strings and single characters only. */
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code
  * uint32_t ulTaskGetIdleRunTimeCounter( void );
@@ -2246,7 +2260,7 @@ void vTaskGetRunTimeStats( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lin
  * frequency configured using the portCONFIGURE_TIMER_FOR_RUN_TIME_STATS() and
  * portGET_RUN_TIME_COUNTER_VALUE() macros.
  *
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup ulTaskGetIdleRunTimeCounter ulTaskGetIdleRunTimeCounter
  * @endcond
  * \ingroup TaskUtils
@@ -2254,11 +2268,13 @@ void vTaskGetRunTimeStats( char * pcWriteBuffer ) PRIVILEGED_FUNCTION;     /*lin
 uint32_t ulTaskGetIdleRunTimeCounter( void ) PRIVILEGED_FUNCTION;
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyIndexed( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify, uint32_t ulValue, eNotifyAction eAction );
  * BaseType_t xTaskNotify( TaskHandle_t xTaskToNotify, uint32_t ulValue, eNotifyAction eAction );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2359,7 +2375,9 @@ uint32_t ulTaskGetIdleRunTimeCounter( void ) PRIVILEGED_FUNCTION;
  * @return Dependent on the value of eAction.  See the description of the
  * eAction parameter.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyIndexed xTaskNotifyIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
@@ -2373,11 +2391,13 @@ BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
     xTaskGenericNotify( ( xTaskToNotify ), ( uxIndexToNotify ), ( ulValue ), ( eAction ), NULL )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyAndQueryIndexed( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify, uint32_t ulValue, eNotifyAction eAction, uint32_t *pulPreviousNotifyValue );
  * BaseType_t xTaskNotifyAndQuery( TaskHandle_t xTaskToNotify, uint32_t ulValue, eNotifyAction eAction, uint32_t *pulPreviousNotifyValue );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2393,7 +2413,9 @@ BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
  * than when the function returns) in the additional pulPreviousNotifyValue
  * parameter.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyAndQueryIndexed xTaskNotifyAndQueryIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 #define xTaskNotifyAndQuery( xTaskToNotify, ulValue, eAction, pulPreviousNotifyValue ) \
@@ -2402,11 +2424,13 @@ BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
     xTaskGenericNotify( ( xTaskToNotify ), ( uxIndexToNotify ), ( ulValue ), ( eAction ), ( pulPreviousNotifyValue ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyIndexedFromISR( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify, uint32_t ulValue, eNotifyAction eAction, BaseType_t *pxHigherPriorityTaskWoken );
  * BaseType_t xTaskNotifyFromISR( TaskHandle_t xTaskToNotify, uint32_t ulValue, eNotifyAction eAction, BaseType_t *pxHigherPriorityTaskWoken );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2511,7 +2535,9 @@ BaseType_t xTaskGenericNotify( TaskHandle_t xTaskToNotify,
  * @return Dependent on the value of eAction.  See the description of the
  * eAction parameter.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyIndexedFromISR xTaskNotifyIndexedFromISR
+ * @endcond
  * \ingroup TaskNotifications
  */
 BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
@@ -2526,11 +2552,13 @@ BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
     xTaskGenericNotifyFromISR( ( xTaskToNotify ), ( uxIndexToNotify ), ( ulValue ), ( eAction ), NULL, ( pxHigherPriorityTaskWoken ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyAndQueryIndexedFromISR( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify, uint32_t ulValue, eNotifyAction eAction, uint32_t *pulPreviousNotificationValue, BaseType_t *pxHigherPriorityTaskWoken );
  * BaseType_t xTaskNotifyAndQueryFromISR( TaskHandle_t xTaskToNotify, uint32_t ulValue, eNotifyAction eAction, uint32_t *pulPreviousNotificationValue, BaseType_t *pxHigherPriorityTaskWoken );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2546,7 +2574,9 @@ BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
  * function is called rather than at the time the function returns) in the
  * additional pulPreviousNotifyValue parameter.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyAndQueryIndexedFromISR xTaskNotifyAndQueryIndexedFromISR
+ * @endcond
  * \ingroup TaskNotifications
  */
 #define xTaskNotifyAndQueryIndexedFromISR( xTaskToNotify, uxIndexToNotify, ulValue, eAction, pulPreviousNotificationValue, pxHigherPriorityTaskWoken ) \
@@ -2555,12 +2585,14 @@ BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
     xTaskGenericNotifyFromISR( ( xTaskToNotify ), ( tskDEFAULT_INDEX_TO_NOTIFY ), ( ulValue ), ( eAction ), ( pulPreviousNotificationValue ), ( pxHigherPriorityTaskWoken ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyWaitIndexed( UBaseType_t uxIndexToWaitOn, uint32_t ulBitsToClearOnEntry, uint32_t ulBitsToClearOnExit, uint32_t *pulNotificationValue, TickType_t xTicksToWait );
  *
  * BaseType_t xTaskNotifyWait( uint32_t ulBitsToClearOnEntry, uint32_t ulBitsToClearOnExit, uint32_t *pulNotificationValue, TickType_t xTicksToWait );
  * @endcode
+ * @endcond
  *
  * Waits for a direct to task notification to be pending at a given index within
  * an array of direct to task notifications.
@@ -2655,7 +2687,9 @@ BaseType_t xTaskGenericNotifyFromISR( TaskHandle_t xTaskToNotify,
  * already pending when xTaskNotifyWait was called) then pdPASS is
  * returned.  Otherwise pdFAIL is returned.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyWaitIndexed xTaskNotifyWaitIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
@@ -2669,11 +2703,13 @@ BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
     xTaskGenericNotifyWait( ( uxIndexToWaitOn ), ( ulBitsToClearOnEntry ), ( ulBitsToClearOnExit ), ( pulNotificationValue ), ( xTicksToWait ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyGiveIndexed( TaskHandle_t xTaskToNotify, UBaseType_t uxIndexToNotify );
  * BaseType_t xTaskNotifyGive( TaskHandle_t xTaskToNotify );
  * @endcode
+ * @endcond
  *
  * Sends a direct to task notification to a particular index in the target
  * task's notification array in a manner similar to giving a counting semaphore.
@@ -2737,7 +2773,9 @@ BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
  * @return xTaskNotifyGive() is a macro that calls xTaskNotify() with the
  * eAction parameter set to eIncrement - so pdPASS is always returned.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyGiveIndexed xTaskNotifyGiveIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 #define xTaskNotifyGive( xTaskToNotify ) \
@@ -2746,11 +2784,13 @@ BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
     xTaskGenericNotify( ( xTaskToNotify ), ( uxIndexToNotify ), ( 0 ), eIncrement, NULL )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * void vTaskNotifyGiveIndexedFromISR( TaskHandle_t xTaskHandle, UBaseType_t uxIndexToNotify, BaseType_t *pxHigherPriorityTaskWoken );
  * void vTaskNotifyGiveFromISR( TaskHandle_t xTaskHandle, BaseType_t *pxHigherPriorityTaskWoken );
  * @endcode
+ * @endcond
  *
  * A version of xTaskNotifyGiveIndexed() that can be called from an interrupt
  * service routine (ISR).
@@ -2821,7 +2861,9 @@ BaseType_t xTaskGenericNotifyWait( UBaseType_t uxIndexToWaitOn,
  * requested from an ISR is dependent on the port - see the documentation page
  * for the port in use.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup vTaskNotifyGiveIndexedFromISR vTaskNotifyGiveIndexedFromISR
+ * @endcond
  * \ingroup TaskNotifications
  */
 void vTaskGenericNotifyGiveFromISR( TaskHandle_t xTaskToNotify,
@@ -2833,12 +2875,14 @@ void vTaskGenericNotifyGiveFromISR( TaskHandle_t xTaskToNotify,
     vTaskGenericNotifyGiveFromISR( ( xTaskToNotify ), ( uxIndexToNotify ), ( pxHigherPriorityTaskWoken ) );
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * uint32_t ulTaskNotifyTakeIndexed( UBaseType_t uxIndexToWaitOn, BaseType_t xClearCountOnExit, TickType_t xTicksToWait );
  *
  * uint32_t ulTaskNotifyTake( BaseType_t xClearCountOnExit, TickType_t xTicksToWait );
  * @endcode
+ * @endcond
  *
  * Waits for a direct to task notification on a particular index in the calling
  * task's notification array in a manner similar to taking a counting semaphore.
@@ -2927,7 +2971,9 @@ void vTaskGenericNotifyGiveFromISR( TaskHandle_t xTaskToNotify,
  * @return The task's notification count before it is either cleared to zero or
  * decremented (see the xClearCountOnExit parameter).
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup ulTaskNotifyTakeIndexed ulTaskNotifyTakeIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 uint32_t ulTaskGenericNotifyTake( UBaseType_t uxIndexToWaitOn,
@@ -2939,12 +2985,14 @@ uint32_t ulTaskGenericNotifyTake( UBaseType_t uxIndexToWaitOn,
     ulTaskGenericNotifyTake( ( uxIndexToWaitOn ), ( xClearCountOnExit ), ( xTicksToWait ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * BaseType_t xTaskNotifyStateClearIndexed( TaskHandle_t xTask, UBaseType_t uxIndexToCLear );
  *
  * BaseType_t xTaskNotifyStateClear( TaskHandle_t xTask );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -2992,7 +3040,9 @@ uint32_t ulTaskGenericNotifyTake( UBaseType_t uxIndexToWaitOn,
  * @return pdTRUE if the task's notification state was set to
  * eNotWaitingNotification, otherwise pdFALSE.
  *
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskNotifyStateClearIndexed xTaskNotifyStateClearIndexed
+ * @endcond
  * \ingroup TaskNotifications
  */
 BaseType_t xTaskGenericNotifyStateClear( TaskHandle_t xTask,
@@ -3003,12 +3053,14 @@ BaseType_t xTaskGenericNotifyStateClear( TaskHandle_t xTask,
     xTaskGenericNotifyStateClear( ( xTask ), ( uxIndexToClear ) )
 
 /**
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task. h
  * @code{c}
  * uint32_t ulTaskNotifyValueClearIndexed( TaskHandle_t xTask, UBaseType_t uxIndexToClear, uint32_t ulBitsToClear );
  *
  * uint32_t ulTaskNotifyValueClear( TaskHandle_t xTask, uint32_t ulBitsToClear );
  * @endcode
+ * @endcond
  *
  * See https://www.FreeRTOS.org/RTOS-task-notifications.html for details.
  *
@@ -3057,7 +3109,9 @@ BaseType_t xTaskGenericNotifyStateClear( TaskHandle_t xTask,
  *
  * @return The value of the target task's notification value before the bits
  * specified by ulBitsToClear were cleared.
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup ulTaskNotifyValueClear ulTaskNotifyValueClear
+ * @endcond
  * \ingroup TaskNotifications
  */
 uint32_t ulTaskGenericNotifyValueClear( TaskHandle_t xTask,
@@ -3069,7 +3123,7 @@ uint32_t ulTaskGenericNotifyValueClear( TaskHandle_t xTask,
     ulTaskGenericNotifyValueClear( ( xTask ), ( uxIndexToClear ), ( ulBitsToClear ) )
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * void vTaskSetTimeOutState( TimeOut_t * const pxTimeOut );
@@ -3082,14 +3136,14 @@ uint32_t ulTaskGenericNotifyValueClear( TaskHandle_t xTask,
  * is to be captured.  The captured time includes the tick count and the number
  * of times the tick count has overflowed since the system first booted.
  * \defgroup vTaskSetTimeOutState vTaskSetTimeOutState
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \ingroup TaskCtrl
  * @endcond
  */
 void vTaskSetTimeOutState( TimeOut_t * const pxTimeOut ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code
  * BaseType_t xTaskCheckForTimeOut( TimeOut_t * const pxTimeOut, TickType_t * const pxTicksToWait );
@@ -3170,7 +3224,7 @@ void vTaskSetTimeOutState( TimeOut_t * const pxTimeOut ) PRIVILEGED_FUNCTION;
  *      return uxReceived;
  *  }
  * @endcode
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \defgroup xTaskCheckForTimeOut xTaskCheckForTimeOut
  * @endcond
  * \ingroup TaskCtrl
@@ -3179,7 +3233,7 @@ BaseType_t xTaskCheckForTimeOut( TimeOut_t * const pxTimeOut,
                                  TickType_t * const pxTicksToWait ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * task.h
  * @code{c}
  * BaseType_t xTaskCatchUpTicks( TickType_t xTicksToCatchUp );
@@ -3204,7 +3258,7 @@ BaseType_t xTaskCheckForTimeOut( TimeOut_t * const pxTimeOut,
  * blocked state and a context switch being performed.  Otherwise pdFALSE.
  *
  * \defgroup xTaskCatchUpTicks xTaskCatchUpTicks
- * @cond
+ * @cond !DOC_SINGLE_GROUP
  * \ingroup TaskCtrl
  * @endcond
  */
@@ -3214,7 +3268,7 @@ BaseType_t xTaskCatchUpTicks( TickType_t xTicksToCatchUp ) PRIVILEGED_FUNCTION;
 /*-----------------------------------------------------------
  * SCHEDULER INTERNALS AVAILABLE FOR PORTING PURPOSES
  *----------------------------------------------------------*/
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 /*
  * Return the handle of the task running on a certain CPU. Because of
  * the nature of SMP processing, there is no guarantee that this
@@ -3335,8 +3389,8 @@ void vTaskPlaceOnEventListRestricted( List_t * const pxEventList,
  * making the call, otherwise pdFALSE.
  */
 BaseType_t xTaskRemoveFromEventList( const List_t * const pxEventList ) PRIVILEGED_FUNCTION;
-BaseType_t xTaskRemoveFromUnorderedEventList( ListItem_t * pxEventListItem,
-                                              const TickType_t xItemValue ) PRIVILEGED_FUNCTION;
+void vTaskRemoveFromUnorderedEventList( ListItem_t * pxEventListItem,
+                                        const TickType_t xItemValue ) PRIVILEGED_FUNCTION;
 
 /*
  * THIS FUNCTION MUST NOT BE USED FROM APPLICATION CODE.  IT IS ONLY
@@ -3399,11 +3453,6 @@ void vTaskPriorityDisinheritAfterTimeout( TaskHandle_t const pxMutexHolder,
  */
 UBaseType_t uxTaskGetTaskNumber( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
 
-/*
- * Get the current core affinity of a task
- */
-BaseType_t xTaskGetAffinity( TaskHandle_t xTask ) PRIVILEGED_FUNCTION;
-
 /*
  * Set the uxTaskNumber of the task referenced by the xTask parameter to
  * uxHandle.
diff --git a/tools/sdk/esp32s2/include/freertos/include/freertos/timers.h b/tools/sdk/esp32s2/include/freertos/include/freertos/timers.h
index a8bc4f38c78..af6dcb23501 100644
--- a/tools/sdk/esp32s2/include/freertos/include/freertos/timers.h
+++ b/tools/sdk/esp32s2/include/freertos/include/freertos/timers.h
@@ -450,7 +450,7 @@ void vTimerSetTimerID( TimerHandle_t xTimer,
 BaseType_t xTimerIsTimerActive( TimerHandle_t xTimer ) PRIVILEGED_FUNCTION;
 
 /**
- * @cond
+ * @cond !DOC_EXCLUDE_HEADER_SECTION
  * TaskHandle_t xTimerGetTimerDaemonTaskHandle( void );
  * @endcond
  *
@@ -1315,7 +1315,7 @@ TickType_t xTimerGetPeriod( TimerHandle_t xTimer ) PRIVILEGED_FUNCTION;
  */
 TickType_t xTimerGetExpiryTime( TimerHandle_t xTimer ) PRIVILEGED_FUNCTION;
 
-/** @cond */
+/** @cond !DOC_EXCLUDE_HEADER_SECTION */
 
 /*
  * Functions beyond this part are not part of the public API and are intended
@@ -1339,7 +1339,7 @@ BaseType_t xTimerGenericCommand( TimerHandle_t xTimer,
 #if ( configSUPPORT_STATIC_ALLOCATION == 1 )
 
     /**
-     * @cond
+     * @cond !DOC_EXCLUDE_HEADER_SECTION
      * task.h
      * @code{c}
      * void vApplicationGetTimerTaskMemory( StaticTask_t ** ppxTimerTaskTCBBuffer, StackType_t ** ppxTimerTaskStackBuffer, uint32_t *pulTimerTaskStackSize )
diff --git a/tools/sdk/esp32s2/include/freertos/port/xtensa/include/freertos/portmacro.h b/tools/sdk/esp32s2/include/freertos/port/xtensa/include/freertos/portmacro.h
index 248c86d15c7..47187379c6c 100644
--- a/tools/sdk/esp32s2/include/freertos/port/xtensa/include/freertos/portmacro.h
+++ b/tools/sdk/esp32s2/include/freertos/port/xtensa/include/freertos/portmacro.h
@@ -24,317 +24,449 @@
  *
  * 1 tab == 4 spaces!
  */
+
 #ifndef PORTMACRO_H
 #define PORTMACRO_H
 
-#ifdef __cplusplus
-extern "C" {
-#endif
-
 #ifndef __ASSEMBLER__
 
-#include <sdkconfig.h>
+#include "sdkconfig.h"
 #include <stdint.h>
 #include <stdlib.h>
 #include <stdbool.h>
 #include <stdarg.h>
-#include <xtensa/hal.h>
 #include <xtensa/config/core.h>
-#include <xtensa/config/system.h>	/* required for XSHAL_CLIB */
-#include <xtensa/xtruntime.h>
-#include "esp_private/crosscore_int.h"
-#include "esp_timer.h"              /* required for FreeRTOS run time stats */
-#include "esp_system.h"
-#include "esp_newlib.h"
+#include <xtensa/hal.h>             /* required for xthal_get_ccount. [refactor-todo] use cpu_hal instead */
+#include <xtensa/xtruntime.h>       /* required for XTOS_SET_INTLEVEL. [refactor-todo] add common intr functions to esp_hw_support */
+#include "xt_instr_macros.h"
 #include "soc/spinlock.h"
-#include <esp_heap_caps.h>
+#include "hal/cpu_hal.h"
+#include "esp_private/crosscore_int.h"
+#include "esp_attr.h"
+#include "esp_timer.h"              /* required for esp_timer_get_time. [refactor-todo] make this common between archs */
+#include "esp_newlib.h"             /* required for esp_reent_init() in tasks.c */
+#include "esp_heap_caps.h"
 #include "esp_rom_sys.h"
-#include "sdkconfig.h"
-#include "freertos/xtensa_api.h"
-#include "esp_system.h"
-#include "soc/cpu.h"
-#include <limits.h>
+#include "esp_system.h"             /* required by esp_get_...() functions in portable.h. [refactor-todo] Update portable.h */
+#include "portbenchmark.h"
 
+/* [refactor-todo] These includes are not directly used in this file. They are kept into to prevent a breaking change. Remove these. */
+#include <limits.h>
+#include <xtensa/config/system.h>
+#include <xtensa/xtensa_api.h>
+#include "soc/cpu.h"
 #ifdef CONFIG_LEGACY_INCLUDE_COMMON_HEADERS
 #include "soc/soc_memory_layout.h"
 #endif
 
-/*-----------------------------------------------------------
- * Port specific definitions.
- *
- * The settings in this file configure FreeRTOS correctly for the
- * given hardware and compiler.
- *
- * These settings should not be altered.
- *-----------------------------------------------------------
- */
+#ifdef __cplusplus
+extern "C" {
+#endif
+
 
-#include "esp_system.h"
-#include "hal/cpu_hal.h"
-#include "xt_instr_macros.h"
 
-/* Type definitions. */
-#define portCHAR		int8_t
-#define portFLOAT		float
-#define portDOUBLE		double
-#define portLONG		int32_t
-#define portSHORT		int16_t
-#define portSTACK_TYPE	uint8_t
-#define portBASE_TYPE	int
+/* --------------------------------------------------- Port Types ------------------------------------------------------
+ * - Port specific types.
+ * - The settings in this file configure FreeRTOS correctly for the given hardware and compiler.
+ * - These settings should not be altered.
+ * - The port types must come before first as they are used further down the file
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-typedef portSTACK_TYPE			StackType_t;
-typedef portBASE_TYPE			BaseType_t;
-typedef unsigned portBASE_TYPE	UBaseType_t;
+#define portCHAR                    int8_t
+#define portFLOAT                   float
+#define portDOUBLE                  double
+#define portLONG                    int32_t
+#define portSHORT                   int16_t
+#define portSTACK_TYPE              uint8_t
+#define portBASE_TYPE               int
+
+typedef portSTACK_TYPE              StackType_t;
+typedef portBASE_TYPE               BaseType_t;
+typedef unsigned portBASE_TYPE      UBaseType_t;
 
 #if( configUSE_16_BIT_TICKS == 1 )
-	typedef uint16_t TickType_t;
-	#define portMAX_DELAY ( TickType_t ) 0xffff
+typedef uint16_t TickType_t;
+#define portMAX_DELAY ( TickType_t ) 0xffff
 #else
-	typedef uint32_t TickType_t;
-	#define portMAX_DELAY ( TickType_t ) 0xffffffffUL
+typedef uint32_t TickType_t;
+#define portMAX_DELAY ( TickType_t ) 0xffffffffUL
 #endif
-/*-----------------------------------------------------------*/
 
-// portbenchmark
-#include "portbenchmark.h"
+/* Task function macros as described on the FreeRTOS.org WEB site. */
+#define portTASK_FUNCTION_PROTO( vFunction, pvParameters )  void vFunction( void *pvParameters )
+#define portTASK_FUNCTION( vFunction, pvParameters )        void vFunction( void *pvParameters )
 
-#include "sdkconfig.h"
-#include "esp_attr.h"
-#include "portmacro_priv.h"
-
-// Cleaner solution allows nested interrupts disabling and restoring via local registers or stack.
-// They can be called from interrupts too.
-// WARNING: Only applies to current CPU. See notes above.
-static inline unsigned portENTER_CRITICAL_NESTED(void) {
-	unsigned state = XTOS_SET_INTLEVEL(XCHAL_EXCM_LEVEL);
-	portbenchmarkINTERRUPT_DISABLE();
-	return state;
-}
-#define portEXIT_CRITICAL_NESTED(state)   do { portbenchmarkINTERRUPT_RESTORE(state); XTOS_RESTORE_JUST_INTLEVEL(state); } while (0)
 
-/*
-Modifications to portENTER_CRITICAL.
 
-For an introduction, see "Critical Sections & Disabling Interrupts" in docs/api-guides/freertos-smp.rst
+/* ----------------------------------------------- Port Configurations -------------------------------------------------
+ * - Configurations values supplied by each port
+ * - Required by FreeRTOS
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-The original portENTER_CRITICAL only disabled the ISRs. This is enough for single-CPU operation: by
-disabling the interrupts, there is no task switch so no other tasks can meddle in the data, and because
-interrupts are disabled, ISRs can't corrupt data structures either.
+#define portCRITICAL_NESTING_IN_TCB     0
+#define portSTACK_GROWTH                ( -1 )
+#define portTICK_PERIOD_MS              ( ( TickType_t ) 1000 / configTICK_RATE_HZ )
+#define portBYTE_ALIGNMENT              4
+#define portNOP()                       XT_NOP()
 
-For multiprocessing, things get a bit more hairy. First of all, disabling the interrupts doesn't stop
-the tasks or ISRs on the other processors meddling with our CPU. For tasks, this is solved by adding
-a spinlock to the portENTER_CRITICAL macro. A task running on the other CPU accessing the same data will
-spinlock in the portENTER_CRITICAL code until the first CPU is done.
 
-For ISRs, we now also need muxes: while portENTER_CRITICAL disabling interrupts will stop ISRs on the same
-CPU from meddling with the data, it does not stop interrupts on the other cores from interfering with the
-data. For this, we also use a spinlock in the routines called by the ISR, but these spinlocks
-do not disable the interrupts (because they already are).
 
-This all assumes that interrupts are either entirely disabled or enabled. Interrupt priority levels
-will break this scheme.
+/* ---------------------------------------------- Forward Declarations -------------------------------------------------
+ * - Forward declarations of all the port functions and macros need to implement the FreeRTOS porting interface
+ * - These must come before definition/declaration of the FreeRTOS porting interface
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-Remark: For the ESP32, portENTER_CRITICAL and portENTER_CRITICAL_ISR both alias vPortEnterCritical, meaning
-that either function can be called both from ISR as well as task context. This is not standard FreeRTOS
-behaviour; please keep this in mind if you need any compatibility with other FreeRTOS implementations.
-*/
-/* "mux" data structure (spinlock) */
-typedef spinlock_t portMUX_TYPE;
+// --------------------- Interrupts ------------------------
 
-#define portMUX_FREE_VAL		SPINLOCK_FREE
-#define portMUX_NO_TIMEOUT      SPINLOCK_WAIT_FOREVER  /* When passed for 'timeout_cycles', spin forever if necessary */
-#define portMUX_TRY_LOCK        SPINLOCK_NO_WAIT       /* Try to acquire the spinlock a single time only */
-#define portMUX_INITIALIZER_UNLOCKED  SPINLOCK_INITIALIZER
+/**
+ * @brief Checks if the current core is in an ISR context
+ *
+ * - ISR context consist of Low/Mid priority ISR, or time tick ISR
+ * - High priority ISRs aren't detected here, but they normally cannot call C code, so that should not be an issue anyway.
+ *
+ * @note [refactor-todo] Check if this should be inlined
+ * @return
+ *  - pdTRUE if in ISR
+ *  - pdFALSE otherwise
+ */
+BaseType_t xPortInIsrContext(void);
 
-#define portCRITICAL_NESTING_IN_TCB 0
+/**
+ * @brief Asserts if in ISR context
+ *
+ * - Asserts on xPortInIsrContext() internally
+ *
+ * @note [refactor-todo] Check if this API is still required
+ * @note [refactor-todo] Check if this should be inlined
+ */
+void vPortAssertIfInISR(void);
 
-static inline void __attribute__((always_inline)) vPortCPUInitializeMutex(portMUX_TYPE *mux)
-{
-    spinlock_initialize(mux);
-}
+/**
+ * @brief Check if in ISR context from High priority ISRs
+ *
+ * - Called from High priority ISR
+ * - Checks if the previous context (before high priority interrupt) was in ISR context (meaning low/med priority)
+ *
+ * @note [refactor-todo] Check if this should be inlined
+ * @return
+ *  - pdTRUE if in previous in ISR context
+ *  - pdFALSE otherwise
+ */
+BaseType_t xPortInterruptedFromISRContext(void);
 
-static inline void __attribute__((always_inline)) vPortCPUAcquireMutex(portMUX_TYPE *mux)
-{
-    spinlock_acquire(mux, portMUX_NO_TIMEOUT);
-}
+/**
+ * @brief Disable interrupts in a nested manner
+ *
+ * - Cleaner solution allows nested interrupts disabling and restoring via local registers or stack.
+ * - They can be called from interrupts too.
+ * - WARNING Only applies to current CPU.
+ * @note [refactor-todo] Define this as portSET_INTERRUPT_MASK_FROM_ISR() instead
+ * @return unsigned Previous interrupt state
+ */
+static inline unsigned __attribute__((always_inline)) portENTER_CRITICAL_NESTED(void);
+
+/* ---------------------- Spinlocks ------------------------
+ * - Modifications made to critical sections to support SMP
+ * - See "Critical Sections & Disabling Interrupts" in docs/api-guides/freertos-smp.rst for more details
+ * - Remark: For the ESP32, portENTER_CRITICAL and portENTER_CRITICAL_ISR both alias vPortEnterCritical, meaning that
+ *           either function can be called both from ISR as well as task context. This is not standard FreeRTOS
+ *           behaviorr; please keep this in mind if you need any compatibility with other FreeRTOS implementations.
+ * @note [refactor-todo] Check if these comments are still true
+ * ------------------------------------------------------ */
+
+typedef spinlock_t                      portMUX_TYPE;           /**< Spinlock type used by FreeRTOS critical sections */
+#define portMUX_INITIALIZER_UNLOCKED    SPINLOCK_INITIALIZER    /**< Spinlock initializer */
+#define portMUX_FREE_VAL                SPINLOCK_FREE           /**< Spinlock is free. [refactor-todo] check if this is still required */
+#define portMUX_NO_TIMEOUT              SPINLOCK_WAIT_FOREVER   /**< When passed for 'timeout_cycles', spin forever if necessary. [refactor-todo] check if this is still required */
+#define portMUX_TRY_LOCK                SPINLOCK_NO_WAIT        /**< Try to acquire the spinlock a single time only. [refactor-todo] check if this is still required */
 
-static inline bool __attribute__((always_inline)) vPortCPUAcquireMutexTimeout(portMUX_TYPE *mux, int timeout)
-{
-    return (spinlock_acquire(mux, timeout));
-}
+/**
+ * @brief Initialize a spinlock
+ *
+ * - Initializes a spinlock that is used by FreeRTOS SMP critical sections
+ *
+ * @param[in] mux Spinlock
+ */
+static inline void __attribute__((always_inline)) vPortCPUInitializeMutex(portMUX_TYPE *mux);
 
-static inline void __attribute__((always_inline)) vPortCPUReleaseMutex(portMUX_TYPE *mux)
-{
-    spinlock_release(mux);
-}
+/**
+ * @brief Acquire a spinlock
+ *
+ * @note [refactor-todo] check if we still need this
+ *
+ * @param[in] mux Spinlock
+ */
+static inline void __attribute__((always_inline)) vPortCPUAcquireMutex(portMUX_TYPE *mux);
 
+/**
+ * @brief Acquire a spinlock but with a specified timeout
+ *
+ * @note [refactor-todo] check if we still need this
+ * @note [refactor-todo] Check if this function should be renamed (due to bool return type)
+ *
+ * @param[in] mux Spinlock
+ * @param timeout
+ * @return true Spinlock acquired
+ * @return false Timed out
+ */
+static inline bool __attribute__((always_inline)) vPortCPUAcquireMutexTimeout(portMUX_TYPE *mux, int timeout);
+
+/**
+ * @brief Release a spinlock
+ *
+ * @note [refactor-todo] check if we still need this
+ *
+ * @param[in] mux Spinlock
+ */
+static inline void __attribute__((always_inline)) vPortCPUReleaseMutex(portMUX_TYPE *mux);
+
+/**
+ * @brief Wrapper for atomic compare-and-set instruction
+ *
+ * This subroutine will atomically compare *addr to 'compare'. If *addr == compare, *addr is set to *set. *set is
+ * updated with the previous value of *addr (either 'compare' or some other value.)
+ *
+ * @warning From the ISA docs: in some (unspecified) cases, the s32c1i instruction may return the "bitwise inverse" of
+ *          the old mem if the mem wasn't written. This doesn't seem to happen on the ESP32 (portMUX assertions would
+ *          fail).
+ *
+ * @note [refactor-todo] check if we still need this
+ * @note [refactor-todo] Check if this function should be renamed (due to void return type)
+ *
+ * @param[inout] addr Pointer to target address
+ * @param[in] compare Compare value
+ * @param[inout] set Pointer to set value
+ */
+static inline void __attribute__((always_inline)) uxPortCompareSet(volatile uint32_t *addr, uint32_t compare, uint32_t *set);
+
+/**
+ * @brief Wrapper for atomic compare-and-set instruction in external RAM
+ *
+ * Atomic compare-and-set but the target address is placed in external RAM
+ *
+ * @note [refactor-todo] check if we still need this
+ *
+ * @param[inout] addr Pointer to target address
+ * @param[in] compare Compare value
+ * @param[inout] set Pointer to set value
+ */
+static inline void uxPortCompareSetExtram(volatile uint32_t *addr, uint32_t compare, uint32_t *set);
+
+// ------------------ Critical Sections --------------------
+
+/**
+ * @brief Enter a SMP critical section
+ *
+ * - Disable interrupts
+ * - Takes spinlock
+ * - Can be nested
+ *
+ * @param[in] mux Spinlock
+ */
 void vPortEnterCritical(portMUX_TYPE *mux);
+
+/**
+ * @brief Exit a SMP critical section
+ *
+ * - Releases spinlock
+ * - Reenables interrupts
+ * - Can be nested
+ *
+ * @param[in] mux Spinlock
+ */
 void vPortExitCritical(portMUX_TYPE *mux);
 
-#define portASSERT_IF_IN_ISR()  vPortAssertIfInISR()
-void vPortAssertIfInISR(void);
+/**
+ * @brief FreeRTOS compliant version of enter critical
+ *
+ * - Ensures that critical section is only entered from task context
+ *
+ * @param[in] mux Spinlock
+ */
+static inline void __attribute__((always_inline)) vPortEnterCriticalCompliance(portMUX_TYPE *mux);
 
-/*
- * Returns true if the current core is in ISR context; low prio ISR, med prio ISR or timer tick ISR. High prio ISRs
- * aren't detected here, but they normally cannot call C code, so that should not be an issue anyway.
+/**
+ * @brief FreeRTOS compliant version of exit critical
+ *
+ * @param[in] mux Spinlock
  */
-BaseType_t xPortInIsrContext(void);
+static inline void __attribute__((always_inline)) vPortExitCriticalCompliance(portMUX_TYPE *mux);
 
-static inline void __attribute__((always_inline)) vPortEnterCriticalCompliance(portMUX_TYPE *mux)
-{
-    if(!xPortInIsrContext()) {
-        vPortEnterCritical(mux);
-    } else {
-        esp_rom_printf("%s:%d (%s)- port*_CRITICAL called from ISR context!\n",
-                       __FILE__, __LINE__, __FUNCTION__);
-        abort();
-    }
-}
+/**
+ * @brief Safe version of enter critical
+ *
+ * - This function can be used to enter a critical section from both task and ISR contexts
+ *
+ * @param[in] mux Spinlock
+ */
+static inline void __attribute__((always_inline)) vPortEnterCriticalSafe(portMUX_TYPE *mux);
 
-static inline void __attribute__((always_inline)) vPortExitCriticalCompliance(portMUX_TYPE *mux)
-{
-    if(!xPortInIsrContext()) {
-        vPortExitCritical(mux);
-    } else {
-        esp_rom_printf("%s:%d (%s)- port*_CRITICAL called from ISR context!\n",
-                       __FILE__, __LINE__, __FUNCTION__);
-        abort();
-    }
-}
+/**
+ * @brief Safe version of exit critical
+ *
+ * @param[in] mux Spinlock
+ */
+static inline void __attribute__((always_inline)) vPortExitCriticalSafe(portMUX_TYPE *mux);
 
-#ifdef CONFIG_FREERTOS_CHECK_PORT_CRITICAL_COMPLIANCE
-/* Calling port*_CRITICAL from ISR context would cause an assert failure.
- * If the parent function is called from both ISR and Non-ISR context then call port*_CRITICAL_SAFE
+// ---------------------- Yielding -------------------------
+
+/**
+ * @brief Perform a solicited context switch
+ *
+ * - Defined in portasm.S
+ *
+ * @note [refactor-todo] The rest of ESP-IDF should call taskYield() instead
  */
-#define portENTER_CRITICAL(mux)        vPortEnterCriticalCompliance(mux)
-#define portEXIT_CRITICAL(mux)         vPortExitCriticalCompliance(mux)
-#else
-#define portENTER_CRITICAL(mux)        vPortEnterCritical(mux)
-#define portEXIT_CRITICAL(mux)         vPortExitCritical(mux)
-#endif
+void vPortYield( void );
 
-#define portENTER_CRITICAL_ISR(mux)    vPortEnterCritical(mux)
-#define portEXIT_CRITICAL_ISR(mux)     vPortExitCritical(mux)
+/**
+ * @brief
+ *
+ * @note [refactor-todo] Refactor this to avoid va_args
+ * @param argc
+ * @param ... Variable arguments to allow for IDF prototype without arguments, and vanilla version WITH argument
+ */
+void vPortEvaluateYieldFromISR(int argc, ...);
 
-static inline void __attribute__((always_inline)) vPortEnterCriticalSafe(portMUX_TYPE *mux)
-{
-    if (xPortInIsrContext()) {
-        portENTER_CRITICAL_ISR(mux);
-    } else {
-        portENTER_CRITICAL(mux);
-    }
-}
+/**
+ * @brief Yields the other core
+ *
+ * - Send an interrupt to another core in order to make the task running on it yield for a higher-priority task.
+ * - Can be used to yield current core as well
+ *
+ * @note [refactor-todo] Put this into private macros as its only called from task.c and is not public API
+ * @param coreid ID of core to yield
+ */
+void vPortYieldOtherCore(BaseType_t coreid);
 
-static inline void __attribute__((always_inline)) vPortExitCriticalSafe(portMUX_TYPE *mux)
-{
-    if (xPortInIsrContext()) {
-        portEXIT_CRITICAL_ISR(mux);
-    } else {
-        portEXIT_CRITICAL(mux);
-    }
-}
+/**
+ * @brief Checks if the current core can yield
+ *
+ * - A core cannot yield if its in an ISR or in a critical section
+ *
+ * @note [refactor-todo] See if this can be separated from port macro
+ * @return true Core can yield
+ * @return false Core cannot yield
+ */
+static inline bool IRAM_ATTR xPortCanYield(void);
 
-#define portENTER_CRITICAL_SAFE(mux)  vPortEnterCriticalSafe(mux)
-#define portEXIT_CRITICAL_SAFE(mux)  vPortExitCriticalSafe(mux)
+// ------------------- Hook Functions ----------------------
 
-/*
- * Wrapper for the Xtensa compare-and-set instruction. This subroutine will atomically compare
- * *addr to 'compare'. If *addr == compare, *addr is set to *set. *set is updated with the previous
- * value of *addr (either 'compare' or some other value.)
+extern void esp_vApplicationIdleHook(void);     /* Required by tasks.c */
+extern void esp_vApplicationTickHook(void);     /* Required by tasks.c */
+
+/**
+ * @brief Hook function called on entry to tickless idle
+ *
+ * - Implemented in pm_impl.c
  *
- * Warning: From the ISA docs: in some (unspecified) cases, the s32c1i instruction may return the
- * *bitwise inverse* of the old mem if the mem wasn't written. This doesn't seem to happen on the
- * ESP32 (portMUX assertions would fail).
+ * @param xExpectedIdleTime Expected idle time
  */
-static inline void __attribute__((always_inline)) uxPortCompareSet(volatile uint32_t *addr, uint32_t compare, uint32_t *set) {
-    compare_and_set_native(addr, compare, set);
-}
+void vApplicationSleep(TickType_t xExpectedIdleTime);
 
-// Critical section management. NW-TODO: replace XTOS_SET_INTLEVEL with more efficient version, if any?
-// These cannot be nested. They should be used with a lot of care and cannot be called from interrupt level.
-//
-// Only applies to one CPU. See notes above & below for reasons not to use these.
-#define portDISABLE_INTERRUPTS()      do { XTOS_SET_INTLEVEL(XCHAL_EXCM_LEVEL); portbenchmarkINTERRUPT_DISABLE(); } while (0)
-#define portENABLE_INTERRUPTS()       do { portbenchmarkINTERRUPT_RESTORE(0); XTOS_SET_INTLEVEL(0); } while (0)
+// ----------------------- System --------------------------
 
+/**
+ * @brief Get the tick rate per second
+ *
+ * @note [refactor-todo] make this inline
+ * @return uint32_t Tick rate in Hz
+ */
+uint32_t xPortGetTickRateHz(void);
 
-// These FreeRTOS versions are similar to the nested versions above
-#define portSET_INTERRUPT_MASK_FROM_ISR()            portENTER_CRITICAL_NESTED()
-#define portCLEAR_INTERRUPT_MASK_FROM_ISR(state)     portEXIT_CRITICAL_NESTED(state)
+/**
+ * @brief Set a watchpoint to watch the last 32 bytes of the stack
+ *
+ * Callback to set a watchpoint on the end of the stack. Called every context switch to change the stack watchpoint
+ * around.
+ *
+ * @param pxStackStart Pointer to the start of the stack
+ */
+void vPortSetStackWatchpoint( void *pxStackStart );
 
-//Because the ROM routines don't necessarily handle a stack in external RAM correctly, we force
-//the stack memory to always be internal.
-#define portTcbMemoryCaps (MALLOC_CAP_INTERNAL|MALLOC_CAP_8BIT)
-#define portStackMemoryCaps (MALLOC_CAP_INTERNAL|MALLOC_CAP_8BIT)
-#define pvPortMallocTcbMem(size) heap_caps_malloc(size, portTcbMemoryCaps)
-#define pvPortMallocStackMem(size)  heap_caps_malloc(size, portStackMemoryCaps)
-
-//xTaskCreateStatic uses these functions to check incoming memory.
-#define portVALID_TCB_MEM(ptr) (esp_ptr_internal(ptr) && esp_ptr_byte_accessible(ptr))
-#ifdef CONFIG_SPIRAM_ALLOW_STACK_EXTERNAL_MEMORY
-#define portVALID_STACK_MEM(ptr) esp_ptr_byte_accessible(ptr)
-#else
-#define portVALID_STACK_MEM(ptr) (esp_ptr_internal(ptr) && esp_ptr_byte_accessible(ptr))
-#endif
+/**
+ * @brief Get the current core's ID
+ *
+ * @note [refactor-todo] IDF should call a FreeRTOS like macro instead of port function directly
+ * @return BaseType_t Core ID
+ */
+static inline BaseType_t IRAM_ATTR xPortGetCoreID(void);
 
 
-static inline void uxPortCompareSetExtram(volatile uint32_t *addr, uint32_t compare, uint32_t *set)
-{
-#ifdef CONFIG_SPIRAM
-    compare_and_set_extram(addr, compare, set);
-#endif
-}
 
+/* ------------------------------------------- FreeRTOS Porting Interface ----------------------------------------------
+ * - Contains all the mappings of the macros required by FreeRTOS
+ * - Most come after forward declare as porting macros map to declared functions
+ * - Maps to forward declared functions
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-/*-----------------------------------------------------------*/
+// ----------------------- Memory --------------------------
 
-/* Architecture specifics. */
-#define portSTACK_GROWTH			( -1 )
-#define portTICK_PERIOD_MS			( ( TickType_t ) 1000 / configTICK_RATE_HZ )
-#define portBYTE_ALIGNMENT			4
-#define portNOP()					XT_NOP()
-/*-----------------------------------------------------------*/
+/**
+ * @brief Task memory allocation macros
+ *
+ * @note Because the ROM routines don't necessarily handle a stack in external RAM correctly, we force the stack
+ * memory to always be internal.
+ * @note [refactor-todo] Update portable.h to match v10.4.3 to use new malloc prototypes
+ */
+#define portTcbMemoryCaps               (MALLOC_CAP_INTERNAL|MALLOC_CAP_8BIT)
+#define portStackMemoryCaps             (MALLOC_CAP_INTERNAL|MALLOC_CAP_8BIT)
+#define pvPortMallocTcbMem(size)        heap_caps_malloc(size, portTcbMemoryCaps)
+#define pvPortMallocStackMem(size)      heap_caps_malloc(size, portStackMemoryCaps)
 
-/* Fine resolution time */
-#define portGET_RUN_TIME_COUNTER_VALUE()  xthal_get_ccount()
-//ccount or esp_timer are initialized elsewhere
-#define portCONFIGURE_TIMER_FOR_RUN_TIME_STATS()
+// --------------------- Interrupts ------------------------
 
-#ifdef CONFIG_FREERTOS_RUN_TIME_STATS_USING_ESP_TIMER
-/* Coarse resolution time (us) */
-#define portALT_GET_RUN_TIME_COUNTER_VALUE(x)     do {x = (uint32_t)esp_timer_get_time();} while(0)
-#endif
+#define portEXIT_CRITICAL_NESTED(state)     do { portbenchmarkINTERRUPT_RESTORE(state); XTOS_RESTORE_JUST_INTLEVEL(state); } while (0)
 
-void vPortYield( void );
-void vPortEvaluateYieldFromISR(int argc, ...);
-void _frxt_setup_switch( void );
+/**
+ * - Only applies to current core
+ * - These cannot be nested. They should be used with a lot of care and cannot be called from interrupt level.
+ *
+ * @note [refactor-todo] replace XTOS_SET_INTLEVEL with more efficient version, if any?
+ */
+#define portDISABLE_INTERRUPTS()            do { XTOS_SET_INTLEVEL(XCHAL_EXCM_LEVEL); portbenchmarkINTERRUPT_DISABLE(); } while (0)
+#define portENABLE_INTERRUPTS()             do { portbenchmarkINTERRUPT_RESTORE(0); XTOS_SET_INTLEVEL(0); } while (0)
 
 /**
- * Macro to count number of arguments of a __VA_ARGS__ used to support portYIELD_FROM_ISR with,
- * or without arguments. The macro counts only 0 or 1 arguments.
+ * ISR versions to enable/disable interrupts
+ */
+#define portSET_INTERRUPT_MASK_FROM_ISR()            portENTER_CRITICAL_NESTED()
+#define portCLEAR_INTERRUPT_MASK_FROM_ISR(state)     portEXIT_CRITICAL_NESTED(state)
+
+#define portASSERT_IF_IN_ISR() vPortAssertIfInISR()
+
+// ------------------ Critical Sections --------------------
+
+/**
+ * @brief FreeRTOS critical section macros
  *
- * In the future, we want to switch to C++20. We also want to become compatible with clang.
- * Hence, we provide two versions of the following macros which are using variadic arguments.
- * The first one is using the GNU extension ##__VA_ARGS__. The second one is using the C++20 feature __VA_OPT__(,).
- * This allows users to compile their code with standard C++20 enabled instead of the GNU extension.
- * Below C++20, we haven't found any good alternative to using ##__VA_ARGS__.
+ * - Added a spinlock argument for SMP
+ * - Can be nested
+ * - Compliance versions will assert if regular critical section API is used in ISR context
+ * - Safe versions can be called from either contexts
  */
-#if defined(__cplusplus) && (__cplusplus >  201703L)
-#define portGET_ARGUMENT_COUNT(...) portGET_ARGUMENT_COUNT_INNER(0 __VA_OPT__(,) __VA_ARGS__,1,0)
+#ifdef CONFIG_FREERTOS_CHECK_PORT_CRITICAL_COMPLIANCE
+#define portENTER_CRITICAL(mux)         vPortEnterCriticalCompliance(mux)
+#define portEXIT_CRITICAL(mux)          vPortExitCriticalCompliance(mux)
 #else
-#define portGET_ARGUMENT_COUNT(...) portGET_ARGUMENT_COUNT_INNER(0, ##__VA_ARGS__,1,0)
-#endif
-#define portGET_ARGUMENT_COUNT_INNER(zero, one, count, ...) count
+#define portENTER_CRITICAL(mux)         vPortEnterCritical(mux)
+#define portEXIT_CRITICAL(mux)          vPortExitCritical(mux)
+#endif /* CONFIG_FREERTOS_CHECK_PORT_CRITICAL_COMPLIANCE */
+#define portENTER_CRITICAL_ISR(mux)     vPortEnterCritical(mux)
+#define portEXIT_CRITICAL_ISR(mux)      vPortExitCritical(mux)
+#define portENTER_CRITICAL_SAFE(mux)    vPortEnterCriticalSafe(mux)
+#define portEXIT_CRITICAL_SAFE(mux)     vPortExitCriticalSafe(mux)
 
-_Static_assert(portGET_ARGUMENT_COUNT() == 0, "portGET_ARGUMENT_COUNT() result does not match for 0 arguments");
-_Static_assert(portGET_ARGUMENT_COUNT(1) == 1, "portGET_ARGUMENT_COUNT() result does not match for 1 argument");
+// ---------------------- Yielding -------------------------
 
-#define portYIELD()	vPortYield()
+#define portYIELD() vPortYield()
 
 /**
  * @note    The macro below could be used when passing a single argument, or without any argument,
  *          it was developed to support both usages of portYIELD inside of an ISR. Any other usage form
- *          might result in undesired behaviour
+ *          might result in undesired behavior
+ *
+ * @note [refactor-todo] Refactor this to avoid va_args
  */
 #if defined(__cplusplus) && (__cplusplus >  201703L)
 #define portYIELD_FROM_ISR(...) vPortEvaluateYieldFromISR(portGET_ARGUMENT_COUNT(__VA_ARGS__) __VA_OPT__(,) __VA_ARGS__)
@@ -351,121 +483,137 @@ _Static_assert(portGET_ARGUMENT_COUNT(1) == 1, "portGET_ARGUMENT_COUNT() result
 */
 #define portYIELD_WITHIN_API() esp_crosscore_int_send_yield(xPortGetCoreID())
 
-/*-----------------------------------------------------------*/
-
-/* Task function macros as described on the FreeRTOS.org WEB site. */
-#define portTASK_FUNCTION_PROTO( vFunction, pvParameters ) void vFunction( void *pvParameters )
-#define portTASK_FUNCTION( vFunction, pvParameters ) void vFunction( void *pvParameters )
-
-// When coprocessors are defined, we to maintain a pointer to coprocessors area.
-// We currently use a hack: redefine field xMPU_SETTINGS in TCB block as a structure that can hold:
-// MPU wrappers, coprocessor area pointer, trace code structure, and more if needed.
-// The field is normally used for memory protection. FreeRTOS should create another general purpose field.
-typedef struct {
-	#if XCHAL_CP_NUM > 0
-	volatile StackType_t* coproc_area; // Pointer to coprocessor save area; MUST BE FIRST
-	#endif
-
-	#if portUSING_MPU_WRAPPERS
-	// Define here mpu_settings, which is port dependent
-	int mpu_setting; // Just a dummy example here; MPU not ported to Xtensa yet
-	#endif
-
-	#if configUSE_TRACE_FACILITY_2
-	struct {
-		// Cf. porttraceStamp()
-		int taskstamp;        /* Stamp from inside task to see where we are */
-		int taskstampcount;   /* A counter usually incremented when we restart the task's loop */
-	} porttrace;
-	#endif
-} xMPU_SETTINGS;
-
-// Main hack to use MPU_wrappers even when no MPU is defined (warning: mpu_setting should not be accessed; otherwise move this above xMPU_SETTINGS)
-#if (XCHAL_CP_NUM > 0 || configUSE_TRACE_FACILITY_2) && !portUSING_MPU_WRAPPERS   // If MPU wrappers not used, we still need to allocate coproc area
-	#undef portUSING_MPU_WRAPPERS
-	#define portUSING_MPU_WRAPPERS 1   // Enable it to allocate coproc area
-	#define MPU_WRAPPERS_H             // Override mpu_wrapper.h to disable unwanted code
-	#define PRIVILEGED_FUNCTION
-	#define PRIVILEGED_DATA
-#endif
-
-extern void esp_vApplicationIdleHook( void );
-extern void esp_vApplicationTickHook( void );
+// ------------------- Hook Functions ----------------------
 
 #ifndef CONFIG_FREERTOS_LEGACY_HOOKS
-#define vApplicationIdleHook    esp_vApplicationIdleHook
-#define vApplicationTickHook    esp_vApplicationTickHook
+#define vApplicationIdleHook esp_vApplicationIdleHook
+#define vApplicationTickHook esp_vApplicationTickHook
 #endif /* !CONFIG_FREERTOS_LEGACY_HOOKS */
 
-void vApplicationSleep( TickType_t xExpectedIdleTime );
+#define portSUPPRESS_TICKS_AND_SLEEP(idleTime) vApplicationSleep(idleTime)
 
-#define portSUPPRESS_TICKS_AND_SLEEP( idleTime ) vApplicationSleep( idleTime )
+// ------------------- Run Time Stats ----------------------
 
-void _xt_coproc_release(volatile void * coproc_sa_base);
+#define portCONFIGURE_TIMER_FOR_RUN_TIME_STATS()
 
-/* Architecture specific optimisations. */
-#if configUSE_PORT_OPTIMISED_TASK_SELECTION == 1
+/**
+ * - Fine resolution uses ccount
+ * - ALT is coarse and uses esp_timer
+ * @note [refactor-todo] Make fine and alt timers mutually exclusive
+ */
+#define portGET_RUN_TIME_COUNTER_VALUE() xthal_get_ccount()
+#ifdef CONFIG_FREERTOS_RUN_TIME_STATS_USING_ESP_TIMER
+#define portALT_GET_RUN_TIME_COUNTER_VALUE(x) do {x = (uint32_t)esp_timer_get_time();} while(0)
+#endif
 
+// -------------- Optimized Task Selection -----------------
+
+#if configUSE_PORT_OPTIMISED_TASK_SELECTION == 1
 /* Check the configuration. */
 #if( configMAX_PRIORITIES > 32 )
-    #error configUSE_PORT_OPTIMISED_TASK_SELECTION can only be set to 1 when configMAX_PRIORITIES is less than or equal to 32.  It is very rare that a system requires more than 10 to 15 different priorities as tasks that share a priority will time slice.
+#error configUSE_PORT_OPTIMISED_TASK_SELECTION can only be set to 1 when configMAX_PRIORITIES is less than or equal to 32.  It is very rare that a system requires more than 10 to 15 different priorities as tasks that share a priority will time slice.
 #endif
 
 /* Store/clear the ready priorities in a bit map. */
 #define portRECORD_READY_PRIORITY( uxPriority, uxReadyPriorities ) ( uxReadyPriorities ) |= ( 1UL << ( uxPriority ) )
 #define portRESET_READY_PRIORITY( uxPriority, uxReadyPriorities ) ( uxReadyPriorities ) &= ~( 1UL << ( uxPriority ) )
+#define portGET_HIGHEST_PRIORITY( uxTopPriority, uxReadyPriorities ) uxTopPriority = ( 31 - __builtin_clz( ( uxReadyPriorities ) ) )
+#endif /* configUSE_PORT_OPTIMISED_TASK_SELECTION */
 
-/*-----------------------------------------------------------*/
 
-#define portGET_HIGHEST_PRIORITY( uxTopPriority, uxReadyPriorities ) uxTopPriority = ( 31 - __builtin_clz( ( uxReadyPriorities ) ) )
 
-#endif /* configUSE_PORT_OPTIMISED_TASK_SELECTION */
+/* --------------------------------------------- Inline Implementations ------------------------------------------------
+ * - Implementation of inline functions of the forward declares
+ * - Should come after forward declare and FreeRTOS Porting interface, as implementation may use both.
+ * - For implementation of non-inlined functions, see port.c
+ * ------------------------------------------------------------------------------------------------------------------ */
 
-/*
- * Send an interrupt to another core in order to make the task running
- * on it yield for a higher-priority task.
- */
+// --------------------- Interrupts ------------------------
 
-void vPortYieldOtherCore( BaseType_t coreid) ;
+static inline unsigned portENTER_CRITICAL_NESTED(void)
+{
+    unsigned state = XTOS_SET_INTLEVEL(XCHAL_EXCM_LEVEL);
+    portbenchmarkINTERRUPT_DISABLE();
+    return state;
+}
 
-/*
- Callback to set a watchpoint on the end of the stack. Called every context switch to change the stack
- watchpoint around.
- */
-void vPortSetStackWatchpoint( void* pxStackStart );
+// ---------------------- Spinlocks ------------------------
 
-/*
- * Returns true if the current core is in ISR context; low prio ISR, med prio ISR or timer tick ISR. High prio ISRs
- * aren't detected here, but they normally cannot call C code, so that should not be an issue anyway.
- */
-BaseType_t xPortInIsrContext(void);
+static inline void __attribute__((always_inline)) vPortCPUInitializeMutex(portMUX_TYPE *mux)
+{
+    spinlock_initialize(mux);
+}
 
-/*
- * This function will be called in High prio ISRs. Returns true if the current core was in ISR context
- * before calling into high prio ISR context.
- */
-BaseType_t xPortInterruptedFromISRContext(void);
+static inline void __attribute__((always_inline)) vPortCPUAcquireMutex(portMUX_TYPE *mux)
+{
+    spinlock_acquire(mux, portMUX_NO_TIMEOUT);
+}
 
-/*
- * The structures and methods of manipulating the MPU are contained within the
- * port layer.
- *
- * Fills the xMPUSettings structure with the memory region information
- * contained in xRegions.
- */
-#if( portUSING_MPU_WRAPPERS == 1 )
-    struct xMEMORY_REGION;
-    void vPortStoreTaskMPUSettings( xMPU_SETTINGS *xMPUSettings, const struct xMEMORY_REGION * const xRegions, StackType_t *pxBottomOfStack, uint32_t usStackDepth ) PRIVILEGED_FUNCTION;
-    void vPortReleaseTaskMPUSettings( xMPU_SETTINGS *xMPUSettings );
+static inline bool __attribute__((always_inline)) vPortCPUAcquireMutexTimeout(portMUX_TYPE *mux, int timeout)
+{
+    return (spinlock_acquire(mux, timeout));
+}
+
+static inline void __attribute__((always_inline)) vPortCPUReleaseMutex(portMUX_TYPE *mux)
+{
+    spinlock_release(mux);
+}
+
+static inline void __attribute__((always_inline)) uxPortCompareSet(volatile uint32_t *addr, uint32_t compare, uint32_t *set)
+{
+    compare_and_set_native(addr, compare, set);
+}
+
+static inline void uxPortCompareSetExtram(volatile uint32_t *addr, uint32_t compare, uint32_t *set)
+{
+#ifdef CONFIG_SPIRAM
+    compare_and_set_extram(addr, compare, set);
 #endif
+}
 
-/* Multi-core: get current core ID */
-static inline BaseType_t IRAM_ATTR xPortGetCoreID(void) {
-    return cpu_hal_get_core_id();
+// ------------------ Critical Sections --------------------
+
+static inline void __attribute__((always_inline)) vPortEnterCriticalCompliance(portMUX_TYPE *mux)
+{
+    if (!xPortInIsrContext()) {
+        vPortEnterCritical(mux);
+    } else {
+        esp_rom_printf("%s:%d (%s)- port*_CRITICAL called from ISR context!\n",
+                       __FILE__, __LINE__, __FUNCTION__);
+        abort();
+    }
 }
 
-/* Get tick rate per second */
-uint32_t xPortGetTickRateHz(void);
+static inline void __attribute__((always_inline)) vPortExitCriticalCompliance(portMUX_TYPE *mux)
+{
+    if (!xPortInIsrContext()) {
+        vPortExitCritical(mux);
+    } else {
+        esp_rom_printf("%s:%d (%s)- port*_CRITICAL called from ISR context!\n",
+                       __FILE__, __LINE__, __FUNCTION__);
+        abort();
+    }
+}
+
+static inline void __attribute__((always_inline)) vPortEnterCriticalSafe(portMUX_TYPE *mux)
+{
+    if (xPortInIsrContext()) {
+        portENTER_CRITICAL_ISR(mux);
+    } else {
+        portENTER_CRITICAL(mux);
+    }
+}
+
+static inline void __attribute__((always_inline)) vPortExitCriticalSafe(portMUX_TYPE *mux)
+{
+    if (xPortInIsrContext()) {
+        portEXIT_CRITICAL_ISR(mux);
+    } else {
+        portEXIT_CRITICAL(mux);
+    }
+}
+
+// ---------------------- Yielding -------------------------
 
 static inline bool IRAM_ATTR xPortCanYield(void)
 {
@@ -484,22 +632,115 @@ static inline bool IRAM_ATTR xPortCanYield(void)
     return ((ps_reg & PS_INTLEVEL_MASK) == 0);
 }
 
-// porttrace
-#if configUSE_TRACE_FACILITY_2
-#include "porttrace.h"
+// ----------------------- System --------------------------
+
+static inline BaseType_t IRAM_ATTR xPortGetCoreID(void)
+{
+    return (uint32_t) cpu_hal_get_core_id();
+}
+
+
+
+/* ------------------------------------------------------ Misc ---------------------------------------------------------
+ * - Miscellaneous porting macros
+ * - These are not port of the FreeRTOS porting interface, but are used by other FreeRTOS dependent components
+ * - [refactor-todo] Remove dependency on MPU wrappers by modifying TCB
+ * ------------------------------------------------------------------------------------------------------------------ */
+
+// -------------------- Co-Processor -----------------------
+
+// When coprocessors are defined, we maintain a pointer to coprocessors area.
+// We currently use a hack: redefine field xMPU_SETTINGS in TCB block as a structure that can hold:
+// MPU wrappers, coprocessor area pointer, trace code structure, and more if needed.
+// The field is normally used for memory protection. FreeRTOS should create another general purpose field.
+typedef struct {
+#if XCHAL_CP_NUM > 0
+    volatile StackType_t *coproc_area; // Pointer to coprocessor save area; MUST BE FIRST
+#endif
+
+#if portUSING_MPU_WRAPPERS
+    // Define here mpu_settings, which is port dependent
+    int mpu_setting; // Just a dummy example here; MPU not ported to Xtensa yet
 #endif
+} xMPU_SETTINGS;
 
-// configASSERT_2 if requested
-#if configASSERT_2
-#include <stdio.h>
-void exit(int);
-#define configASSERT( x )   if (!(x)) { porttracePrint(-1); printf("\nAssertion failed in %s:%d\n", __FILE__, __LINE__); exit(-1); }
+// Main hack to use MPU_wrappers even when no MPU is defined (warning: mpu_setting should not be accessed; otherwise move this above xMPU_SETTINGS)
+#if (XCHAL_CP_NUM > 0) && !portUSING_MPU_WRAPPERS   // If MPU wrappers not used, we still need to allocate coproc area
+#undef portUSING_MPU_WRAPPERS
+#define portUSING_MPU_WRAPPERS 1   // Enable it to allocate coproc area
+#define MPU_WRAPPERS_H             // Override mpu_wrapper.h to disable unwanted code
+#define PRIVILEGED_FUNCTION
+#define PRIVILEGED_DATA
 #endif
 
-#endif // __ASSEMBLER__
+void _xt_coproc_release(volatile void *coproc_sa_base);
+
+/*
+ * The structures and methods of manipulating the MPU are contained within the
+ * port layer.
+ *
+ * Fills the xMPUSettings structure with the memory region information
+ * contained in xRegions.
+ */
+#if( portUSING_MPU_WRAPPERS == 1 )
+struct xMEMORY_REGION;
+void vPortStoreTaskMPUSettings( xMPU_SETTINGS *xMPUSettings, const struct xMEMORY_REGION *const xRegions, StackType_t *pxBottomOfStack, uint32_t usStackDepth ) PRIVILEGED_FUNCTION;
+void vPortReleaseTaskMPUSettings( xMPU_SETTINGS *xMPUSettings );
+#endif
+
+// -------------------- VA_ARGS Yield ----------------------
+
+/**
+ * Macro to count number of arguments of a __VA_ARGS__ used to support portYIELD_FROM_ISR with,
+ * or without arguments. The macro counts only 0 or 1 arguments.
+ *
+ * In the future, we want to switch to C++20. We also want to become compatible with clang.
+ * Hence, we provide two versions of the following macros which are using variadic arguments.
+ * The first one is using the GNU extension ##__VA_ARGS__. The second one is using the C++20 feature __VA_OPT__(,).
+ * This allows users to compile their code with standard C++20 enabled instead of the GNU extension.
+ * Below C++20, we haven't found any good alternative to using ##__VA_ARGS__.
+ */
+#if defined(__cplusplus) && (__cplusplus >  201703L)
+#define portGET_ARGUMENT_COUNT(...) portGET_ARGUMENT_COUNT_INNER(0 __VA_OPT__(,) __VA_ARGS__,1,0)
+#else
+#define portGET_ARGUMENT_COUNT(...) portGET_ARGUMENT_COUNT_INNER(0, ##__VA_ARGS__,1,0)
+#endif
+#define portGET_ARGUMENT_COUNT_INNER(zero, one, count, ...) count
+
+_Static_assert(portGET_ARGUMENT_COUNT() == 0, "portGET_ARGUMENT_COUNT() result does not match for 0 arguments");
+_Static_assert(portGET_ARGUMENT_COUNT(1) == 1, "portGET_ARGUMENT_COUNT() result does not match for 1 argument");
+
+// -------------------- Heap Related -----------------------
+
+/**
+ * @brief Checks if a given piece of memory can be used to store a task's TCB
+ *
+ * - Defined in port_common.c
+ *
+ * @param ptr Pointer to memory
+ * @return true Memory can be used to store a TCB
+ * @return false Otherwise
+ */
+bool xPortCheckValidTCBMem(const void *ptr);
+
+/**
+ * @brief Checks if a given piece of memory can be used to store a task's stack
+ *
+ * - Defined in port_common.c
+ *
+ * @param ptr Pointer to memory
+ * @return true Memory can be used to store a task stack
+ * @return false Otherwise
+ */
+bool xPortcheckValidStackMem(const void *ptr);
+
+#define portVALID_TCB_MEM(ptr) xPortCheckValidTCBMem(ptr)
+#define portVALID_STACK_MEM(ptr) xPortcheckValidStackMem(ptr)
 
 #ifdef __cplusplus
 }
 #endif
 
+#endif // __ASSEMBLER__
+
 #endif /* PORTMACRO_H */
diff --git a/tools/sdk/esp32s2/include/freertos/port/xtensa/include/freertos/portmacro_priv.h b/tools/sdk/esp32s2/include/freertos/port/xtensa/include/freertos/portmacro_priv.h
deleted file mode 100644
index 843456b5aa8..00000000000
--- a/tools/sdk/esp32s2/include/freertos/port/xtensa/include/freertos/portmacro_priv.h
+++ /dev/null
@@ -1,78 +0,0 @@
-/*
-    FreeRTOS V8.2.0 - Copyright (C) 2015 Real Time Engineers Ltd.
-    All rights reserved
-
-    VISIT http://www.FreeRTOS.org TO ENSURE YOU ARE USING THE LATEST VERSION.
-
-    ***************************************************************************
-     *                                                                       *
-     *    FreeRTOS provides completely free yet professionally developed,    *
-     *    robust, strictly quality controlled, supported, and cross          *
-     *    platform software that has become a de facto standard.             *
-     *                                                                       *
-     *    Help yourself get started quickly and support the FreeRTOS         *
-     *    project by purchasing a FreeRTOS tutorial book, reference          *
-     *    manual, or both from: http://www.FreeRTOS.org/Documentation        *
-     *                                                                       *
-     *    Thank you!                                                         *
-     *                                                                       *
-    ***************************************************************************
-
-    This file is part of the FreeRTOS distribution.
-
-    FreeRTOS is free software; you can redistribute it and/or modify it under
-    the terms of the GNU General Public License (version 2) as published by the
-    Free Software Foundation >>!AND MODIFIED BY!<< the FreeRTOS exception.
-
-    >>!   NOTE: The modification to the GPL is included to allow you to     !<<
-    >>!   distribute a combined work that includes FreeRTOS without being   !<<
-    >>!   obliged to provide the source code for proprietary components     !<<
-    >>!   outside of the FreeRTOS kernel.                                   !<<
-
-    FreeRTOS is distributed in the hope that it will be useful, but WITHOUT ANY
-    WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
-    FOR A PARTICULAR PURPOSE.  Full license text is available from the following
-    link: http://www.freertos.org/a00114.html
-
-    1 tab == 4 spaces!
-
-    ***************************************************************************
-     *                                                                       *
-     *    Having a problem?  Start by reading the FAQ "My application does   *
-     *    not run, what could be wrong?"                                     *
-     *                                                                       *
-     *    http://www.FreeRTOS.org/FAQHelp.html                               *
-     *                                                                       *
-    ***************************************************************************
-
-    http://www.FreeRTOS.org - Documentation, books, training, latest versions,
-    license and Real Time Engineers Ltd. contact details.
-
-    http://www.FreeRTOS.org/plus - A selection of FreeRTOS ecosystem products,
-    including FreeRTOS+Trace - an indispensable productivity tool, a DOS
-    compatible FAT file system, and our tiny thread aware UDP/IP stack.
-
-    http://www.OpenRTOS.com - Real Time Engineers ltd license FreeRTOS to High
-    Integrity Systems to sell under the OpenRTOS brand.  Low cost OpenRTOS
-    licenses offer ticketed support, indemnification and middleware.
-
-    http://www.SafeRTOS.com - High Integrity Systems also provide a safety
-    engineered and independently SIL3 certified version for use in safety and
-    mission critical applications that require provable dependability.
-
-    1 tab == 4 spaces!
-*/
-
-
-/* This header holds the macros for porting which should only be used inside FreeRTOS */
-
-#pragma once
-#include "soc/soc_memory_layout.h"
-
-//xTaskCreateStatic uses these functions to check incoming memory.
-#define portVALID_TCB_MEM(ptr) (esp_ptr_internal(ptr) && esp_ptr_byte_accessible(ptr))
-#ifdef CONFIG_SPIRAM_ALLOW_STACK_EXTERNAL_MEMORY
-#define portVALID_STACK_MEM(ptr) esp_ptr_byte_accessible(ptr)
-#else
-#define portVALID_STACK_MEM(ptr) (esp_ptr_internal(ptr) && esp_ptr_byte_accessible(ptr))
-#endif
diff --git a/tools/sdk/esp32s2/include/hal/esp32s2/include/hal/rmt_ll.h b/tools/sdk/esp32s2/include/hal/esp32s2/include/hal/rmt_ll.h
index 958fe7aef5d..ba73eaa2b75 100644
--- a/tools/sdk/esp32s2/include/hal/esp32s2/include/hal/rmt_ll.h
+++ b/tools/sdk/esp32s2/include/hal/esp32s2/include/hal/rmt_ll.h
@@ -1,16 +1,9 @@
-// Copyright 2019 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2019-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
 #pragma once
 
 #include <stdbool.h>
@@ -22,6 +15,8 @@
 extern "C" {
 #endif
 
+#define RMT_LL_MAX_LOOP_COUNT          (1023)/*!< Max loop count that hardware is supported */
+
 #define RMT_LL_HW_BASE  (&RMT)
 #define RMT_LL_MEM_BASE (&RMTMEM)
 
diff --git a/tools/sdk/esp32s2/include/hal/include/hal/i2s_hal.h b/tools/sdk/esp32s2/include/hal/include/hal/i2s_hal.h
index a08813db808..037970fa2b3 100644
--- a/tools/sdk/esp32s2/include/hal/include/hal/i2s_hal.h
+++ b/tools/sdk/esp32s2/include/hal/include/hal/i2s_hal.h
@@ -1,16 +1,8 @@
-// Copyright 2020 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2020-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 /*******************************************************************************
  * NOTICE
@@ -176,28 +168,28 @@ void i2s_hal_enable_slave_fd_mode(i2s_hal_context_t *hal);
  *
  * @param hal Context of the HAL layer
  */
-#define i2s_hal_start_tx(hal)               i2s_ll_tx_start((hal)->dev)
+void i2s_hal_start_tx(i2s_hal_context_t *hal);
 
 /**
  * @brief Start I2S rx
  *
  * @param hal Context of the HAL layer
  */
-#define i2s_hal_start_rx(hal)               i2s_ll_rx_start((hal)->dev)
+void i2s_hal_start_rx(i2s_hal_context_t *hal);
 
 /**
  * @brief Stop I2S tx
  *
  * @param hal Context of the HAL layer
  */
-#define i2s_hal_stop_tx(hal)                i2s_ll_tx_stop((hal)->dev)
+void i2s_hal_stop_tx(i2s_hal_context_t *hal);
 
 /**
  * @brief Stop I2S rx
  *
  * @param hal Context of the HAL layer
  */
-#define i2s_hal_stop_rx(hal)                i2s_ll_rx_stop((hal)->dev)
+void i2s_hal_stop_rx(i2s_hal_context_t *hal);
 
 /**
  * @brief Set the received data length to trigger `in_suc_eof` interrupt.
diff --git a/tools/sdk/esp32s2/include/hal/include/hal/lcd_hal.h b/tools/sdk/esp32s2/include/hal/include/hal/lcd_hal.h
index 76e28dda0e4..db255b3d1e4 100644
--- a/tools/sdk/esp32s2/include/hal/include/hal/lcd_hal.h
+++ b/tools/sdk/esp32s2/include/hal/include/hal/lcd_hal.h
@@ -1,22 +1,8 @@
-// Copyright 2021 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
-
-/*******************************************************************************
- * NOTICE
- * The HAL is not public api, don't use in application code.
- * See readme.md in soc/README.md
- ******************************************************************************/
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
diff --git a/tools/sdk/esp32s2/include/hal/include/hal/lcd_types.h b/tools/sdk/esp32s2/include/hal/include/hal/lcd_types.h
index 69dab14801d..01e6d0c2949 100644
--- a/tools/sdk/esp32s2/include/hal/include/hal/lcd_types.h
+++ b/tools/sdk/esp32s2/include/hal/include/hal/lcd_types.h
@@ -1,16 +1,8 @@
-// Copyright 2021 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
@@ -21,15 +13,12 @@ extern "C" {
 /**
  * @brief LCD clock source
  * @note User should select the clock source based on the real requirement:
- * ╔═════════════════════╦══════════════════════════╦════════════════════════════╗
- * ║ LCD clock source    ║ Features                 ║ Power Management           ║
- * ╠═════════════════════╬══════════════════════════╬════════════════════════════╣
- * ║ LCD_CLK_SRC_PLL160M ║ High resolution, fixed   ║ ESP_PM_APB_FREQ_MAX lock   ║
- * ╠═════════════════════╬══════════════════════════╬════════════════════════════╣
- * ║ LCD_CLK_SRC_APLL    ║ Configurable resolution  ║ ESP_PM_NO_LIGHT_SLEEP lock ║
- * ╠═════════════════════╬══════════════════════════╬════════════════════════════╣
- * ║ LCD_CLK_SRC_XTAL    ║ Medium resolution, fixed ║ No PM lock                 ║
- * ╚═════════════════════╩══════════════════════════╩════════════════════════════╝
+ *
+ * | LCD clock source    | Features                 | Power Management           |
+ * |---------------------|--------------------------|----------------------------|
+ * | LCD_CLK_SRC_PLL160M | High resolution, fixed   | ESP_PM_APB_FREQ_MAX lock   |
+ * | LCD_CLK_SRC_APLL    | Configurable resolution  | ESP_PM_NO_LIGHT_SLEEP lock |
+ * | LCD_CLK_SRC_XTAL    | Medium resolution, fixed | No PM lock                 |
  */
 typedef enum {
     LCD_CLK_SRC_PLL160M, /*!< Select PLL160M as the source clock */
diff --git a/tools/sdk/esp32s2/include/hal/include/hal/touch_sensor_types.h b/tools/sdk/esp32s2/include/hal/include/hal/touch_sensor_types.h
index ec027bf8705..9085f5eecd8 100644
--- a/tools/sdk/esp32s2/include/hal/include/hal/touch_sensor_types.h
+++ b/tools/sdk/esp32s2/include/hal/include/hal/touch_sensor_types.h
@@ -1,16 +1,8 @@
-// Copyright 2015-2020 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
@@ -155,12 +147,23 @@ typedef enum {
     TOUCH_PAD_INTR_MASK_INACTIVE = BIT(2),  /*!<Inactive for one of the enabled channels. */
     TOUCH_PAD_INTR_MASK_SCAN_DONE = BIT(3), /*!<Measurement done for all the enabled channels. */
     TOUCH_PAD_INTR_MASK_TIMEOUT = BIT(4),   /*!<Timeout for one of the enabled channels. */
+#if SOC_TOUCH_PROXIMITY_MEAS_DONE_SUPPORTED
+    TOUCH_PAD_INTR_MASK_PROXI_MEAS_DONE = BIT(5),   /*!<For proximity sensor, when the number of measurements reaches the set count of measurements, an interrupt will be generated. */
+    TOUCH_PAD_INTR_MASK_MAX
+#define TOUCH_PAD_INTR_MASK_ALL (TOUCH_PAD_INTR_MASK_TIMEOUT    \
+                                | TOUCH_PAD_INTR_MASK_SCAN_DONE \
+                                | TOUCH_PAD_INTR_MASK_INACTIVE  \
+                                | TOUCH_PAD_INTR_MASK_ACTIVE    \
+                                | TOUCH_PAD_INTR_MASK_DONE      \
+                                | TOUCH_PAD_INTR_MASK_PROXI_MEAS_DONE) /*!<All touch interrupt type enable. */
+#else
     TOUCH_PAD_INTR_MASK_MAX
 #define TOUCH_PAD_INTR_MASK_ALL (TOUCH_PAD_INTR_MASK_TIMEOUT    \
                                 | TOUCH_PAD_INTR_MASK_SCAN_DONE \
                                 | TOUCH_PAD_INTR_MASK_INACTIVE  \
                                 | TOUCH_PAD_INTR_MASK_ACTIVE    \
                                 | TOUCH_PAD_INTR_MASK_DONE) /*!<All touch interrupt type enable. */
+#endif
 } touch_pad_intr_mask_t;
 FLAG_ATTR(touch_pad_intr_mask_t)
 
diff --git a/tools/sdk/esp32s2/include/lwip/port/esp32/include/lwipopts.h b/tools/sdk/esp32s2/include/lwip/port/esp32/include/lwipopts.h
index 4bc2ad13960..b9cfe17aca9 100644
--- a/tools/sdk/esp32s2/include/lwip/port/esp32/include/lwipopts.h
+++ b/tools/sdk/esp32s2/include/lwip/port/esp32/include/lwipopts.h
@@ -266,6 +266,12 @@
 
 #endif
 
+/**
+ * CONFIG_LWIP_DHCP_OPTIONS_LEN: The total length of outgoing DHCP option msg. If you have many options
+ * and options value is too long, you can configure the length according to your requirements
+ */
+#define DHCP_OPTIONS_LEN                CONFIG_LWIP_DHCP_OPTIONS_LEN
+
 /*
    ------------------------------------
    ---------- AUTOIP options ----------
diff --git a/tools/sdk/esp32s2/include/mdns/include/mdns.h b/tools/sdk/esp32s2/include/mdns/include/mdns.h
index 4bb288d4174..35c8d538abc 100644
--- a/tools/sdk/esp32s2/include/mdns/include/mdns.h
+++ b/tools/sdk/esp32s2/include/mdns/include/mdns.h
@@ -187,7 +187,10 @@ esp_err_t mdns_instance_name_set(const char * instance_name);
  * @note The value length of txt items will be automatically decided by strlen
  *
  * @param  instance_name    instance name to set. If NULL,
- *                          global instance name or hostname will be used
+ *                          global instance name or hostname will be used.
+ *                          Note that MDNS_MULTIPLE_INSTANCE config option
+ *                          needs to be enabled for adding multiple instances
+ *                          with the same instance type.
  * @param  service_type     service type (_http, _ftp, etc)
  * @param  proto            service protocol (_tcp, _udp)
  * @param  port             service port
@@ -209,6 +212,9 @@ esp_err_t mdns_service_add(const char * instance_name, const char * service_type
  *
  * @param  instance_name    instance name to set. If NULL,
  *                          global instance name or hostname will be used
+ *                          Note that MDNS_MULTIPLE_INSTANCE config option
+ *                          needs to be enabled for adding multiple instances
+ *                          with the same instance type.
  * @param  service_type     service type (_http, _ftp, etc)
  * @param  proto            service protocol (_tcp, _udp)
  * @param  hostname         service hostname. If NULL, local hostname will be used.
@@ -238,6 +244,22 @@ esp_err_t mdns_service_add_for_host(const char * instance_name, const char * ser
  */
 bool mdns_service_exists(const char * service_type, const char * proto, const char * hostname);
 
+
+/**
+ * @brief  Check whether a service has been added.
+ *
+ * @param  instance         instance name
+ * @param  service_type     service type (_http, _ftp, etc)
+ * @param  proto            service protocol (_tcp, _udp)
+ * @param  hostname         service hostname. If NULL, checks for the local hostname.
+ *
+ * @return
+ *     - true   Correspondding service has been added.
+ *     - false  Service not found.
+ */
+bool mdns_service_exists_with_instance(const char *instance, const char *service_type, const char *proto,
+                                       const char *hostname);
+
 /**
  * @brief  Remove service from mDNS server
  *
diff --git a/tools/sdk/esp32s2/include/mqtt/esp-mqtt/include/mqtt_client.h b/tools/sdk/esp32s2/include/mqtt/esp-mqtt/include/mqtt_client.h
index 348d0729f1d..475581b50db 100644
--- a/tools/sdk/esp32s2/include/mqtt/esp-mqtt/include/mqtt_client.h
+++ b/tools/sdk/esp32s2/include/mqtt/esp-mqtt/include/mqtt_client.h
@@ -206,6 +206,7 @@ typedef struct {
     int network_timeout_ms;                 /*!< Abort network operation if it is not completed after this value, in milliseconds (defaults to 10s) */
     bool disable_keepalive;                 /*!< Set disable_keepalive=true to turn off keep-alive mechanism, false by default (keepalive is active by default). Note: setting the config value `keepalive` to `0` doesn't disable keepalive feature, but uses a default keepalive period */
     const char *path;                       /*!< Path in the URI*/
+    int message_retransmit_timeout;         /*!< timeout for retansmit of failded packet */
 } esp_mqtt_client_config_t;
 
 /**
@@ -375,6 +376,7 @@ esp_err_t esp_mqtt_client_destroy(esp_mqtt_client_handle_t client);
  * @param config    mqtt configuration structure
  *
  * @return ESP_ERR_NO_MEM if failed to allocate
+ *         ESP_ERR_INVALID_ARG if conflicts on transport configuration.
  *         ESP_OK on success
  */
 esp_err_t esp_mqtt_set_config(esp_mqtt_client_handle_t client, const esp_mqtt_client_config_t *config);
diff --git a/tools/sdk/esp32s2/include/openssl/include/openssl/ssl.h b/tools/sdk/esp32s2/include/openssl/include/openssl/ssl.h
index 4a3376c0dbf..9a33e0d33b2 100644
--- a/tools/sdk/esp32s2/include/openssl/include/openssl/ssl.h
+++ b/tools/sdk/esp32s2/include/openssl/include/openssl/ssl.h
@@ -1,16 +1,13 @@
-// Copyright 2015-2016 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2015-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+// Allow for this warning suppression only in IDF_CI_BUILD
+#if !defined(ESP_OPENSSL_SUPPRESS_LEGACY_WARNING) || !defined(IDF_CI_BUILD)
+#warning "OpenSSL component will be removed from ESP-IDF in v5.0, please use esp_tls instead"
+#endif
 
 #ifndef _SSL_H_
 #define _SSL_H_
diff --git a/tools/sdk/esp32s2/include/protocomm/include/transports/protocomm_ble.h b/tools/sdk/esp32s2/include/protocomm/include/transports/protocomm_ble.h
index a9e7b6a209e..f899f95b188 100644
--- a/tools/sdk/esp32s2/include/protocomm/include/transports/protocomm_ble.h
+++ b/tools/sdk/esp32s2/include/protocomm/include/transports/protocomm_ble.h
@@ -1,16 +1,8 @@
-// Copyright 2018 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2018-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
@@ -86,6 +78,10 @@ typedef struct protocomm_ble_config {
      * Pointer to the Name-UUID lookup table
      */
     protocomm_ble_name_uuid_t *nu_lookup;
+
+    /* BLE bonding */
+    unsigned ble_bonding:1;
+
 } protocomm_ble_config_t;
 
 /**
diff --git a/tools/sdk/esp32s2/include/soc/esp32s2/include/soc/rtc.h b/tools/sdk/esp32s2/include/soc/esp32s2/include/soc/rtc.h
index 1877bb1cfe8..11b13fb6aaf 100644
--- a/tools/sdk/esp32s2/include/soc/esp32s2/include/soc/rtc.h
+++ b/tools/sdk/esp32s2/include/soc/esp32s2/include/soc/rtc.h
@@ -120,11 +120,14 @@ extern "C" {
 set sleep_init default param
 */
 #define RTC_CNTL_DBG_ATTEN_LIGHTSLEEP_DEFAULT  6
+#define RTC_CNTL_DBG_ATTEN_LIGHTSLEEP_NODROP  0
 #define RTC_CNTL_DBG_ATTEN_DEEPSLEEP_DEFAULT  15
 #define RTC_CNTL_DBG_ATTEN_MONITOR_DEFAULT  0
 #define RTC_CNTL_BIASSLP_MONITOR_DEFAULT  0
+#define RTC_CNTL_BIASSLP_SLEEP_ON  0
 #define RTC_CNTL_BIASSLP_SLEEP_DEFAULT  1
 #define RTC_CNTL_PD_CUR_MONITOR_DEFAULT  1
+#define RTC_CNTL_PD_CUR_SLEEP_ON  0
 #define RTC_CNTL_PD_CUR_SLEEP_DEFAULT  1
 
 #define APLL_SDM_STOP_VAL_1         0x09
@@ -658,6 +661,7 @@ typedef struct {
     uint32_t rtc_slowmem_pd_en : 1;     //!< power down RTC slow memory
     uint32_t rtc_peri_pd_en : 1;        //!< power down RTC peripherals
     uint32_t wifi_pd_en : 1;            //!< power down WiFi
+    uint32_t int_8m_pd_en : 1;          //!< Power down Internal 8M oscillator
     uint32_t deep_slp : 1;              //!< power down digital domain
     uint32_t wdt_flashboot_mod_en : 1;  //!< enable WDT flashboot mode
     uint32_t dig_dbias_wak : 3;         //!< set bias for digital domain, in active mode
@@ -665,6 +669,7 @@ typedef struct {
     uint32_t rtc_dbias_wak : 3;         //!< set bias for RTC domain, in active mode
     uint32_t rtc_dbias_slp : 3;         //!< set bias for RTC domain, in sleep mode
     uint32_t vddsdio_pd_en : 1;         //!< power down VDDSDIO regulator
+    uint32_t xtal_fpu : 1;              //!< keep main XTAL powered up in sleep
     uint32_t deep_slp_reject : 1;
     uint32_t light_slp_reject : 1;
 } rtc_sleep_config_t;
@@ -677,6 +682,7 @@ typedef struct {
  *
  * @param RTC_SLEEP_PD_x flags combined using bitwise OR
  */
+#define is_dslp(pd_flags)   ((pd_flags) & RTC_SLEEP_PD_DIG)
 #define RTC_SLEEP_CONFIG_DEFAULT(sleep_flags) { \
     .lslp_mem_inf_fpu = 0, \
     .rtc_mem_inf_follow_cpu = ((sleep_flags) & RTC_SLEEP_PD_RTC_MEM_FOLLOW_CPU) ? 1 : 0, \
@@ -684,13 +690,21 @@ typedef struct {
     .rtc_slowmem_pd_en = ((sleep_flags) & RTC_SLEEP_PD_RTC_SLOW_MEM) ? 1 : 0, \
     .rtc_peri_pd_en = ((sleep_flags) & RTC_SLEEP_PD_RTC_PERIPH) ? 1 : 0, \
     .wifi_pd_en = ((sleep_flags) & RTC_SLEEP_PD_WIFI) ? 1 : 0, \
+    .int_8m_pd_en = is_dslp(sleep_flags) ? 1 : ((sleep_flags) & RTC_SLEEP_PD_INT_8M) ? 1 : 0, \
     .deep_slp = ((sleep_flags) & RTC_SLEEP_PD_DIG) ? 1 : 0, \
     .wdt_flashboot_mod_en = 0, \
     .dig_dbias_wak = RTC_CNTL_DIG_DBIAS_1V10, \
-    .dig_dbias_slp = RTC_CNTL_DIG_DBIAS_0V90, \
+    .dig_dbias_slp = is_dslp(sleep_flags)                   ? RTC_CNTL_DIG_DBIAS_0V90 \
+                   : !((sleep_flags) & RTC_SLEEP_PD_INT_8M) ? RTC_CNTL_DIG_DBIAS_1V10 \
+                   : !((sleep_flags) & RTC_SLEEP_PD_XTAL)   ? RTC_CNTL_DIG_DBIAS_1V10 \
+                   : RTC_CNTL_DIG_DBIAS_0V90, \
     .rtc_dbias_wak = RTC_CNTL_DBIAS_1V10, \
-    .rtc_dbias_slp = RTC_CNTL_DBIAS_1V00, \
+    .rtc_dbias_slp = is_dslp(sleep_flags)                   ? RTC_CNTL_DBIAS_1V00 \
+                   : !((sleep_flags) & RTC_SLEEP_PD_INT_8M) ? RTC_CNTL_DBIAS_1V10 \
+                   : !((sleep_flags) & RTC_SLEEP_PD_XTAL)   ? RTC_CNTL_DBIAS_1V10 \
+                   : RTC_CNTL_DBIAS_1V00, \
     .vddsdio_pd_en = ((sleep_flags) & RTC_SLEEP_PD_VDDSDIO) ? 1 : 0, \
+    .xtal_fpu = is_dslp(sleep_flags) ? 0 : ((sleep_flags) & RTC_SLEEP_PD_XTAL) ? 0 : 1, \
     .deep_slp_reject = 1, \
     .light_slp_reject = 1 \
 };
@@ -702,6 +716,8 @@ typedef struct {
 #define RTC_SLEEP_PD_RTC_MEM_FOLLOW_CPU BIT(4)  //!< RTC FAST and SLOW memories are automatically powered up and down along with the CPU
 #define RTC_SLEEP_PD_VDDSDIO            BIT(5)  //!< Power down VDDSDIO regulator
 #define RTC_SLEEP_PD_WIFI               BIT(6)
+#define RTC_SLEEP_PD_INT_8M             BIT(7)  //!< Power down Internal 8M oscillator
+#define RTC_SLEEP_PD_XTAL               BIT(8)  //!< Power down main XTAL
 
 /**
  * @brief Prepare the chip to enter sleep mode
diff --git a/tools/sdk/esp32s2/include/soc/esp32s2/include/soc/soc_caps.h b/tools/sdk/esp32s2/include/soc/esp32s2/include/soc/soc_caps.h
index b33af28f369..620c5dc9ac8 100644
--- a/tools/sdk/esp32s2/include/soc/esp32s2/include/soc/soc_caps.h
+++ b/tools/sdk/esp32s2/include/soc/esp32s2/include/soc/soc_caps.h
@@ -1,16 +1,8 @@
-// Copyright 2020 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2020-2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 /*
  * Soc capabilities file, describing the following chip attributes:
@@ -242,12 +234,12 @@
 #define SOC_TIMER_GROUP_TOTAL_TIMERS (SOC_TIMER_GROUPS * SOC_TIMER_GROUP_TIMERS_PER_GROUP)
 
 /*-------------------------- TOUCH SENSOR CAPS -------------------------------*/
-#define SOC_TOUCH_SENSOR_NUM                (15)    /*! 15 Touch channels */
-#define SOC_TOUCH_PROXIMITY_CHANNEL_NUM     (3)  /* Sopport touch proximity channel number. */
+#define SOC_TOUCH_VERSION_2                 (1)     /*!<Hardware version of touch sensor */
+#define SOC_TOUCH_SENSOR_NUM                (15)    /*!<15 Touch channels */
+#define SOC_TOUCH_PROXIMITY_CHANNEL_NUM     (3)     /* Sopport touch proximity channel number. */
 
 #define SOC_TOUCH_PAD_THRESHOLD_MAX         (0x1FFFFF)  /*!<If set touch threshold max value, The touch sensor can't be in touched status */
 #define SOC_TOUCH_PAD_MEASURE_WAIT_MAX      (0xFF)  /*!<The timer frequency is 8Mhz, the max value is 0xff */
-#define SOC_TOUCH_PAD_WAKE_SUPPORTED        (1)     /*!<Supports waking up from touch pad trigger */
 
 /*-------------------------- TWAI CAPS ---------------------------------------*/
 #define SOC_TWAI_BRP_MIN                2
@@ -338,5 +330,9 @@
 
 #define SOC_PM_SUPPORT_WIFI_PD          (1)
 
+#define SOC_PM_SUPPORT_TOUCH_SENSOR_WAKEUP        (1)     /*!<Supports waking up from touch pad trigger */
+
+/*-------------------------- COEXISTENCE HARDWARE PTI CAPS -------------------------------*/
+#define SOC_COEX_HW_PTI                 (1)
 /* ---------------------------- Compatibility ------------------------------- */
 // No contents
diff --git a/tools/sdk/esp32s2/include/soc/include/soc/lcd_periph.h b/tools/sdk/esp32s2/include/soc/include/soc/lcd_periph.h
index d7d3ebc68bf..7bb7c530cf2 100644
--- a/tools/sdk/esp32s2/include/soc/include/soc/lcd_periph.h
+++ b/tools/sdk/esp32s2/include/soc/include/soc/lcd_periph.h
@@ -1,16 +1,8 @@
-// Copyright 2021 Espressif Systems (Shanghai) PTE LTD
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
 
 #pragma once
 
diff --git a/tools/sdk/esp32s2/include/spi_flash/include/esp_partition.h b/tools/sdk/esp32s2/include/spi_flash/include/esp_partition.h
index 36baea29860..550bf6b6b26 100644
--- a/tools/sdk/esp32s2/include/spi_flash/include/esp_partition.h
+++ b/tools/sdk/esp32s2/include/spi_flash/include/esp_partition.h
@@ -181,7 +181,9 @@ esp_partition_iterator_t esp_partition_next(esp_partition_iterator_t iterator);
 /**
  * @brief Release partition iterator
  *
- * @param iterator Iterator obtained using esp_partition_find. Must be non-NULL.
+ * @param iterator Iterator obtained using esp_partition_find.
+ *                 The iterator is allowed to be NULL, so it is not necessary to check its value
+ *                 before calling this function.
  *
  */
 void esp_partition_iterator_release(esp_partition_iterator_t iterator);
diff --git a/tools/sdk/esp32c3/include/spi_flash/include/spi_flash_private.h b/tools/sdk/esp32s2/include/spi_flash/include/esp_private/spi_flash_os.h
similarity index 88%
rename from tools/sdk/esp32c3/include/spi_flash/include/spi_flash_private.h
rename to tools/sdk/esp32s2/include/spi_flash/include/esp_private/spi_flash_os.h
index 9d822fb9475..e302df86ea2 100644
--- a/tools/sdk/esp32c3/include/spi_flash/include/spi_flash_private.h
+++ b/tools/sdk/esp32s2/include/spi_flash/include/esp_private/spi_flash_os.h
@@ -4,7 +4,9 @@
  * SPDX-License-Identifier: Apache-2.0
  */
 
-
+/**
+ * System level MSPI APIs (private)
+ */
 /**
  * Currently the MSPI timing tuning related APIs are designed to be private.
  * Because:
@@ -17,6 +19,10 @@
  */
 #pragma once
 
+#include <stdint.h>
+#include <stdbool.h>
+#include "sdkconfig.h"
+#include "esp_err.h"
 #if CONFIG_IDF_TARGET_ESP32
 #include "esp32/rom/spi_flash.h"
 #elif CONFIG_IDF_TARGET_ESP32S2
@@ -34,9 +40,9 @@ extern "C" {
 #endif
 
 /**
- * @brief Register ROM functions and init flash device registers to make use of octal flash
+ * @brief To setup Flash chip
  */
-esp_err_t esp_opiflash_init(void);
+esp_err_t spi_flash_init_chip_state(void);
 
 /**
  * @brief Make MSPI work under 20Mhz
@@ -88,6 +94,12 @@ void spi_timing_get_flash_timing_param(spi_flash_hal_timing_config_t *out_timing
  */
 bool spi_timine_config_flash_is_tuned(void);
 
+/**
+ * @brief Set Flash chip specifically required MSPI register settings here
+ */
+void spi_flash_set_vendor_required_regs(void);
+
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/tools/sdk/esp32s2/include/wpa_supplicant/esp_supplicant/include/esp_mbo.h b/tools/sdk/esp32s2/include/wpa_supplicant/esp_supplicant/include/esp_mbo.h
new file mode 100644
index 00000000000..4292213943f
--- /dev/null
+++ b/tools/sdk/esp32s2/include/wpa_supplicant/esp_supplicant/include/esp_mbo.h
@@ -0,0 +1,64 @@
+/*
+ * SPDX-FileCopyrightText: 2021 Espressif Systems (Shanghai) CO LTD
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+#ifndef _ESP_MBO_H
+#define _ESP_MBO_H
+
+#include <stdbool.h>
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/**
+  * enum non_pref_chan_reason: Reason for non preference of channel
+  */
+enum non_pref_chan_reason {
+	NON_PREF_CHAN_REASON_UNSPECIFIED = 0,
+	NON_PREF_CHAN_REASON_RSSI = 1,
+	NON_PREF_CHAN_REASON_EXT_INTERFERENCE = 2,
+	NON_PREF_CHAN_REASON_INT_INTERFERENCE = 3,
+};
+
+/**
+  * @brief  Channel structure for non preferred channel
+  *
+  * @param  reason: enum non_pref_chan_reason
+  * @param  oper_class: operating class for the channel
+  * @param  chan: channel number
+  * @param  preference: channel preference
+  */
+struct non_pref_chan {
+	enum non_pref_chan_reason reason;
+	uint8_t oper_class;
+	uint8_t chan;
+	uint8_t preference;
+};
+
+/**
+  * @brief  Array structure for non preferred channel struct
+  *
+  * @param  non_pref_chan_num: channel count
+  * @param  chan: array of non_pref_chan type
+  */
+struct non_pref_chan_s {
+	size_t non_pref_chan_num;
+	struct non_pref_chan chan[];
+};
+
+/**
+  * @brief  Update channel preference for MBO IE
+  *
+  * @param  non_pref_chan: Non preference channel list
+  *
+  * @return
+  *    - 0: success else failure
+  */
+int esp_mbo_update_non_pref_chan(struct non_pref_chan_s *non_pref_chan);
+
+#ifdef __cplusplus
+}
+#endif
+#endif
diff --git a/tools/sdk/esp32s2/include/wpa_supplicant/esp_supplicant/include/esp_wnm.h b/tools/sdk/esp32s2/include/wpa_supplicant/esp_supplicant/include/esp_wnm.h
index a1dcfb655c5..4301385d2cf 100644
--- a/tools/sdk/esp32s2/include/wpa_supplicant/esp_supplicant/include/esp_wnm.h
+++ b/tools/sdk/esp32s2/include/wpa_supplicant/esp_supplicant/include/esp_wnm.h
@@ -1,17 +1,7 @@
-/**
- * Copyright 2020 Espressif Systems (Shanghai) PTE LTD
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *   http://www.apache.org/licenses/LICENSE-2.0
+/*
+ * SPDX-FileCopyrightText: 2020-2021 Espressif Systems (Shanghai) CO LTD
  *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
+ * SPDX-License-Identifier: Apache-2.0
  */
 
 #ifndef _ESP_WNM_H
@@ -29,11 +19,13 @@ enum btm_query_reason {
 	REASON_UNSPECIFIED = 0,
 	REASON_FRAME_LOSS = 1,
 	REASON_DELAY = 2,
-	REASON_QOS_CAPACITY = 3,
-	REASON_FIRST_ASSOC = 4,
-	REASON_LOAD_BALALNCE = 5,
-	REASON_BETTER_AP = 6,
-	REASON_CURRENT_DEAUTH = 7,
+	REASON_BANDWIDTH = 3,
+	REASON_LOAD_BALANCE = 4,
+	REASON_RSSI = 5,
+	REASON_RETRANSMISSIONS = 6,
+	REASON_INTERFERENCE = 7,
+	REASON_GRAY_ZONE = 8,
+	REASON_PREMIUM_AP = 9,
 };
 
 /**
diff --git a/tools/sdk/esp32s2/ld/libcat_face_detect.a b/tools/sdk/esp32s2/ld/libcat_face_detect.a
index 1bc5162b09f..98527ca429a 100644
Binary files a/tools/sdk/esp32s2/ld/libcat_face_detect.a and b/tools/sdk/esp32s2/ld/libcat_face_detect.a differ
diff --git a/tools/sdk/esp32s2/ld/libdl.a b/tools/sdk/esp32s2/ld/libdl.a
index 1e6c8c48421..efe06e32edb 100644
Binary files a/tools/sdk/esp32s2/ld/libdl.a and b/tools/sdk/esp32s2/ld/libdl.a differ
diff --git a/tools/sdk/esp32s2/ld/libhuman_face_detect.a b/tools/sdk/esp32s2/ld/libhuman_face_detect.a
index 72fd6f5eb49..2e04b6d2a42 100644
Binary files a/tools/sdk/esp32s2/ld/libhuman_face_detect.a and b/tools/sdk/esp32s2/ld/libhuman_face_detect.a differ
diff --git a/tools/sdk/esp32s2/ld/sections.ld b/tools/sdk/esp32s2/ld/sections.ld
index 94bf03d6a9f..cf784476531 100644
--- a/tools/sdk/esp32s2/ld/sections.ld
+++ b/tools/sdk/esp32s2/ld/sections.ld
@@ -205,7 +205,7 @@ SECTIONS
     *libesp_system.a:ubsan.*(.literal .literal.* .text .text.*)
     *libfreertos.a:(EXCLUDE_FILE(*libfreertos.a:port.* *libfreertos.a:port_common.*) .literal EXCLUDE_FILE(*libfreertos.a:port.* *libfreertos.a:port_common.*) .literal.* EXCLUDE_FILE(*libfreertos.a:port.* *libfreertos.a:port_common.*) .text EXCLUDE_FILE(*libfreertos.a:port.* *libfreertos.a:port_common.*) .text.*)
     *libfreertos.a:port.*(.literal.pxPortInitialiseStack .literal.unlikely.vPortEndScheduler .literal.vApplicationStackOverflowHook .literal.vPortAssertIfInISR .literal.vPortEnterCritical .literal.vPortExitCritical .literal.vPortSetStackWatchpoint .literal.vPortYieldOtherCore .literal.xPortInIsrContext .literal.xPortStartScheduler .text .text.pxPortInitialiseStack .text.unlikely.vPortEndScheduler .text.vApplicationStackOverflowHook .text.vPortAssertIfInISR .text.vPortEnterCritical .text.vPortExitCritical .text.vPortSetStackWatchpoint .text.vPortYieldOtherCore .text.xPortGetTickRateHz .text.xPortInIsrContext .text.xPortStartScheduler)
-    *libfreertos.a:port_common.*(.literal.esp_startup_start_app_common .text .text.esp_startup_start_app_common)
+    *libfreertos.a:port_common.*(.literal.esp_startup_start_app_common .literal.xPortCheckValidTCBMem .literal.xPortcheckValidStackMem .text .text.esp_startup_start_app_common .text.xPortCheckValidTCBMem .text.xPortcheckValidStackMem)
     *libgcc.a:_divsf3.*(.literal .literal.* .text .text.*)
     *libgcc.a:lib2funcs.*(.literal .literal.* .text .text.*)
     *libgcov.a:(.literal .literal.* .text .text.*)
@@ -371,8 +371,8 @@ SECTIONS
     *(.ext_ram.bss*)
 
     *(.bss .bss.*)
-    *(.dynbss .dynsbss .gnu.linkonce.b .gnu.linkonce.b.* .gnu.linkonce.sb .gnu.linkonce.sb.* .gnu.linkonce.sb2 .gnu.linkonce.sb2.* .sbss .sbss.* .sbss2 .sbss2.* .scommon .share.mem)
     *(.ext_ram.bss .ext_ram.bss.*)
+    *(.dynbss .dynsbss .gnu.linkonce.b .gnu.linkonce.b.* .gnu.linkonce.sb .gnu.linkonce.sb.* .gnu.linkonce.sb2 .gnu.linkonce.sb2.* .sbss .sbss.* .sbss2 .sbss2.* .scommon .share.mem)
     *(COMMON)
     _bt_bss_start = ABSOLUTE(.);
     *libbt.a:(.bss .bss.* COMMON)
diff --git a/tools/sdk/esp32s2/lib/libapp_trace.a b/tools/sdk/esp32s2/lib/libapp_trace.a
index 65b1f96e83b..ac3a85ea72f 100644
Binary files a/tools/sdk/esp32s2/lib/libapp_trace.a and b/tools/sdk/esp32s2/lib/libapp_trace.a differ
diff --git a/tools/sdk/esp32s2/lib/libapp_update.a b/tools/sdk/esp32s2/lib/libapp_update.a
index 36db9c85c17..38f7d7e635e 100644
Binary files a/tools/sdk/esp32s2/lib/libapp_update.a and b/tools/sdk/esp32s2/lib/libapp_update.a differ
diff --git a/tools/sdk/esp32s2/lib/libarduino_tinyusb.a b/tools/sdk/esp32s2/lib/libarduino_tinyusb.a
index e004a2db069..184eb360c23 100644
Binary files a/tools/sdk/esp32s2/lib/libarduino_tinyusb.a and b/tools/sdk/esp32s2/lib/libarduino_tinyusb.a differ
diff --git a/tools/sdk/esp32s2/lib/libasio.a b/tools/sdk/esp32s2/lib/libasio.a
index 3a23faf503f..b1dc1b216ee 100644
Binary files a/tools/sdk/esp32s2/lib/libasio.a and b/tools/sdk/esp32s2/lib/libasio.a differ
diff --git a/tools/sdk/esp32s2/lib/libbootloader_support.a b/tools/sdk/esp32s2/lib/libbootloader_support.a
index ffa6ff3b490..22f8033876c 100644
Binary files a/tools/sdk/esp32s2/lib/libbootloader_support.a and b/tools/sdk/esp32s2/lib/libbootloader_support.a differ
diff --git a/tools/sdk/esp32s2/lib/libcoap.a b/tools/sdk/esp32s2/lib/libcoap.a
index 6a5173eb19d..a0093c629d2 100644
Binary files a/tools/sdk/esp32s2/lib/libcoap.a and b/tools/sdk/esp32s2/lib/libcoap.a differ
diff --git a/tools/sdk/esp32s2/lib/libcoexist.a b/tools/sdk/esp32s2/lib/libcoexist.a
index 0f146c2c899..cfc9be30fc2 100644
Binary files a/tools/sdk/esp32s2/lib/libcoexist.a and b/tools/sdk/esp32s2/lib/libcoexist.a differ
diff --git a/tools/sdk/esp32s2/lib/libconsole.a b/tools/sdk/esp32s2/lib/libconsole.a
index 3f95dd5e32d..aa0ca4a1a40 100644
Binary files a/tools/sdk/esp32s2/lib/libconsole.a and b/tools/sdk/esp32s2/lib/libconsole.a differ
diff --git a/tools/sdk/esp32s2/lib/libcore.a b/tools/sdk/esp32s2/lib/libcore.a
index e017a1006a3..0ea53525c5b 100644
Binary files a/tools/sdk/esp32s2/lib/libcore.a and b/tools/sdk/esp32s2/lib/libcore.a differ
diff --git a/tools/sdk/esp32s2/lib/libcxx.a b/tools/sdk/esp32s2/lib/libcxx.a
index 9202ad02b03..99c85f299ad 100644
Binary files a/tools/sdk/esp32s2/lib/libcxx.a and b/tools/sdk/esp32s2/lib/libcxx.a differ
diff --git a/tools/sdk/esp32s2/lib/libdriver.a b/tools/sdk/esp32s2/lib/libdriver.a
index 271f07c3d00..3f7b5b90508 100644
Binary files a/tools/sdk/esp32s2/lib/libdriver.a and b/tools/sdk/esp32s2/lib/libdriver.a differ
diff --git a/tools/sdk/esp32s2/lib/libefuse.a b/tools/sdk/esp32s2/lib/libefuse.a
index 45e85b5b281..994398fcf35 100644
Binary files a/tools/sdk/esp32s2/lib/libefuse.a and b/tools/sdk/esp32s2/lib/libefuse.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp-tls.a b/tools/sdk/esp32s2/lib/libesp-tls.a
index 0fbddfec7bd..8c8e740c521 100644
Binary files a/tools/sdk/esp32s2/lib/libesp-tls.a and b/tools/sdk/esp32s2/lib/libesp-tls.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp32-camera.a b/tools/sdk/esp32s2/lib/libesp32-camera.a
index b6dff3e52a3..6bd84c93d7a 100644
Binary files a/tools/sdk/esp32s2/lib/libesp32-camera.a and b/tools/sdk/esp32s2/lib/libesp32-camera.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_common.a b/tools/sdk/esp32s2/lib/libesp_common.a
index 3d55f543596..adbc942d3e1 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_common.a and b/tools/sdk/esp32s2/lib/libesp_common.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_eth.a b/tools/sdk/esp32s2/lib/libesp_eth.a
index 5a3c5d1ec1f..d8752b6e178 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_eth.a and b/tools/sdk/esp32s2/lib/libesp_eth.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_event.a b/tools/sdk/esp32s2/lib/libesp_event.a
index 3ba390e2ef4..b5e063cbfd9 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_event.a and b/tools/sdk/esp32s2/lib/libesp_event.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_gdbstub.a b/tools/sdk/esp32s2/lib/libesp_gdbstub.a
index 2432b1aeced..1ba4aa80d12 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_gdbstub.a and b/tools/sdk/esp32s2/lib/libesp_gdbstub.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_hid.a b/tools/sdk/esp32s2/lib/libesp_hid.a
index 82ea9ff4cbb..a0753a2ec8a 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_hid.a and b/tools/sdk/esp32s2/lib/libesp_hid.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_http_client.a b/tools/sdk/esp32s2/lib/libesp_http_client.a
index 557786f2310..598dda23545 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_http_client.a and b/tools/sdk/esp32s2/lib/libesp_http_client.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_http_server.a b/tools/sdk/esp32s2/lib/libesp_http_server.a
index 7dcb8e9bdb5..6e7ef78c017 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_http_server.a and b/tools/sdk/esp32s2/lib/libesp_http_server.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_https_ota.a b/tools/sdk/esp32s2/lib/libesp_https_ota.a
index 31fadcfb272..9b721d649eb 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_https_ota.a and b/tools/sdk/esp32s2/lib/libesp_https_ota.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_https_server.a b/tools/sdk/esp32s2/lib/libesp_https_server.a
index fe3cf5581d8..89979aa87ca 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_https_server.a and b/tools/sdk/esp32s2/lib/libesp_https_server.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_hw_support.a b/tools/sdk/esp32s2/lib/libesp_hw_support.a
index 2dacebef4e3..f2d2229e29b 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_hw_support.a and b/tools/sdk/esp32s2/lib/libesp_hw_support.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_ipc.a b/tools/sdk/esp32s2/lib/libesp_ipc.a
index a846df70856..a5245e30694 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_ipc.a and b/tools/sdk/esp32s2/lib/libesp_ipc.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_lcd.a b/tools/sdk/esp32s2/lib/libesp_lcd.a
index 4bcdbecb1ca..0c3a9418984 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_lcd.a and b/tools/sdk/esp32s2/lib/libesp_lcd.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_littlefs.a b/tools/sdk/esp32s2/lib/libesp_littlefs.a
index 16cb7dea9b6..66b7958bcad 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_littlefs.a and b/tools/sdk/esp32s2/lib/libesp_littlefs.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_local_ctrl.a b/tools/sdk/esp32s2/lib/libesp_local_ctrl.a
index fabc0fe34f0..e92c6014500 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_local_ctrl.a and b/tools/sdk/esp32s2/lib/libesp_local_ctrl.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_netif.a b/tools/sdk/esp32s2/lib/libesp_netif.a
index e30fa30a9b5..3f3fd194596 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_netif.a and b/tools/sdk/esp32s2/lib/libesp_netif.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_phy.a b/tools/sdk/esp32s2/lib/libesp_phy.a
index 10c41de3f84..fbb2b1f358a 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_phy.a and b/tools/sdk/esp32s2/lib/libesp_phy.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_pm.a b/tools/sdk/esp32s2/lib/libesp_pm.a
index ffa717deb5d..f156c9906df 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_pm.a and b/tools/sdk/esp32s2/lib/libesp_pm.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_ringbuf.a b/tools/sdk/esp32s2/lib/libesp_ringbuf.a
index 46a0abb1b61..fbbd39850e8 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_ringbuf.a and b/tools/sdk/esp32s2/lib/libesp_ringbuf.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_serial_slave_link.a b/tools/sdk/esp32s2/lib/libesp_serial_slave_link.a
index eb9d9aa7d49..79368ffd203 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_serial_slave_link.a and b/tools/sdk/esp32s2/lib/libesp_serial_slave_link.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_system.a b/tools/sdk/esp32s2/lib/libesp_system.a
index 338875d18ca..b97bf823467 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_system.a and b/tools/sdk/esp32s2/lib/libesp_system.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_timer.a b/tools/sdk/esp32s2/lib/libesp_timer.a
index c869d63ec01..604761d623e 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_timer.a and b/tools/sdk/esp32s2/lib/libesp_timer.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_websocket_client.a b/tools/sdk/esp32s2/lib/libesp_websocket_client.a
index 86d2344ffa3..d1d4cce573d 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_websocket_client.a and b/tools/sdk/esp32s2/lib/libesp_websocket_client.a differ
diff --git a/tools/sdk/esp32s2/lib/libesp_wifi.a b/tools/sdk/esp32s2/lib/libesp_wifi.a
index 8b21c2fc5a0..9ecdc54b417 100644
Binary files a/tools/sdk/esp32s2/lib/libesp_wifi.a and b/tools/sdk/esp32s2/lib/libesp_wifi.a differ
diff --git a/tools/sdk/esp32s2/lib/libespcoredump.a b/tools/sdk/esp32s2/lib/libespcoredump.a
index 25b0cc93cad..dbff1ec0f16 100644
Binary files a/tools/sdk/esp32s2/lib/libespcoredump.a and b/tools/sdk/esp32s2/lib/libespcoredump.a differ
diff --git a/tools/sdk/esp32s2/lib/libespnow.a b/tools/sdk/esp32s2/lib/libespnow.a
index debe3c57249..1c9c4851315 100644
Binary files a/tools/sdk/esp32s2/lib/libespnow.a and b/tools/sdk/esp32s2/lib/libespnow.a differ
diff --git a/tools/sdk/esp32s2/lib/libfatfs.a b/tools/sdk/esp32s2/lib/libfatfs.a
index 9a65e421336..355140be2b9 100644
Binary files a/tools/sdk/esp32s2/lib/libfatfs.a and b/tools/sdk/esp32s2/lib/libfatfs.a differ
diff --git a/tools/sdk/esp32s2/lib/libfreemodbus.a b/tools/sdk/esp32s2/lib/libfreemodbus.a
index 50511f3c123..b9e0fca5abb 100644
Binary files a/tools/sdk/esp32s2/lib/libfreemodbus.a and b/tools/sdk/esp32s2/lib/libfreemodbus.a differ
diff --git a/tools/sdk/esp32s2/lib/libfreertos.a b/tools/sdk/esp32s2/lib/libfreertos.a
index 733f6a3ee3a..ee14407fac1 100644
Binary files a/tools/sdk/esp32s2/lib/libfreertos.a and b/tools/sdk/esp32s2/lib/libfreertos.a differ
diff --git a/tools/sdk/esp32s2/lib/libhal.a b/tools/sdk/esp32s2/lib/libhal.a
index c13b15b5ff4..6ce696e646a 100644
Binary files a/tools/sdk/esp32s2/lib/libhal.a and b/tools/sdk/esp32s2/lib/libhal.a differ
diff --git a/tools/sdk/esp32s2/lib/libheap.a b/tools/sdk/esp32s2/lib/libheap.a
index 7ed09f63f9c..673282ace1a 100644
Binary files a/tools/sdk/esp32s2/lib/libheap.a and b/tools/sdk/esp32s2/lib/libheap.a differ
diff --git a/tools/sdk/esp32s2/lib/liblog.a b/tools/sdk/esp32s2/lib/liblog.a
index cd6ad238403..4abce7dcab0 100644
Binary files a/tools/sdk/esp32s2/lib/liblog.a and b/tools/sdk/esp32s2/lib/liblog.a differ
diff --git a/tools/sdk/esp32s2/lib/liblwip.a b/tools/sdk/esp32s2/lib/liblwip.a
index 84c9a7cbc7c..6fb9552493a 100644
Binary files a/tools/sdk/esp32s2/lib/liblwip.a and b/tools/sdk/esp32s2/lib/liblwip.a differ
diff --git a/tools/sdk/esp32s2/lib/libmbedcrypto.a b/tools/sdk/esp32s2/lib/libmbedcrypto.a
index c11c0cb50c2..f23f00429ed 100644
Binary files a/tools/sdk/esp32s2/lib/libmbedcrypto.a and b/tools/sdk/esp32s2/lib/libmbedcrypto.a differ
diff --git a/tools/sdk/esp32s2/lib/libmbedtls.a b/tools/sdk/esp32s2/lib/libmbedtls.a
index c69d43d797a..88c24274720 100644
Binary files a/tools/sdk/esp32s2/lib/libmbedtls.a and b/tools/sdk/esp32s2/lib/libmbedtls.a differ
diff --git a/tools/sdk/esp32s2/lib/libmdns.a b/tools/sdk/esp32s2/lib/libmdns.a
index 863f659c447..caa89e815b2 100644
Binary files a/tools/sdk/esp32s2/lib/libmdns.a and b/tools/sdk/esp32s2/lib/libmdns.a differ
diff --git a/tools/sdk/esp32s2/lib/libmesh.a b/tools/sdk/esp32s2/lib/libmesh.a
index 8f4fbde53cd..4a43e09b1c2 100644
Binary files a/tools/sdk/esp32s2/lib/libmesh.a and b/tools/sdk/esp32s2/lib/libmesh.a differ
diff --git a/tools/sdk/esp32s2/lib/libmqtt.a b/tools/sdk/esp32s2/lib/libmqtt.a
index 56d63e439e0..12e59e01c3f 100644
Binary files a/tools/sdk/esp32s2/lib/libmqtt.a and b/tools/sdk/esp32s2/lib/libmqtt.a differ
diff --git a/tools/sdk/esp32s2/lib/libnet80211.a b/tools/sdk/esp32s2/lib/libnet80211.a
index 93351e991c8..69800846e43 100644
Binary files a/tools/sdk/esp32s2/lib/libnet80211.a and b/tools/sdk/esp32s2/lib/libnet80211.a differ
diff --git a/tools/sdk/esp32s2/lib/libnewlib.a b/tools/sdk/esp32s2/lib/libnewlib.a
index 1f724357d57..a9b0adb894b 100644
Binary files a/tools/sdk/esp32s2/lib/libnewlib.a and b/tools/sdk/esp32s2/lib/libnewlib.a differ
diff --git a/tools/sdk/esp32s2/lib/libnvs_flash.a b/tools/sdk/esp32s2/lib/libnvs_flash.a
index a75e4c5782c..7c3ecb087bf 100644
Binary files a/tools/sdk/esp32s2/lib/libnvs_flash.a and b/tools/sdk/esp32s2/lib/libnvs_flash.a differ
diff --git a/tools/sdk/esp32s2/lib/libopenssl.a b/tools/sdk/esp32s2/lib/libopenssl.a
index 0c02d23769a..ae0543e096c 100644
Binary files a/tools/sdk/esp32s2/lib/libopenssl.a and b/tools/sdk/esp32s2/lib/libopenssl.a differ
diff --git a/tools/sdk/esp32s2/lib/libpp.a b/tools/sdk/esp32s2/lib/libpp.a
index 3aa7310ee69..f6b7fc13fee 100644
Binary files a/tools/sdk/esp32s2/lib/libpp.a and b/tools/sdk/esp32s2/lib/libpp.a differ
diff --git a/tools/sdk/esp32s2/lib/libprotocomm.a b/tools/sdk/esp32s2/lib/libprotocomm.a
index 42d5c8885f5..7d5dcb48c40 100644
Binary files a/tools/sdk/esp32s2/lib/libprotocomm.a and b/tools/sdk/esp32s2/lib/libprotocomm.a differ
diff --git a/tools/sdk/esp32s2/lib/libpthread.a b/tools/sdk/esp32s2/lib/libpthread.a
index 9bbcb0516fd..58b2c22ac9e 100644
Binary files a/tools/sdk/esp32s2/lib/libpthread.a and b/tools/sdk/esp32s2/lib/libpthread.a differ
diff --git a/tools/sdk/esp32s2/lib/libsdmmc.a b/tools/sdk/esp32s2/lib/libsdmmc.a
index ba4320bc74d..d94647ebb3e 100644
Binary files a/tools/sdk/esp32s2/lib/libsdmmc.a and b/tools/sdk/esp32s2/lib/libsdmmc.a differ
diff --git a/tools/sdk/esp32s2/lib/libsmartconfig.a b/tools/sdk/esp32s2/lib/libsmartconfig.a
index 29134ba9039..d1b81cbf1ea 100644
Binary files a/tools/sdk/esp32s2/lib/libsmartconfig.a and b/tools/sdk/esp32s2/lib/libsmartconfig.a differ
diff --git a/tools/sdk/esp32s2/lib/libsoc.a b/tools/sdk/esp32s2/lib/libsoc.a
index 57560c731e3..ad9543d0bf5 100644
Binary files a/tools/sdk/esp32s2/lib/libsoc.a and b/tools/sdk/esp32s2/lib/libsoc.a differ
diff --git a/tools/sdk/esp32s2/lib/libspi_flash.a b/tools/sdk/esp32s2/lib/libspi_flash.a
index 7d19d683fd9..a5db338b9a2 100644
Binary files a/tools/sdk/esp32s2/lib/libspi_flash.a and b/tools/sdk/esp32s2/lib/libspi_flash.a differ
diff --git a/tools/sdk/esp32s2/lib/libspiffs.a b/tools/sdk/esp32s2/lib/libspiffs.a
index eb8c65fa83b..79b1af01df6 100644
Binary files a/tools/sdk/esp32s2/lib/libspiffs.a and b/tools/sdk/esp32s2/lib/libspiffs.a differ
diff --git a/tools/sdk/esp32s2/lib/libtcp_transport.a b/tools/sdk/esp32s2/lib/libtcp_transport.a
index 993d07d6661..b844a197f2a 100644
Binary files a/tools/sdk/esp32s2/lib/libtcp_transport.a and b/tools/sdk/esp32s2/lib/libtcp_transport.a differ
diff --git a/tools/sdk/esp32s2/lib/libtcpip_adapter.a b/tools/sdk/esp32s2/lib/libtcpip_adapter.a
index 6c20dd87b02..6215e82396e 100644
Binary files a/tools/sdk/esp32s2/lib/libtcpip_adapter.a and b/tools/sdk/esp32s2/lib/libtcpip_adapter.a differ
diff --git a/tools/sdk/esp32s2/lib/libtouch_element.a b/tools/sdk/esp32s2/lib/libtouch_element.a
index 8720a0c1b40..0bb6a1cf082 100644
Binary files a/tools/sdk/esp32s2/lib/libtouch_element.a and b/tools/sdk/esp32s2/lib/libtouch_element.a differ
diff --git a/tools/sdk/esp32s2/lib/libulp.a b/tools/sdk/esp32s2/lib/libulp.a
index b69794e91a2..5b910490675 100644
Binary files a/tools/sdk/esp32s2/lib/libulp.a and b/tools/sdk/esp32s2/lib/libulp.a differ
diff --git a/tools/sdk/esp32s2/lib/libusb.a b/tools/sdk/esp32s2/lib/libusb.a
index c6004586a2f..a3a69f8c0e2 100644
Binary files a/tools/sdk/esp32s2/lib/libusb.a and b/tools/sdk/esp32s2/lib/libusb.a differ
diff --git a/tools/sdk/esp32s2/lib/libvfs.a b/tools/sdk/esp32s2/lib/libvfs.a
index 94902934bd2..6d837302c88 100644
Binary files a/tools/sdk/esp32s2/lib/libvfs.a and b/tools/sdk/esp32s2/lib/libvfs.a differ
diff --git a/tools/sdk/esp32s2/lib/libwapi.a b/tools/sdk/esp32s2/lib/libwapi.a
index 91bddc368c8..8914130654f 100644
Binary files a/tools/sdk/esp32s2/lib/libwapi.a and b/tools/sdk/esp32s2/lib/libwapi.a differ
diff --git a/tools/sdk/esp32s2/lib/libwear_levelling.a b/tools/sdk/esp32s2/lib/libwear_levelling.a
index f27dd314042..2731e946188 100644
Binary files a/tools/sdk/esp32s2/lib/libwear_levelling.a and b/tools/sdk/esp32s2/lib/libwear_levelling.a differ
diff --git a/tools/sdk/esp32s2/lib/libwifi_provisioning.a b/tools/sdk/esp32s2/lib/libwifi_provisioning.a
index 40d6d4fc7f7..b2316845a13 100644
Binary files a/tools/sdk/esp32s2/lib/libwifi_provisioning.a and b/tools/sdk/esp32s2/lib/libwifi_provisioning.a differ
diff --git a/tools/sdk/esp32s2/lib/libwpa_supplicant.a b/tools/sdk/esp32s2/lib/libwpa_supplicant.a
index 5afc5fd0100..40da40383dd 100644
Binary files a/tools/sdk/esp32s2/lib/libwpa_supplicant.a and b/tools/sdk/esp32s2/lib/libwpa_supplicant.a differ
diff --git a/tools/sdk/esp32s2/lib/libxtensa.a b/tools/sdk/esp32s2/lib/libxtensa.a
index c7ff5169849..5d9d565b6fd 100644
Binary files a/tools/sdk/esp32s2/lib/libxtensa.a and b/tools/sdk/esp32s2/lib/libxtensa.a differ
diff --git a/tools/sdk/esp32s2/sdkconfig b/tools/sdk/esp32s2/sdkconfig
index be0df95cf98..240cd9d014c 100644
--- a/tools/sdk/esp32s2/sdkconfig
+++ b/tools/sdk/esp32s2/sdkconfig
@@ -93,6 +93,7 @@ CONFIG_ESPTOOLPY_FLASHMODE_QIO=y
 # CONFIG_ESPTOOLPY_FLASHMODE_QOUT is not set
 # CONFIG_ESPTOOLPY_FLASHMODE_DIO is not set
 # CONFIG_ESPTOOLPY_FLASHMODE_DOUT is not set
+CONFIG_ESPTOOLPY_FLASH_SAMPLE_MODE_STR=y
 CONFIG_ESPTOOLPY_FLASHMODE="dio"
 # CONFIG_ESPTOOLPY_FLASHFREQ_120M is not set
 CONFIG_ESPTOOLPY_FLASHFREQ_80M=y
@@ -203,7 +204,7 @@ CONFIG_TINYUSB_CDC_TX_BUFSIZE=64
 #
 CONFIG_TINYUSB_MSC_ENABLED=y
 CONFIG_TINYUSB_DESC_MSC_STRING="Espressif MSC Device"
-CONFIG_TINYUSB_MSC_BUFSIZE=512
+CONFIG_TINYUSB_MSC_BUFSIZE=4096
 # end of Mass Storage (MSC) driver
 
 #
@@ -223,6 +224,15 @@ CONFIG_TINYUSB_MIDI_RX_BUFSIZE=64
 CONFIG_TINYUSB_MIDI_TX_BUFSIZE=64
 # end of MIDI driver
 
+#
+# VIDEO driver
+#
+CONFIG_TINYUSB_VIDEO_ENABLED=y
+CONFIG_TINYUSB_DESC_VIDEO_STRING="Espressif VIDEO Device"
+CONFIG_TINYUSB_VIDEO_STREAMING_BUFSIZE=64
+CONFIG_TINYUSB_VIDEO_STREAMING_IFS=1
+# end of VIDEO driver
+
 #
 # DFU Runtime driver
 #
@@ -658,6 +668,7 @@ CONFIG_ESP_WIFI_FTM_ENABLE=y
 CONFIG_ESP_WIFI_FTM_INITIATOR_SUPPORT=y
 CONFIG_ESP_WIFI_FTM_RESPONDER_SUPPORT=y
 # CONFIG_ESP_WIFI_STA_DISCONNECTED_PM_ENABLE is not set
+# CONFIG_ESP_WIFI_EXTERNAL_COEXIST_ENABLE is not set
 # end of Wi-Fi
 
 #
@@ -867,6 +878,7 @@ CONFIG_LWIP_TCPIP_RECVMBOX_SIZE=32
 # CONFIG_LWIP_DHCP_DOES_ARP_CHECK is not set
 # CONFIG_LWIP_DHCP_DISABLE_CLIENT_ID is not set
 CONFIG_LWIP_DHCP_RESTORE_LAST_IP=y
+CONFIG_LWIP_DHCP_OPTIONS_LEN=68
 
 #
 # DHCP server
@@ -1123,6 +1135,7 @@ CONFIG_MDNS_SERVICE_ADD_TIMEOUT_MS=2000
 # CONFIG_MDNS_STRICT_MODE is not set
 CONFIG_MDNS_TIMER_PERIOD_MS=100
 # CONFIG_MDNS_NETWORKING_SOCKET is not set
+CONFIG_MDNS_MULTIPLE_INSTANCE=y
 # end of mDNS
 
 #
@@ -1426,7 +1439,7 @@ CONFIG_USB_CDC_RX_BUFSIZE=64
 CONFIG_USB_CDC_TX_BUFSIZE=64
 CONFIG_USB_MSC_ENABLED=y
 CONFIG_USB_DESC_MSC_STRING="Espressif MSC Device"
-CONFIG_USB_MSC_BUFSIZE=512
+CONFIG_USB_MSC_BUFSIZE=4096
 CONFIG_USB_DESC_HID_STRING="Espressif HID Device"
 CONFIG_USB_DEBUG_LEVEL=0
 # CONFIG_COMPILER_OPTIMIZATION_LEVEL_DEBUG is not set
@@ -1475,6 +1488,7 @@ CONFIG_TASK_WDT_PANIC=y
 CONFIG_TASK_WDT_TIMEOUT_S=5
 # CONFIG_TASK_WDT_CHECK_IDLE_TASK_CPU0 is not set
 CONFIG_TIMER_TASK_STACK_SIZE=4096
+# CONFIG_EXTERNAL_COEX_ENABLE is not set
 # CONFIG_ESP32_ENABLE_COREDUMP_TO_FLASH is not set
 # CONFIG_ESP32_ENABLE_COREDUMP_TO_UART is not set
 CONFIG_ESP32_ENABLE_COREDUMP_TO_NONE=y
@@ -1539,7 +1553,7 @@ CONFIG_USB_DESC_CDC_STRING="Espressif CDC Device"
 CONFIG_USB_DESC_MSC_STRING="Espressif MSC Device"
 CONFIG_USB_DESC_HID_STRING="Espressif HID Device"
 CONFIG_USB_MSC_ENABLED=y
-CONFIG_USB_MSC_BUFSIZE=512
+CONFIG_USB_MSC_BUFSIZE=4096
 CONFIG_USB_CDC_ENABLED=y
 CONFIG_USB_CDC_RX_BUFSIZE=64
 CONFIG_USB_CDC_TX_BUFSIZE=64