Auto merge of #82183 - michaelwoerister:lazier-defpathhash-loading2, r=wesleywiser

Simplify lazy DefPathHash decoding by using an on-disk hash table.

This PR simplifies the logic around mapping `DefPathHash` values encountered during incremental compilation to valid `DefId`s in the current session. It is able to do so by using an on-disk hash table encoding that allows for looking up values directly, i.e. without deserializing the entire table.

The main simplification comes from not having to keep track of `DefPathHashes` being used during the compilation session.

Author: bors

Cargo.lock

@@ -2321,6 +2321,15 @@ dependencies = [
  "rustc-std-workspace-core",
 ]
 
+[[package]]
+name = "odht"
+version = "0.2.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0b18a8d1c919d3e7b5c49708d08ef7d60bc2150a7c3a8244257c54ca3f625010"
+dependencies = [
+ "cfg-if 1.0.0",
+]
+
 [[package]]
 name = "once_cell"
 version = "1.7.2"
@@ -3859,6 +3868,7 @@ version = "0.0.0"
 name = "rustc_hir"
 version = "0.0.0"
 dependencies = [
+ "odht",
  "rustc_ast",
  "rustc_data_structures",
  "rustc_feature",
@@ -4045,6 +4055,7 @@ name = "rustc_metadata"
 version = "0.0.0"
 dependencies = [
  "libc",
+ "odht",
  "rustc_ast",
  "rustc_attr",
  "rustc_data_structures",

compiler/rustc_hir/Cargo.toml

@@ -17,3 +17,4 @@ rustc_serialize = { path = "../rustc_serialize" }
 rustc_ast = { path = "../rustc_ast" }
 tracing = "0.1"
 smallvec = { version = "1.6.1", features = ["union", "may_dangle"] }
+odht = { version = "0.2.1", features = ["nightly"] }

compiler/rustc_hir/src/def_path_hash_map.rs

@@ -0,0 +1,37 @@
+use rustc_data_structures::fingerprint::Fingerprint;
+use rustc_span::def_id::{DefIndex, DefPathHash};
+
+#[derive(Clone, Default)]
+pub struct Config;
+
+impl odht::Config for Config {
+    type Key = DefPathHash;
+    type Value = DefIndex;
+
+    type EncodedKey = [u8; 16];
+    type EncodedValue = [u8; 4];
+
+    type H = odht::UnHashFn;
+
+    #[inline]
+    fn encode_key(k: &DefPathHash) -> [u8; 16] {
+        k.0.to_le_bytes()
+    }
+
+    #[inline]
+    fn encode_value(v: &DefIndex) -> [u8; 4] {
+        v.as_u32().to_le_bytes()
+    }
+
+    #[inline]
+    fn decode_key(k: &[u8; 16]) -> DefPathHash {
+        DefPathHash(Fingerprint::from_le_bytes(*k))
+    }
+
+    #[inline]
+    fn decode_value(v: &[u8; 4]) -> DefIndex {
+        DefIndex::from_u32(u32::from_le_bytes(*v))
+    }
+}
+
+pub type DefPathHashMap = odht::HashTableOwned<Config>;

compiler/rustc_hir/src/definitions.rs

@@ -6,11 +6,11 @@
 
 pub use crate::def_id::DefPathHash;
 use crate::def_id::{CrateNum, DefIndex, LocalDefId, StableCrateId, CRATE_DEF_INDEX, LOCAL_CRATE};
+use crate::def_path_hash_map::DefPathHashMap;
 use crate::hir;
 
 use rustc_data_structures::fx::FxHashMap;
 use rustc_data_structures::stable_hasher::StableHasher;
-use rustc_data_structures::unhash::UnhashMap;
 use rustc_index::vec::IndexVec;
 use rustc_span::hygiene::ExpnId;
 use rustc_span::symbol::{kw, sym, Symbol};
@@ -28,7 +28,7 @@ use tracing::debug;
 pub struct DefPathTable {
     index_to_key: IndexVec<DefIndex, DefKey>,
     def_path_hashes: IndexVec<DefIndex, DefPathHash>,
-    def_path_hash_to_index: UnhashMap<DefPathHash, DefIndex>,
+    def_path_hash_to_index: DefPathHashMap,
 }
 
 impl DefPathTable {
@@ -44,7 +44,7 @@ impl DefPathTable {
 
         // Check for hash collisions of DefPathHashes. These should be
         // exceedingly rare.
-        if let Some(existing) = self.def_path_hash_to_index.insert(def_path_hash, index) {
+        if let Some(existing) = self.def_path_hash_to_index.insert(&def_path_hash, &index) {
             let def_path1 = DefPath::make(LOCAL_CRATE, existing, |idx| self.def_key(idx));
             let def_path2 = DefPath::make(LOCAL_CRATE, index, |idx| self.def_key(idx));
 
@@ -87,7 +87,7 @@ impl DefPathTable {
 
     pub fn enumerated_keys_and_path_hashes(
         &self,
-    ) -> impl Iterator<Item = (DefIndex, &DefKey, &DefPathHash)> + '_ {
+    ) -> impl Iterator<Item = (DefIndex, &DefKey, &DefPathHash)> + ExactSizeIterator + '_ {
         self.index_to_key
             .iter_enumerated()
             .map(move |(index, key)| (index, key, &self.def_path_hashes[index]))
@@ -110,6 +110,9 @@ pub struct Definitions {
     expansions_that_defined: FxHashMap<LocalDefId, ExpnId>,
 
     def_id_to_span: IndexVec<LocalDefId, Span>,
+
+    /// The [StableCrateId] of the local crate.
+    stable_crate_id: StableCrateId,
 }
 
 /// A unique identifier that we can use to lookup a definition
@@ -356,6 +359,7 @@ impl Definitions {
             hir_id_to_def_id: Default::default(),
             expansions_that_defined: Default::default(),
             def_id_to_span,
+            stable_crate_id,
         }
     }
 
@@ -439,11 +443,17 @@ impl Definitions {
     }
 
     #[inline(always)]
-    pub fn local_def_path_hash_to_def_id(&self, hash: DefPathHash) -> Option<LocalDefId> {
+    pub fn local_def_path_hash_to_def_id(&self, hash: DefPathHash) -> LocalDefId {
+        debug_assert!(hash.stable_crate_id() == self.stable_crate_id);
         self.table
             .def_path_hash_to_index
             .get(&hash)
-            .map(|&local_def_index| LocalDefId { local_def_index })
+            .map(|local_def_index| LocalDefId { local_def_index })
+            .unwrap()
+    }
+
+    pub fn def_path_hash_to_def_index_map(&self) -> &DefPathHashMap {
+        &self.table.def_path_hash_to_index
     }
 }
 

compiler/rustc_hir/src/lib.rs

@@ -16,6 +16,7 @@ extern crate rustc_data_structures;
 
 mod arena;
 pub mod def;
+pub mod def_path_hash_map;
 pub mod definitions;
 pub use rustc_span::def_id;
 mod hir;

compiler/rustc_metadata/Cargo.toml

@@ -8,6 +8,7 @@ doctest = false
 
 [dependencies]
 libc = "0.2"
+odht = { version = "0.2.1", features = ["nightly"] }
 snap = "1"
 tracing = "0.1"
 smallvec = { version = "1.6.1", features = ["union", "may_dangle"] }

compiler/rustc_metadata/src/creader.rs

@@ -45,7 +45,7 @@ pub struct CStore {
 
     /// This map is used to verify we get no hash conflicts between
     /// `StableCrateId` values.
-    stable_crate_ids: FxHashMap<StableCrateId, CrateNum>,
+    pub(crate) stable_crate_ids: FxHashMap<StableCrateId, CrateNum>,
 
     /// Unused externs of the crate
     unused_externs: Vec<Symbol>,