Merge pull request rust-lang#54 from AdamNiederer/vmul

Add vmulpd and vmulps

Author: alexcrichton

src/x86/avx.rs

@@ -18,6 +18,23 @@ pub fn _mm256_add_ps(a: f32x8, b: f32x8) -> f32x8 {
     a + b
 }
 
+/// Add packed double-precision (64-bit) floating-point elements
+/// in `a` and `b`.
+#[inline(always)]
+#[target_feature = "+avx"]
+#[cfg_attr(test, assert_instr(vmulpd))]
+pub fn _mm256_mul_pd(a: f64x4, b: f64x4) -> f64x4 {
+    a * b
+}
+
+/// Add packed single-precision (32-bit) floating-point elements in `a` and `b`.
+#[inline(always)]
+#[target_feature = "+avx"]
+#[cfg_attr(test, assert_instr(vmulps))]
+pub fn _mm256_mul_ps(a: f32x8, b: f32x8) -> f32x8 {
+    a * b
+}
+
 /// Alternatively add and subtract packed double-precision (64-bit)
 /// floating-point elements in `a` to/from packed elements in `b`.
 #[inline(always)]
@@ -122,6 +139,24 @@ mod tests {
         assert_eq!(r, e);
     }
 
+    #[simd_test = "avx"]
+    fn _mm256_mul_pd() {
+        let a = f64x4::new(1.0, 2.0, 3.0, 4.0);
+        let b = f64x4::new(5.0, 6.0, 7.0, 8.0);
+        let r = avx::_mm256_mul_pd(a, b);
+        let e = f64x4::new(5.0, 12.0, 21.0, 32.0);
+        assert_eq!(r, e);
+    }
+
+    #[simd_test = "avx"]
+    fn _mm256_mul_ps() {
+        let a = f32x8::new(1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0);
+        let b = f32x8::new(9.0, 10.0, 11.0, 12.0, 13.0, 14.0, 15.0, 16.0);
+        let r = avx::_mm256_mul_ps(a, b);
+        let e = f32x8::new(9.0, 20.0, 33.0, 48.0, 65.0, 84.0, 105.0, 128.0);
+        assert_eq!(r, e);
+    }
+
     #[simd_test = "avx"]
     fn _mm256_addsub_pd() {
         let a = f64x4::new(1.0, 2.0, 3.0, 4.0);