Add new SIMD multiplication instructions to C/JS API (#3782)

Adds C/JS APIs for the SIMD instructions
* Q15MulrSatSVecI16x8
* ExtMulLowSVecI16x8
* ExtMulHighSVecI16x8
* ExtMulLowUVecI16x8
* ExtMulHighUVecI16x8
* ExtMulLowSVecI32x4
* ExtMulHighSVecI32x4
* ExtMulLowUVecI32x4
* ExtMulHighUVecI32x4
* ExtMulLowSVecI64x2
* ExtMulHighSVecI64x2
* ExtMulLowUVecI64x2
* ExtMulHighUVecI64x2

2 files changed, 169 insertions, 0 deletions

diff --git a/test/binaryen.js/kitchen-sink.js b/test/binaryen.js/kitchen-sink.js
index 7a191a90d..78fa53e22 100644
--- a/test/binaryen.js/kitchen-sink.js
+++ b/test/binaryen.js/kitchen-sink.js
@@ -423,6 +423,11 @@ function test_core() {
     module.i16x8.max_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.i16x8.max_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.i16x8.avgr_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i16x8.q15mulr_sat_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i16x8.extmul_low_i8x16_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i16x8.extmul_high_i8x16_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i16x8.extmul_low_i8x16_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i16x8.extmul_high_i8x16_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.i32x4.add(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.i32x4.sub(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.i32x4.mul(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
@@ -431,9 +436,17 @@ function test_core() {
     module.i32x4.max_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.i32x4.max_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.i32x4.dot_i16x8_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i32x4.extmul_low_i16x8_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i32x4.extmul_high_i16x8_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i32x4.extmul_low_i16x8_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i32x4.extmul_high_i16x8_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.i64x2.add(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.i64x2.sub(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.i64x2.mul(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i64x2.extmul_low_i32x4_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i64x2.extmul_high_i32x4_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i64x2.extmul_low_i32x4_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
+    module.i64x2.extmul_high_i32x4_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.f32x4.add(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.f32x4.sub(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
     module.f32x4.mul(module.v128.const(v128_bytes), module.v128.const(v128_bytes)),
diff --git a/test/binaryen.js/kitchen-sink.js.txt b/test/binaryen.js/kitchen-sink.js.txt
index f9c4bc5a9..51db40ade 100644
--- a/test/binaryen.js/kitchen-sink.js.txt
+++ b/test/binaryen.js/kitchen-sink.js.txt
@@ -1294,6 +1294,36 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7}
        )
       )
       (drop
+       (i16x8.q15mulr_sat_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i16x8.extmul_low_i8x16_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i16x8.extmul_high_i8x16_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i16x8.extmul_low_i8x16_u
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i16x8.extmul_high_i8x16_u
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
        (i32x4.add
         (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
         (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
@@ -1342,6 +1372,30 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7}
        )
       )
       (drop
+       (i32x4.extmul_low_i16x8_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i32x4.extmul_high_i16x8_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i32x4.extmul_low_i16x8_u
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i32x4.extmul_high_i16x8_u
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
        (i64x2.add
         (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
         (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
@@ -1360,6 +1414,30 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7}
        )
       )
       (drop
+       (i64x2.extmul_low_i32x4_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i64x2.extmul_high_i32x4_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i64x2.extmul_low_i32x4_u
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i64x2.extmul_high_i32x4_u
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
        (f32x4.add
         (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
         (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
@@ -3218,6 +3296,36 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7}
        )
       )
       (drop
+       (i16x8.q15mulr_sat_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i16x8.extmul_low_i8x16_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i16x8.extmul_high_i8x16_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i16x8.extmul_low_i8x16_u
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i16x8.extmul_high_i8x16_u
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
        (i32x4.add
         (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
         (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
@@ -3266,6 +3374,30 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7}
        )
       )
       (drop
+       (i32x4.extmul_low_i16x8_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i32x4.extmul_high_i16x8_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i32x4.extmul_low_i16x8_u
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i32x4.extmul_high_i16x8_u
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
        (i64x2.add
         (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
         (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
@@ -3284,6 +3416,30 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7}
        )
       )
       (drop
+       (i64x2.extmul_low_i32x4_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i64x2.extmul_high_i32x4_s
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i64x2.extmul_low_i32x4_u
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
+       (i64x2.extmul_high_i32x4_u
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+        (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
+       )
+      )
+      (drop
        (f32x4.add
         (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)
         (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)