[wasm-simd][ia32] Implement v128.load32_zero v128.load64_zero

Prototype these two instructions on ia32. They are movss and movsd
respectively, so the implementation is pretty simple, as we support
these instructions already.

Bug: v8:11038
Change-Id: Iebf4afab2bf1edfb4b14a4855d5036677f999ca9
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2486232Reviewed-by: Bill Budge <bbudge@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#70634}

src/compiler/backend/ia32/instruction-selector-ia32.cc

@@ -396,6 +396,12 @@ void InstructionSelector::VisitLoadTransform(Node* node) {
     case LoadTransformation::kS128Load32x2U:
       opcode = kIA32S128Load32x2U;
       break;
+    case LoadTransformation::kS128Load32Zero:
+      opcode = kIA32Movss;
+      break;
+    case LoadTransformation::kS128Load64Zero:
+      opcode = kIA32Movsd;
+      break;
     default:
       UNREACHABLE();
   }

test/cctest/wasm/test-run-wasm-simd.cc

@@ -3697,7 +3697,7 @@ WASM_SIMD_TEST(S128Load32x2S) {
 }
 
 // TODO(v8:10713): Prototyping v128.load32_zero and v128.load64_zero.
-#if V8_TARGET_ARCH_X64 || V8_TARGET_ARCH_ARM64
+#if V8_TARGET_ARCH_X64 || V8_TARGET_ARCH_ARM64 || V8_TARGET_ARCH_IA32
 template <typename S>
 void RunLoadZeroTest(TestExecutionTier execution_tier, LowerSimd lower_simd,
                      WasmOpcode op) {
@@ -3729,7 +3729,7 @@ WASM_SIMD_TEST_NO_LOWERING(S128Load32Zero) {
 WASM_SIMD_TEST_NO_LOWERING(S128Load64Zero) {
   RunLoadZeroTest<int64_t>(execution_tier, lower_simd, kExprS128Load64Zero);
 }
-#endif  // V8_TARGET_ARCH_X64 || V8_TARGET_ARCH_ARM64
+#endif  // V8_TARGET_ARCH_X64 || V8_TARGET_ARCH_ARM64 || V8_TARGET_ARCH_IA32
 
 #if V8_TARGET_ARCH_X64
 // TODO(v8:10975): Prototyping load lane and store lane.