Add several SIMD opcodes to IA32

src/compiler/ia32/code-generator-ia32.cc

Index: src/compiler/ia32/code-generator-ia32.cc
diff --git a/src/compiler/ia32/code-generator-ia32.cc b/src/compiler/ia32/code-generator-ia32.cc
index 8188a92c1e8cd4ee049cdef15f2612e4e1f8869d..8bc4ab9efe3574ab67a5a37489f02356ae40dbcc 100644
--- a/src/compiler/ia32/code-generator-ia32.cc
+++ b/src/compiler/ia32/code-generator-ia32.cc
@@ -1901,6 +1901,30 @@ CodeGenerator::CodeGenResult CodeGenerator::AssembleArchInstruction(
       __ xchg(i.InputRegister(index), operand);
       break;
     }
+    case kIA32Int32x4Create: {
+      XMMRegister dst = i.OutputSimd128Register();
+      __ movd(dst, i.InputOperand(0));
+      __ shufps(dst, dst, 0x0);
+      break;
+    }
+    case kIA32Int32x4ExtractLane: {
+      CpuFeatureScope sse_scope(masm(), SSE4_1);
+      __ Pextrd(i.OutputRegister(), i.InputSimd128Register(0), i.InputInt8(1));
+      break;
+    }
+    case kIA32Int32x4ReplaceLane: {
+      CpuFeatureScope sse_scope(masm(), SSE4_1);
+      __ Pinsrd(i.OutputSimd128Register(), i.InputOperand(2), i.InputInt8(1));
+      break;
+    }
+    case kIA32Int32x4Add: {
+      __ paddd(i.OutputSimd128Register(), i.InputSimd128Register(1));

[bbudge, 2017/02/15 16:07:31]

If it's planned to use AVX, then we should make these macro instructions and use
the macroassembler to switch between implementations.

[Jing, 2017/02/16 05:11:17]

To add AVX version ,we do have some code needed to add/change. For example, all
other movds should also be changed to Movds.
I plan to supplement the SSE instructions as well as their AVX version in next
patch. Is it OK to change the instructions to macro instructions at that time?

[bbudge, 2017/02/17 00:44:39]

OK.

+      break;
+    }
+    case kIA32Int32x4Sub: {
+      __ psubd(i.OutputSimd128Register(), i.InputSimd128Register(1));
+      break;
+    }
     case kCheckedLoadInt8:
       ASSEMBLE_CHECKED_LOAD_INTEGER(movsx_b);
       break;