Experimental implementation: Exposing SIMD instructions into JavaScript

src/ia32/macro-assembler-ia32.cc

Index: src/ia32/macro-assembler-ia32.cc
diff --git a/src/ia32/macro-assembler-ia32.cc b/src/ia32/macro-assembler-ia32.cc
index 2d0720a77d208716b300c8a4346295bb3571a403..9a15eac53aa4fdcba8fcc06481cfdbe8a87576ef 100644
--- a/src/ia32/macro-assembler-ia32.cc
+++ b/src/ia32/macro-assembler-ia32.cc
@@ -1120,12 +1120,13 @@ void MacroAssembler::EnterExitFrameEpilogue(int argc, bool save_doubles) {
   // Optionally save all XMM registers.
   if (save_doubles) {
     CpuFeatureScope scope(this, SSE2);
-    int space = XMMRegister::kNumRegisters * kDoubleSize + argc * kPointerSize;
+    int space = XMMRegister::kNumRegisters * kFloat32x4Size +
+        argc * kPointerSize;
     sub(esp, Immediate(space));
     const int offset = -2 * kPointerSize;
     for (int i = 0; i < XMMRegister::kNumRegisters; i++) {
       XMMRegister reg = XMMRegister::from_code(i);
-      movsd(Operand(ebp, offset - ((i + 1) * kDoubleSize)), reg);
+      movups(Operand(ebp, offset - ((i + 1) * kFloat32x4Size)), reg);
     }
   } else {
     sub(esp, Immediate(argc * kPointerSize));
@@ -1169,7 +1170,7 @@ void MacroAssembler::LeaveExitFrame(bool save_doubles) {
     const int offset = -2 * kPointerSize;
     for (int i = 0; i < XMMRegister::kNumRegisters; i++) {
       XMMRegister reg = XMMRegister::from_code(i);
-      movsd(reg, Operand(ebp, offset - ((i + 1) * kDoubleSize)));
+      movups(reg, Operand(ebp, offset - ((i + 1) * kFloat32x4Size)));
     }
   }
 
@@ -1835,6 +1836,30 @@ void MacroAssembler::AllocateHeapNumber(Register result,
 }
 
 
+void MacroAssembler::AllocateFloat32x4(Register result,
+                                       Register scratch,
+                                       Label* gc_required) {
+  // Allocate heap number in new space.
+  Allocate(Float32x4::kSize, result, scratch, no_reg, gc_required, TAG_OBJECT);
+
+  // Set the map.
+  mov(FieldOperand(result, HeapObject::kMapOffset),
+      Immediate(isolate()->factory()->float32x4_map()));
+}
+
+
+void MacroAssembler::AllocateInt32x4(Register result,
+                                      Register scratch,
+                                      Label* gc_required) {
+  // Allocate heap number in new space.
+  Allocate(Int32x4::kSize, result, scratch, no_reg, gc_required, TAG_OBJECT);
+
+  // Set the map.
+  mov(FieldOperand(result, HeapObject::kMapOffset),
+      Immediate(isolate()->factory()->int32x4_map()));
+}
+
+
 void MacroAssembler::AllocateTwoByteString(Register result,
                                            Register length,
                                            Register scratch1,
@@ -3697,6 +3722,59 @@ void MacroAssembler::JumpIfDictionaryInPrototypeChain(
   j(not_equal, &loop_again);
 }
 
+
+void MacroAssembler::absps(XMMRegister dst) {
+  static const struct V8_ALIGNED(16) {
+    uint32_t a;
+    uint32_t b;
+    uint32_t c;
+    uint32_t d;
+  } float_absolute_constant =
+      { 0x7FFFFFFF, 0x7FFFFFFF, 0x7FFFFFFF, 0x7FFFFFFF };
+  andps(dst,
+        Operand(reinterpret_cast<int32_t>(&float_absolute_constant)));
+}
+
+
+void MacroAssembler::notps(XMMRegister dst) {
+  static const struct V8_ALIGNED(16) {
+    uint32_t a;
+    uint32_t b;
+    uint32_t c;
+    uint32_t d;
+  } float_not_constant =
+      { 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF };
+  xorps(dst,
+        Operand(Operand(reinterpret_cast<int32_t>(&float_not_constant))));
+}
+
+
+void MacroAssembler::negateps(XMMRegister dst) {
+  static const struct V8_ALIGNED(16) {
+    uint32_t a;
+    uint32_t b;
+    uint32_t c;
+    uint32_t d;
+  } float_negate_constant =
+      { 0x80000000, 0x80000000, 0x80000000, 0x80000000 };
+  xorps(dst,
+        Operand(Operand(reinterpret_cast<int32_t>(&float_negate_constant))));
+}
+
+
+void MacroAssembler::pnegd(XMMRegister dst) {
+  static const struct V8_ALIGNED(16) {
+    uint32_t a;
+    uint32_t b;
+    uint32_t c;
+    uint32_t d;
+  } int32_one_constant = { 0x1, 0x1, 0x1, 0x1 };
+  notps(dst);
+  paddd(dst,
+        Operand(Operand(reinterpret_cast<int32_t>(&int32_one_constant))));
+}
+
+
 } }  // namespace v8::internal
 
 #endif  // V8_TARGET_ARCH_IA32