Implemented fast paths for: (#846)

* opt

* Nit.

* opt_p2

* Nit.

5 files changed, 40 insertions, 4 deletions

diff --git a/ARMeilleure/CodeGen/X86/Assembler.cs b/ARMeilleure/CodeGen/X86/Assembler.cs
index 24a122c3..4568253a 100644
--- a/ARMeilleure/CodeGen/X86/Assembler.cs
+++ b/ARMeilleure/CodeGen/X86/Assembler.cs
@@ -103,6 +103,7 @@ namespace ARMeilleure.CodeGen.X86
             Add(X86Instruction.Cvtsi2sd,   new InstructionInfo(BadOp,      BadOp,      BadOp,      BadOp,      0x00000f2a, InstructionFlags.Vex | InstructionFlags.PrefixF2));
             Add(X86Instruction.Cvtsi2ss,   new InstructionInfo(BadOp,      BadOp,      BadOp,      BadOp,      0x00000f2a, InstructionFlags.Vex | InstructionFlags.PrefixF3));
             Add(X86Instruction.Cvtss2sd,   new InstructionInfo(BadOp,      BadOp,      BadOp,      BadOp,      0x00000f5a, InstructionFlags.Vex | InstructionFlags.PrefixF3));
+            Add(X86Instruction.Cvtss2si,   new InstructionInfo(BadOp,      BadOp,      BadOp,      BadOp,      0x00000f2d, InstructionFlags.Vex | InstructionFlags.PrefixF3));
             Add(X86Instruction.Div,        new InstructionInfo(BadOp,      BadOp,      BadOp,      BadOp,      0x060000f7, InstructionFlags.None));
             Add(X86Instruction.Divpd,      new InstructionInfo(BadOp,      BadOp,      BadOp,      BadOp,      0x00000f5e, InstructionFlags.Vex | InstructionFlags.Prefix66));
             Add(X86Instruction.Divps,      new InstructionInfo(BadOp,      BadOp,      BadOp,      BadOp,      0x00000f5e, InstructionFlags.Vex));
@@ -791,6 +792,16 @@ namespace ARMeilleure.CodeGen.X86
             }
         }
 
+        public void WriteInstruction(
+            X86Instruction inst,
+            Operand dest,
+            Operand src1,
+            Operand src2,
+            OperandType type)
+        {
+            WriteInstruction(dest, src1, src2, inst, type);
+        }
+
         public void WriteInstruction(X86Instruction inst, Operand dest, Operand source, byte imm)
         {
             WriteInstruction(dest, null, source, inst);
diff --git a/ARMeilleure/CodeGen/X86/CodeGenerator.cs b/ARMeilleure/CodeGen/X86/CodeGenerator.cs
index d1224363..0268665c 100644
--- a/ARMeilleure/CodeGen/X86/CodeGenerator.cs
+++ b/ARMeilleure/CodeGen/X86/CodeGenerator.cs
@@ -269,11 +269,11 @@ namespace ARMeilleure.CodeGen.X86
                         {
                             if (dest.Type == OperandType.I32)
                             {
-                                context.Assembler.Movd(dest, source); // int _mm_cvtsi128_si32
+                                context.Assembler.Movd(dest, source); // int _mm_cvtsi128_si32(__m128i a)
                             }
                             else /* if (dest.Type == OperandType.I64) */
                             {
-                                context.Assembler.Movq(dest, source); // __int64 _mm_cvtsi128_si64
+                                context.Assembler.Movq(dest, source); // __int64 _mm_cvtsi128_si64(__m128i a)
                             }
                         }
                         else
@@ -305,6 +305,26 @@ namespace ARMeilleure.CodeGen.X86
                         break;
                     }
 
+                    case IntrinsicType.BinaryGpr:
+                    {
+                        Operand dest = operation.Destination;
+                        Operand src1 = operation.GetSource(0);
+                        Operand src2 = operation.GetSource(1);
+
+                        EnsureSameType(dest, src1);
+
+                        if (!HardwareCapabilities.SupportsVexEncoding)
+                        {
+                            EnsureSameReg(dest, src1);
+                        }
+
+                        Debug.Assert(!dest.Type.IsInteger() && src2.Type.IsInteger());
+
+                        context.Assembler.WriteInstruction(info.Inst, dest, src1, src2, src2.Type);
+
+                        break;
+                    }
+
                     case IntrinsicType.BinaryImm:
                     {
                         Operand dest = operation.Destination;
@@ -1070,11 +1090,11 @@ namespace ARMeilleure.CodeGen.X86
 
             if (source.Type == OperandType.I32)
             {
-                context.Assembler.Movd(dest, source);
+                context.Assembler.Movd(dest, source); // (__m128i _mm_cvtsi32_si128(int a))
             }
             else /* if (source.Type == OperandType.I64) */
             {
-                context.Assembler.Movq(dest, source);
+                context.Assembler.Movq(dest, source); // (__m128i _mm_cvtsi64_si128(__int64 a))
             }
         }
 
diff --git a/ARMeilleure/CodeGen/X86/IntrinsicTable.cs b/ARMeilleure/CodeGen/X86/IntrinsicTable.cs
index e87de035..fd3b691d 100644
--- a/ARMeilleure/CodeGen/X86/IntrinsicTable.cs
+++ b/ARMeilleure/CodeGen/X86/IntrinsicTable.cs
@@ -41,8 +41,11 @@ namespace ARMeilleure.CodeGen.X86
             Add(Intrinsic.X86Cvtps2pd,   new IntrinsicInfo(X86Instruction.Cvtps2pd,   IntrinsicType.Unary));
             Add(Intrinsic.X86Cvtsd2si,   new IntrinsicInfo(X86Instruction.Cvtsd2si,   IntrinsicType.UnaryToGpr));
             Add(Intrinsic.X86Cvtsd2ss,   new IntrinsicInfo(X86Instruction.Cvtsd2ss,   IntrinsicType.Binary));
+            Add(Intrinsic.X86Cvtsi2sd,   new IntrinsicInfo(X86Instruction.Cvtsi2sd,   IntrinsicType.BinaryGpr));
             Add(Intrinsic.X86Cvtsi2si,   new IntrinsicInfo(X86Instruction.Movd,       IntrinsicType.UnaryToGpr));
+            Add(Intrinsic.X86Cvtsi2ss,   new IntrinsicInfo(X86Instruction.Cvtsi2ss,   IntrinsicType.BinaryGpr));
             Add(Intrinsic.X86Cvtss2sd,   new IntrinsicInfo(X86Instruction.Cvtss2sd,   IntrinsicType.Binary));
+            Add(Intrinsic.X86Cvtss2si,   new IntrinsicInfo(X86Instruction.Cvtss2si,   IntrinsicType.UnaryToGpr));
             Add(Intrinsic.X86Divpd,      new IntrinsicInfo(X86Instruction.Divpd,      IntrinsicType.Binary));
             Add(Intrinsic.X86Divps,      new IntrinsicInfo(X86Instruction.Divps,      IntrinsicType.Binary));
             Add(Intrinsic.X86Divsd,      new IntrinsicInfo(X86Instruction.Divsd,      IntrinsicType.Binary));
diff --git a/ARMeilleure/CodeGen/X86/IntrinsicType.cs b/ARMeilleure/CodeGen/X86/IntrinsicType.cs
index 4e9b33e1..41c52b59 100644
--- a/ARMeilleure/CodeGen/X86/IntrinsicType.cs
+++ b/ARMeilleure/CodeGen/X86/IntrinsicType.cs
@@ -7,6 +7,7 @@ namespace ARMeilleure.CodeGen.X86
         Unary,
         UnaryToGpr,
         Binary,
+        BinaryGpr,
         BinaryImm,
         Ternary,
         TernaryImm
diff --git a/ARMeilleure/CodeGen/X86/X86Instruction.cs b/ARMeilleure/CodeGen/X86/X86Instruction.cs
index a29e68fb..813730f2 100644
--- a/ARMeilleure/CodeGen/X86/X86Instruction.cs
+++ b/ARMeilleure/CodeGen/X86/X86Instruction.cs
@@ -38,6 +38,7 @@ namespace ARMeilleure.CodeGen.X86
         Cvtsi2sd,
         Cvtsi2ss,
         Cvtss2sd,
+        Cvtss2si,
         Div,
         Divpd,
         Divps,