1 files changed, 36 insertions, 8 deletions

diff --git a/lib/Target/X86/X86InstrInfo.cpp b/lib/Target/X86/X86InstrInfo.cpp
index 1947d35..e555cd1 100644
--- a/lib/Target/X86/X86InstrInfo.cpp
+++ b/lib/Target/X86/X86InstrInfo.cpp
@@ -1018,13 +1018,11 @@ void X86InstrInfo::reMaterialize(MachineBasicBlock &MBB,
   switch (Opc) {
   default: break;
   case X86::MOV8r0:
-  case X86::MOV16r0:
   case X86::MOV32r0: {
     if (!isSafeToClobberEFLAGS(MBB, I)) {
       switch (Opc) {
       default: break;
       case X86::MOV8r0:  Opc = X86::MOV8ri;  break;
-      case X86::MOV16r0: Opc = X86::MOV16ri; break;
       case X86::MOV32r0: Opc = X86::MOV32ri; break;
       }
       Clone = false;
@@ -1880,7 +1878,7 @@ bool X86InstrInfo::copyRegToReg(MachineBasicBlock &MBB,
     if (SrcReg != X86::EFLAGS)
       return false;
     if (DestRC == &X86::GR64RegClass || DestRC == &X86::GR64_NOSPRegClass) {
-      BuildMI(MBB, MI, DL, get(X86::PUSHFQ));
+      BuildMI(MBB, MI, DL, get(X86::PUSHFQ64));
       BuildMI(MBB, MI, DL, get(X86::POP64r), DestReg);
       return true;
     } else if (DestRC == &X86::GR32RegClass ||
@@ -2292,9 +2290,7 @@ X86InstrInfo::foldMemoryOperandImpl(MachineFunction &MF,
     OpcodeTablePtr = &RegOp2MemOpTable2Addr;
     isTwoAddrFold = true;
   } else if (i == 0) { // If operand 0
-    if (MI->getOpcode() == X86::MOV16r0)
-      NewMI = MakeM0Inst(*this, X86::MOV16mi, MOs, MI);
-    else if (MI->getOpcode() == X86::MOV32r0)
+    if (MI->getOpcode() == X86::MOV32r0)
       NewMI = MakeM0Inst(*this, X86::MOV32mi, MOs, MI);
     else if (MI->getOpcode() == X86::MOV8r0)
       NewMI = MakeM0Inst(*this, X86::MOV8mi, MOs, MI);
@@ -2370,6 +2366,23 @@ MachineInstr* X86InstrInfo::foldMemoryOperandImpl(MachineFunction &MF,
   // Check switch flag 
   if (NoFusing) return NULL;
 
+  if (!MF.getFunction()->hasFnAttr(Attribute::OptimizeForSize))
+    switch (MI->getOpcode()) {
+    case X86::CVTSD2SSrr:
+    case X86::Int_CVTSD2SSrr:
+    case X86::CVTSS2SDrr:
+    case X86::Int_CVTSS2SDrr:
+    case X86::RCPSSr:
+    case X86::RCPSSr_Int:
+    case X86::ROUNDSDr_Int:
+    case X86::ROUNDSSr_Int:
+    case X86::RSQRTSSr:
+    case X86::RSQRTSSr_Int:
+    case X86::SQRTSSr:
+    case X86::SQRTSSr_Int:
+      return 0;
+    }
+
   const MachineFrameInfo *MFI = MF.getFrameInfo();
   unsigned Size = MFI->getObjectSize(FrameIndex);
   unsigned Alignment = MFI->getObjectAlignment(FrameIndex);
@@ -2405,6 +2418,23 @@ MachineInstr* X86InstrInfo::foldMemoryOperandImpl(MachineFunction &MF,
   // Check switch flag 
   if (NoFusing) return NULL;
 
+  if (!MF.getFunction()->hasFnAttr(Attribute::OptimizeForSize))
+    switch (MI->getOpcode()) {
+    case X86::CVTSD2SSrr:
+    case X86::Int_CVTSD2SSrr:
+    case X86::CVTSS2SDrr:
+    case X86::Int_CVTSS2SDrr:
+    case X86::RCPSSr:
+    case X86::RCPSSr_Int:
+    case X86::ROUNDSDr_Int:
+    case X86::ROUNDSSr_Int:
+    case X86::RSQRTSSr:
+    case X86::RSQRTSSr_Int:
+    case X86::SQRTSSr:
+    case X86::SQRTSSr_Int:
+      return 0;
+    }
+
   // Determine the alignment of the load.
   unsigned Alignment = 0;
   if (LoadMI->hasOneMemOperand())
@@ -2529,7 +2559,6 @@ bool X86InstrInfo::canFoldMemoryOperand(const MachineInstr *MI,
   } else if (OpNum == 0) { // If operand 0
     switch (Opc) {
     case X86::MOV8r0:
-    case X86::MOV16r0:
     case X86::MOV32r0:
       return true;
     default: break;
@@ -2558,7 +2587,6 @@ bool X86InstrInfo::unfoldMemoryOperand(MachineFunction &MF, MachineInstr *MI,
     MemOp2RegOpTable.find((unsigned*)MI->getOpcode());
   if (I == MemOp2RegOpTable.end())
     return false;
-  DebugLoc dl = MI->getDebugLoc();
   unsigned Opc = I->second.first;
   unsigned Index = I->second.second & 0xf;
   bool FoldedLoad = I->second.second & (1 << 4);