Support segmented stack on Win64

Identical to Win32 method except for 0x14 -> 0x28 and FS -> GS.

Author: klutzy

lib/Target/X86/X86FrameLowering.cpp

@@ -1438,7 +1438,7 @@ X86FrameLowering::adjustForSegmentedStacks(MachineFunction &MF) const {
   if (MF.getFunction()->isVarArg())
     report_fatal_error("Segmented stacks do not support vararg functions.");
   if (!STI.isTargetLinux() && !STI.isTargetDarwin() &&
-      !STI.isTargetWin32() && !STI.isTargetFreeBSD())
+      !STI.isTargetWin32() && !STI.isTargetWin64() && !STI.isTargetFreeBSD())
     report_fatal_error("Segmented stacks not supported on this platform.");
 
   MachineBasicBlock *allocMBB = MF.CreateMachineBasicBlock();
@@ -1489,6 +1489,9 @@ X86FrameLowering::adjustForSegmentedStacks(MachineFunction &MF) const {
     } else if (STI.isTargetDarwin()) {
       TlsReg = X86::GS;
       TlsOffset = 0x60 + 90*8; // See pthread_machdep.h. Steal TLS slot 90.
+    } else if (STI.isTargetWin64()) {
+      TlsReg = X86::GS;
+      TlsOffset = 0x28; // pvArbitrary, reserved for application use
     } else if (STI.isTargetFreeBSD()) {
       TlsReg = X86::FS;
       TlsOffset = 0x18;
@@ -1526,7 +1529,7 @@ X86FrameLowering::adjustForSegmentedStacks(MachineFunction &MF) const {
       BuildMI(checkMBB, DL, TII.get(X86::LEA32r), ScratchReg).addReg(X86::ESP)
         .addImm(1).addReg(0).addImm(-StackSize).addReg(0);
 
-    if (STI.isTargetLinux() || STI.isTargetWin32()) {
+    if (STI.isTargetLinux() || STI.isTargetWin32() || STI.isTargetWin64()) {
       BuildMI(checkMBB, DL, TII.get(X86::CMP32rm)).addReg(ScratchReg)
         .addReg(0).addImm(0).addReg(0).addImm(TlsOffset).addReg(TlsReg);
     } else if (STI.isTargetDarwin()) {