Vendor import of llvm trunk r130700:vendor/llvm/llvm-r130700

http://llvm.org/svn/llvm-project/llvm/trunk@130700

1 files changed, 95 insertions, 66 deletions

diff --git a/lib/Target/X86/X86FrameLowering.cpp b/lib/Target/X86/X86FrameLowering.cpp
index 0a3f931acf93..06d12fc04a34 100644
--- a/lib/Target/X86/X86FrameLowering.cpp
+++ b/lib/Target/X86/X86FrameLowering.cpp
@@ -22,6 +22,7 @@
 #include "llvm/CodeGen/MachineInstrBuilder.h"
 #include "llvm/CodeGen/MachineModuleInfo.h"
 #include "llvm/CodeGen/MachineRegisterInfo.h"
+#include "llvm/MC/MCAsmInfo.h"
 #include "llvm/Target/TargetData.h"
 #include "llvm/Target/TargetOptions.h"
 #include "llvm/Support/CommandLine.h"
@@ -296,7 +297,7 @@ void X86FrameLowering::emitCalleeSavedFrameMoves(MachineFunction &MF,
   // FIXME: This is dirty hack. The code itself is pretty mess right now.
   // It should be rewritten from scratch and generalized sometimes.
 
-  // Determine maximum offset (minumum due to stack growth).
+  // Determine maximum offset (minimum due to stack growth).
   int64_t MaxOffset = 0;
   for (std::vector<CalleeSavedInfo>::const_iterator
          I = CSI.begin(), E = CSI.end(); I != E; ++I)
@@ -551,65 +552,71 @@ void X86FrameLowering::emitPrologue(MachineFunction &MF) const {
   // responsible for adjusting the stack pointer.  Touching the stack at 4K
   // increments is necessary to ensure that the guard pages used by the OS
   // virtual memory manager are allocated in correct sequence.
-  if (NumBytes >= 4096 &&
-      (STI.isTargetCygMing() || STI.isTargetWin32()) &&
-      !STI.isTargetEnvMacho()) {
+  if (NumBytes >= 4096 && STI.isTargetCOFF() && !STI.isTargetEnvMacho()) {
+    const char *StackProbeSymbol;
+    bool isSPUpdateNeeded = false;
+
+    if (Is64Bit) {
+      if (STI.isTargetCygMing())
+        StackProbeSymbol = "___chkstk";
+      else {
+        StackProbeSymbol = "__chkstk";
+        isSPUpdateNeeded = true;
+      }
+    } else if (STI.isTargetCygMing())
+      StackProbeSymbol = "_alloca";
+    else
+      StackProbeSymbol = "_chkstk";
+
     // Check whether EAX is livein for this function.
     bool isEAXAlive = isEAXLiveIn(MF);
 
-    const char *StackProbeSymbol =
-      STI.isTargetWindows() ? "_chkstk" : "_alloca";
-    if (Is64Bit && STI.isTargetCygMing())
-      StackProbeSymbol = "__chkstk";
-    unsigned CallOp = Is64Bit ? X86::CALL64pcrel32 : X86::CALLpcrel32;
-    if (!isEAXAlive) {
-      BuildMI(MBB, MBBI, DL, TII.get(X86::MOV32ri), X86::EAX)
-        .addImm(NumBytes);
-      BuildMI(MBB, MBBI, DL, TII.get(CallOp))
-        .addExternalSymbol(StackProbeSymbol)
-        .addReg(StackPtr,    RegState::Define | RegState::Implicit)
-        .addReg(X86::EFLAGS, RegState::Define | RegState::Implicit);
-    } else {
+    if (isEAXAlive) {
+      // Sanity check that EAX is not livein for this function.
+      // It should not be, so throw an assert.
+      assert(!Is64Bit && "EAX is livein in x64 case!");
+
       // Save EAX
       BuildMI(MBB, MBBI, DL, TII.get(X86::PUSH32r))
         .addReg(X86::EAX, RegState::Kill);
+    }
 
-      // Allocate NumBytes-4 bytes on stack. We'll also use 4 already
-      // allocated bytes for EAX.
+    if (Is64Bit) {
+      // Handle the 64-bit Windows ABI case where we need to call __chkstk.
+      // Function prologue is responsible for adjusting the stack pointer.
+      BuildMI(MBB, MBBI, DL, TII.get(X86::MOV64ri), X86::RAX)
+        .addImm(NumBytes);
+    } else {
+      // Allocate NumBytes-4 bytes on stack in case of isEAXAlive.
+      // We'll also use 4 already allocated bytes for EAX.
       BuildMI(MBB, MBBI, DL, TII.get(X86::MOV32ri), X86::EAX)
-        .addImm(NumBytes - 4);
-      BuildMI(MBB, MBBI, DL, TII.get(CallOp))
-        .addExternalSymbol(StackProbeSymbol)
-        .addReg(StackPtr,    RegState::Define | RegState::Implicit)
-        .addReg(X86::EFLAGS, RegState::Define | RegState::Implicit);
-
-      // Restore EAX
-      MachineInstr *MI = addRegOffset(BuildMI(MF, DL, TII.get(X86::MOV32rm),
-                                              X86::EAX),
-                                      StackPtr, false, NumBytes - 4);
-      MBB.insert(MBBI, MI);
+        .addImm(isEAXAlive ? NumBytes - 4 : NumBytes);
+    }
+
+    BuildMI(MBB, MBBI, DL,
+            TII.get(Is64Bit ? X86::W64ALLOCA : X86::CALLpcrel32))
+      .addExternalSymbol(StackProbeSymbol)
+      .addReg(StackPtr,    RegState::Define | RegState::Implicit)
+      .addReg(X86::EFLAGS, RegState::Define | RegState::Implicit);
+
+    // MSVC x64's __chkstk needs to adjust %rsp.
+    // FIXME: %rax preserves the offset and should be available.
+    if (isSPUpdateNeeded)
+      emitSPUpdate(MBB, MBBI, StackPtr, -(int64_t)NumBytes, Is64Bit,
+                   TII, *RegInfo);
+
+    if (isEAXAlive) {
+        // Restore EAX
+        MachineInstr *MI = addRegOffset(BuildMI(MF, DL, TII.get(X86::MOV32rm),
+                                                X86::EAX),
+                                        StackPtr, false, NumBytes - 4);
+        MBB.insert(MBBI, MI);
     }
-  } else if (NumBytes >= 4096 &&
-             STI.isTargetWin64() &&
-             !STI.isTargetEnvMacho()) {
-    // Sanity check that EAX is not livein for this function.  It should
-    // not be, so throw an assert.
-    assert(!isEAXLiveIn(MF) && "EAX is livein in the Win64 case!");
-
-    // Handle the 64-bit Windows ABI case where we need to call __chkstk.
-    // Function prologue is responsible for adjusting the stack pointer.
-    BuildMI(MBB, MBBI, DL, TII.get(X86::MOV32ri), X86::EAX)
-      .addImm(NumBytes);
-    BuildMI(MBB, MBBI, DL, TII.get(X86::WINCALL64pcrel32))
-      .addExternalSymbol("__chkstk")
-      .addReg(StackPtr, RegState::Define | RegState::Implicit);
-    emitSPUpdate(MBB, MBBI, StackPtr, -(int64_t)NumBytes, Is64Bit,
-                 TII, *RegInfo);
   } else if (NumBytes)
     emitSPUpdate(MBB, MBBI, StackPtr, -(int64_t)NumBytes, Is64Bit,
                  TII, *RegInfo);
 
-  if ((NumBytes || PushedRegs) && needsFrameMoves) {
+  if (( (!HasFP && NumBytes) || PushedRegs) && needsFrameMoves) {
     // Mark end of stack pointer adjustment.
     MCSymbol *Label = MMI.getContext().CreateTempSymbol();
     BuildMI(MBB, MBBI, DL, TII.get(X86::PROLOG_LABEL)).addSym(Label);
@@ -779,7 +786,7 @@ void X86FrameLowering::emitEpilogue(MachineFunction &MF,
     assert(Offset >= 0 && "Offset should never be negative");
 
     if (Offset) {
-      // Check for possible merge with preceeding ADD instruction.
+      // Check for possible merge with preceding ADD instruction.
       Offset += mergeSPUpdates(MBB, MBBI, StackPtr, true);
       emitSPUpdate(MBB, MBBI, StackPtr, Offset, Is64Bit, TII, *RegInfo);
     }
@@ -823,7 +830,7 @@ void X86FrameLowering::emitEpilogue(MachineFunction &MF,
     int delta = -1*X86FI->getTCReturnAddrDelta();
     MBBI = MBB.getLastNonDebugInstr();
 
-    // Check for possible merge with preceeding ADD instruction.
+    // Check for possible merge with preceding ADD instruction.
     delta += mergeSPUpdates(MBB, MBBI, StackPtr, true);
     emitSPUpdate(MBB, MBBI, StackPtr, delta, Is64Bit, TII, *RegInfo);
   }
@@ -892,7 +899,6 @@ bool X86FrameLowering::spillCalleeSavedRegisters(MachineBasicBlock &MBB,
 
   MachineFunction &MF = *MBB.getParent();
 
-  bool isWin64 = STI.isTargetWin64();
   unsigned SlotSize = STI.is64Bit() ? 8 : 4;
   unsigned FPReg = TRI->getFrameRegister(MF);
   unsigned CalleeFrameSize = 0;
@@ -900,25 +906,39 @@ bool X86FrameLowering::spillCalleeSavedRegisters(MachineBasicBlock &MBB,
   const TargetInstrInfo &TII = *MF.getTarget().getInstrInfo();
   X86MachineFunctionInfo *X86FI = MF.getInfo<X86MachineFunctionInfo>();
 
+  // Push GPRs. It increases frame size.
   unsigned Opc = STI.is64Bit() ? X86::PUSH64r : X86::PUSH32r;
   for (unsigned i = CSI.size(); i != 0; --i) {
     unsigned Reg = CSI[i-1].getReg();
+    if (!X86::GR64RegClass.contains(Reg) &&
+        !X86::GR32RegClass.contains(Reg))
+      continue;
     // Add the callee-saved register as live-in. It's killed at the spill.
     MBB.addLiveIn(Reg);
     if (Reg == FPReg)
       // X86RegisterInfo::emitPrologue will handle spilling of frame register.
       continue;
-    if (!X86::VR128RegClass.contains(Reg) && !isWin64) {
-      CalleeFrameSize += SlotSize;
-      BuildMI(MBB, MI, DL, TII.get(Opc)).addReg(Reg, RegState::Kill);
-    } else {
-      const TargetRegisterClass *RC = TRI->getMinimalPhysRegClass(Reg);
-      TII.storeRegToStackSlot(MBB, MI, Reg, true, CSI[i-1].getFrameIdx(),
-                              RC, TRI);
-    }
+    CalleeFrameSize += SlotSize;
+    BuildMI(MBB, MI, DL, TII.get(Opc)).addReg(Reg, RegState::Kill);
   }
 
   X86FI->setCalleeSavedFrameSize(CalleeFrameSize);
+
+  // Make XMM regs spilled. X86 does not have ability of push/pop XMM.
+  // It can be done by spilling XMMs to stack frame.
+  // Note that only Win64 ABI might spill XMMs.
+  for (unsigned i = CSI.size(); i != 0; --i) {
+    unsigned Reg = CSI[i-1].getReg();
+    if (X86::GR64RegClass.contains(Reg) ||
+        X86::GR32RegClass.contains(Reg))
+      continue;
+    // Add the callee-saved register as live-in. It's killed at the spill.
+    MBB.addLiveIn(Reg);
+    const TargetRegisterClass *RC = TRI->getMinimalPhysRegClass(Reg);
+    TII.storeRegToStackSlot(MBB, MI, Reg, true, CSI[i-1].getFrameIdx(),
+                            RC, TRI);
+  }
+
   return true;
 }
 
@@ -933,21 +953,30 @@ bool X86FrameLowering::restoreCalleeSavedRegisters(MachineBasicBlock &MBB,
 
   MachineFunction &MF = *MBB.getParent();
   const TargetInstrInfo &TII = *MF.getTarget().getInstrInfo();
+
+  // Reload XMMs from stack frame.
+  for (unsigned i = 0, e = CSI.size(); i != e; ++i) {
+    unsigned Reg = CSI[i].getReg();
+    if (X86::GR64RegClass.contains(Reg) ||
+        X86::GR32RegClass.contains(Reg))
+      continue;
+    const TargetRegisterClass *RC = TRI->getMinimalPhysRegClass(Reg);
+    TII.loadRegFromStackSlot(MBB, MI, Reg, CSI[i].getFrameIdx(),
+                             RC, TRI);
+  }
+
+  // POP GPRs.
   unsigned FPReg = TRI->getFrameRegister(MF);
-  bool isWin64 = STI.isTargetWin64();
   unsigned Opc = STI.is64Bit() ? X86::POP64r : X86::POP32r;
   for (unsigned i = 0, e = CSI.size(); i != e; ++i) {
     unsigned Reg = CSI[i].getReg();
+    if (!X86::GR64RegClass.contains(Reg) &&
+        !X86::GR32RegClass.contains(Reg))
+      continue;
     if (Reg == FPReg)
       // X86RegisterInfo::emitEpilogue will handle restoring of frame register.
       continue;
-    if (!X86::VR128RegClass.contains(Reg) && !isWin64) {
-      BuildMI(MBB, MI, DL, TII.get(Opc), Reg);
-    } else {
-      const TargetRegisterClass *RC = TRI->getMinimalPhysRegClass(Reg);
-      TII.loadRegFromStackSlot(MBB, MI, Reg, CSI[i].getFrameIdx(),
-                               RC, TRI);
-    }
+    BuildMI(MBB, MI, DL, TII.get(Opc), Reg);
   }
   return true;
 }