PR3739, part 2: Use an explicit store to spill XMM registers. (Previously,

the code tried to use "push", which doesn't exist for XMM registers.)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@72836 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/X86/X86InstrInfo.cpp b/lib/Target/X86/X86InstrInfo.cpp
index 2cd3733f0fb389220cb6d5959b82f2e255d7565f..8a9b7c917b4bf80f2a9322ff0ae08348e1606f12 100644 (file)
--- a/lib/Target/X86/X86InstrInfo.cpp
+++ b/lib/Target/X86/X86InstrInfo.cpp
@@ -2009,16 +2009,24 @@ bool X86InstrInfo::spillCalleeSavedRegisters(MachineBasicBlock &MBB,
 
   MachineFunction &MF = *MBB.getParent();
   X86MachineFunctionInfo *X86FI = MF.getInfo<X86MachineFunctionInfo>();
-  X86FI->setCalleeSavedFrameSize(CSI.size() * SlotSize);
+  unsigned CalleeFrameSize = 0;
   
   unsigned Opc = is64Bit ? X86::PUSH64r : X86::PUSH32r;
   for (unsigned i = CSI.size(); i != 0; --i) {
     unsigned Reg = CSI[i-1].getReg();
+    const TargetRegisterClass *RegClass = CSI[i-1].getRegClass();
     // Add the callee-saved register as live-in. It's killed at the spill.
     MBB.addLiveIn(Reg);
-    BuildMI(MBB, MI, DL, get(Opc))
-      .addReg(Reg, RegState::Kill);
+    if (RegClass != &X86::VR128RegClass) {
+      CalleeFrameSize += SlotSize;
+      BuildMI(MBB, MI, DL, get(Opc))
+        .addReg(Reg, RegState::Kill);
+    } else {
+      storeRegToStackSlot(MBB, MI, Reg, true, CSI[i-1].getFrameIdx(), RegClass);
+    }
   }
+
+  X86FI->setCalleeSavedFrameSize(CalleeFrameSize);
   return true;
 }
 
@@ -2036,7 +2044,12 @@ bool X86InstrInfo::restoreCalleeSavedRegisters(MachineBasicBlock &MBB,
   unsigned Opc = is64Bit ? X86::POP64r : X86::POP32r;
   for (unsigned i = 0, e = CSI.size(); i != e; ++i) {
     unsigned Reg = CSI[i].getReg();
-    BuildMI(MBB, MI, DL, get(Opc), Reg);
+    const TargetRegisterClass *RegClass = CSI[i].getRegClass();
+    if (RegClass != &X86::VR128RegClass) {
+      BuildMI(MBB, MI, DL, get(Opc), Reg);
+    } else {
+      loadRegFromStackSlot(MBB, MI, Reg, CSI[i].getFrameIdx(), RegClass);
+    }
   }
   return true;
 }

diff --git a/test/CodeGen/X86/2009-06-03-Win64SpillXMM.ll b/test/CodeGen/X86/2009-06-03-Win64SpillXMM.ll
new file mode 100644 (file)
index 0000000..33d7972
--- /dev/null
+++ b/test/CodeGen/X86/2009-06-03-Win64SpillXMM.ll
@@ -0,0 +1,12 @@
+; RUN: llvm-as < %s | llc -o %t1 -f
+; RUN: grep "subq.*\\\$40, \\\%rsp" %t1
+; RUN: grep "movaps    \\\%xmm8, \\\(\\\%rsp\\\)" %t1
+; RUN: grep "movaps    \\\%xmm7, 16\\\(\\\%rsp\\\)" %t1
+target triple = "x86_64-mingw64"
+
+define i32 @a() nounwind {
+entry:
+       tail call void asm sideeffect "", "~{xmm7},~{xmm8},~{dirflag},~{fpsr},~{flags}"() nounwind
+       ret i32 undef
+}
+