Even if we don't have 7 bytes of stack space we may need to save and

restore the stack pointer from the frame pointer on thumbv6.

Fixes rdar://8819685

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123196 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/ARM/Thumb1FrameLowering.cpp b/lib/Target/ARM/Thumb1FrameLowering.cpp
index e3f2d2d5981ff036a9f2fab9c4818618c3a0f7c5..8a6dda8214849fe923126b2e971190e0f9ce6755 100644 (file)
--- a/lib/Target/ARM/Thumb1FrameLowering.cpp
+++ b/lib/Target/ARM/Thumb1FrameLowering.cpp
@@ -156,6 +156,12 @@ void Thumb1FrameLowering::emitPrologue(MachineFunction &MF) const {
   // to reference locals.
   if (RegInfo->hasBasePointer(MF))
     BuildMI(MBB, MBBI, dl, TII.get(ARM::tMOVgpr2gpr), BasePtr).addReg(ARM::SP);
+    
+  // If the frame has variable sized objects then the epilogue must restore
+  // the sp from fp. We can assume there's an FP here since hasFP already
+  // checks for hasVarSizedObjects.
+  if (MFI->hasVarSizedObjects())
+    AFI->setShouldRestoreSPFromFP(true);
 }
 
 static bool isCalleeSavedRegister(unsigned Reg, const unsigned *CSRegs) {
@@ -221,7 +227,8 @@ void Thumb1FrameLowering::emitEpilogue(MachineFunction &MF,
     if (AFI->shouldRestoreSPFromFP()) {
       NumBytes = AFI->getFramePtrSpillOffset() - NumBytes;
       // Reset SP based on frame pointer only if the stack frame extends beyond
-      // frame pointer stack slot or target is ELF and the function has FP.
+      // frame pointer stack slot, the target is ELF and the function has FP, or
+      // the target uses var sized objects.
       if (NumBytes) {
         assert(MF.getRegInfo().isPhysRegUsed(ARM::R4) &&
                "No scratch register to restore SP from FP!");

diff --git a/test/CodeGen/ARM/thumb1-varalloc.ll b/test/CodeGen/ARM/thumb1-varalloc.ll
new file mode 100644 (file)
index 0000000..25093fe
--- /dev/null
+++ b/test/CodeGen/ARM/thumb1-varalloc.ll
@@ -0,0 +1,40 @@
+; RUN: llc < %s -mtriple=thumbv6-apple-darwin | FileCheck %s
+; rdar://8819685
+
+@__bar = external hidden global i8*
+@__baz = external hidden global i8*
+
+define i8* @_foo() {
+entry:
+; CHECK: foo:
+
+       %size = alloca i32, align 4
+       %0 = load i8** @__bar, align 4
+       %1 = icmp eq i8* %0, null
+       br i1 %1, label %bb1, label %bb3
+               
+bb1:
+       store i32 1026, i32* %size, align 4
+       %2 = alloca [1026 x i8], align 1
+; CHECK: mov     r0, sp
+; CHECK: adds    r4, r0, r4
+       %3 = getelementptr inbounds [1026 x i8]* %2, i32 0, i32 0
+       %4 = call i32 @_called_func(i8* %3, i32* %size) nounwind
+       %5 = icmp eq i32 %4, 0
+       br i1 %5, label %bb2, label %bb3
+       
+bb2:
+       %6 = call i8* @strdup(i8* %3) nounwind
+       store i8* %6, i8** @__baz, align 4
+       br label %bb3
+       
+bb3:
+       %.0 = phi i8* [ %0, %entry ], [ %6, %bb2 ], [ %3, %bb1 ]
+; CHECK: subs    r4, #5
+; CHECK-NEXT: mov     sp, r4
+; CHECK-NEXT: pop     {r4, r5, r6, r7, pc}
+       ret i8* %.0
+}
+
+declare noalias i8* @strdup(i8* nocapture) nounwind
+declare i32 @_called_func(i8*, i32*) nounwind
\ No newline at end of file