[Sparc] Account for bias in stack readjustment

[jrtc27]

Summary

This was broken long ago in D12208, which failed to account for the fact that 64-bit SPARC uses a stack bias of 2047, and it is the *unbiased* value which should be aligned, not the biased one. This was seen to be an issue with Rust.

[jrtc27]

Ping

[jacob_hansen]

Can you please add a test for this? There is already a test in stack-align.ll that covers 32bit stack realignment, you can probably modify that testcase to cover 64bit also.

Hopefully someone with more SPARC V9 knowledge can chip in to verify the correctness of the patch.

[jrtc27]

I've added a 64-bit test to CodeGen/SPARC/stack-align.ll, and fixed the comment in SparcFrameLowering.cpp for the final instruction (%g1 and %o6 were round the wrong way in the comment).

[jyknight]

Looks good.

[jrtc27]

Ping?

lib/Target/Sparc/SparcFrameLowering.cpp

 }
 
 void SparcFrameLowering::emitPrologue(MachineFunction &MF,
                                       MachineBasicBlock &MBB) const {
   SparcMachineFunctionInfo *FuncInfo = MF.getInfo<SparcMachineFunctionInfo>();
 
   assert(&MF.front() == &MBB && "Shrink-wrapping not yet supported");
   MachineFrameInfo &MFI = MF.getFrameInfo();
+  const SparcSubtarget &Subtarget = MF.getSubtarget<SparcSubtarget>();
   const SparcInstrInfo &TII =
-      *static_cast<const SparcInstrInfo *>(MF.getSubtarget().getInstrInfo());
+      *static_cast<const SparcInstrInfo *>(Subtarget.getInstrInfo());
   const SparcRegisterInfo &RegInfo =
-      *static_cast<const SparcRegisterInfo *>(MF.getSubtarget().getRegisterInfo());
+      *static_cast<const SparcRegisterInfo *>(Subtarget.getRegisterInfo());
   MachineBasicBlock::iterator MBBI = MBB.begin();
   // Debug location must be unknown since the first debug location is used
   // to determine the end of the prologue.
   DebugLoc dl;
   bool NeedsStackRealignment = RegInfo.needsStackRealignment(MF);
 
   // FIXME: unfortunately, returning false from canRealignStack
   // actually just causes needsStackRealignment to return false,
void SparcFrameLowering::emitPrologue(MachineFunction &MF,
   // Add the extra call frame stack size, if needed. (This is the same
   // code as in PrologEpilogInserter, but also gets disabled by
   // targetHandlesStackFrameRounding)
   if (MFI.adjustsStack() && hasReservedCallFrame(MF))
     NumBytes += MFI.getMaxCallFrameSize();
 
   // Adds the SPARC subtarget-specific spill area to the stack
   // size. Also ensures target-required alignment.
-  NumBytes = MF.getSubtarget<SparcSubtarget>().getAdjustedFrameSize(NumBytes);
+  NumBytes = Subtarget.getAdjustedFrameSize(NumBytes);
 
   // Finally, ensure that the size is sufficiently aligned for the
   // data on the stack.
   if (MFI.getMaxAlignment() > 0) {
     NumBytes = alignTo(NumBytes, MFI.getMaxAlignment());
   }
 
   // Update stack size with corrected value.
void SparcFrameLowering::emitPrologue(MachineFunction &MF,
   unsigned regOutRA = RegInfo.getDwarfRegNum(SP::O7, true);
   // Emit ".cfi_register 15, 31".
   CFIIndex = MF.addFrameInst(
       MCCFIInstruction::createRegister(nullptr, regOutRA, regInRA));
   BuildMI(MBB, MBBI, dl, TII.get(TargetOpcode::CFI_INSTRUCTION))
       .addCFIIndex(CFIIndex);
 
   if (NeedsStackRealignment) {
-    // andn %o6, MaxAlign-1, %o6
+    int64_t Bias = Subtarget.getStackPointerBias();
+    unsigned regUnbiased;
+    if (Bias) {
+      // This clobbers G1 which we always know is available here.
+      regUnbiased = SP::G1;
+      // add %o6, BIAS, %g1
+      BuildMI(MBB, MBBI, dl, TII.get(SP::ADDri), regUnbiased)
+        .addReg(SP::O6).addImm(Bias);
+    } else
+      regUnbiased = SP::O6;
+
+    // andn %regUnbiased, MaxAlign-1, %regUnbiased
     int MaxAlign = MFI.getMaxAlignment();
-    BuildMI(MBB, MBBI, dl, TII.get(SP::ANDNri), SP::O6).addReg(SP::O6).addImm(MaxAlign - 1);
+    BuildMI(MBB, MBBI, dl, TII.get(SP::ANDNri), regUnbiased)
+      .addReg(regUnbiased).addImm(MaxAlign - 1);
+
+    if (Bias) {
+      // add %g1, -BIAS, %o6
+      BuildMI(MBB, MBBI, dl, TII.get(SP::ADDri), SP::O6)
+        .addReg(regUnbiased).addImm(-Bias);
+    }
   }
 }
 
 MachineBasicBlock::iterator SparcFrameLowering::
 eliminateCallFramePseudoInstr(MachineFunction &MF, MachineBasicBlock &MBB,
                               MachineBasicBlock::iterator I) const {
   if (!hasReservedCallFrame(MF)) {
     MachineInstr &MI = *I;

test/CodeGen/SPARC/stack-align.ll

-; RUN: llc -march=sparc < %s | FileCheck %s
+; RUN: llc -march=sparc < %s | FileCheck %s --check-prefixes=CHECK,CHECK32
+; RUN: llc -march=sparcv9 < %s | FileCheck %s --check-prefixes=CHECK,CHECK64
 declare void @stack_realign_helper(i32 %a, i32* %b)
 
 ;; This is a function where we have a local variable of 64-byte
 ;; alignment.  We want to see that the stack is aligned (the initial
 ;; andn), that the local var is accessed via stack pointer (to %o1), and that
 ;; the argument is accessed via frame pointer not stack pointer (to %o0).
 
 ;; CHECK-LABEL: stack_realign:
-;; CHECK:      andn %sp, 63, %sp
-;; CHECK-NEXT: ld [%fp+92], %o0
-;; CHECK-NEXT: call stack_realign_helper
-;; CHECK-NEXT: add %sp, 128, %o1
+;; CHECK32:      andn %sp, 63, %sp
+;; CHECK32-NEXT: ld [%fp+92], %o0
+;; CHECK64:      add %sp, 2047, %g1
+;; CHECK64-NEXT: andn %g1, 63, %g1
+;; CHECK64-NEXT: add %g1, -2047, %sp
+;; CHECK64-NEXT: ld [%fp+2227], %o0
+;; CHECK-NEXT:   call stack_realign_helper
+;; CHECK32-NEXT: add %sp, 128, %o1
+;; CHECK64-NEXT: add %sp, 2239, %o1
 
 define void @stack_realign(i32 %a, i32 %b, i32 %c, i32 %d, i32 %e, i32 %f, i32 %g) {
 entry:
   %aligned = alloca i32, align 64
   call void @stack_realign_helper(i32 %g, i32* %aligned)
   ret void
 }