From 4eb2d228e9c676eea67c4a74090f76efa1a058fb Mon Sep 17 00:00:00 2001
From: Weiming Zhao <weimingz@codeaurora.org>
Date: Thu, 20 Mar 2014 23:28:16 +0000
Subject: [PATCH] Fix PR19136: [ARM] Fix Folding SP Update into vpush/vpop

Sicne MBB->computeRegisterLivenes() returns Dead for sub regs like s0,
d0 is used in vpop instead of updating sp, which causes s0 dead before
its use.

This patch checks the liveness of each subreg to make sure the reg is
actually dead.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@204411 91177308-0d34-0410-b5e6-96231b3b80d8
---
 lib/Target/ARM/ARMBaseInstrInfo.cpp   | 16 +++++++++++---
 test/CodeGen/ARM/fold-stack-adjust.ll | 31 +++++++++++++++++++++++++++
 2 files changed, 44 insertions(+), 3 deletions(-)

diff --git a/lib/Target/ARM/ARMBaseInstrInfo.cpp b/lib/Target/ARM/ARMBaseInstrInfo.cpp
index 21b8e6b8705..4c69a90c4b6 100644
--- a/lib/Target/ARM/ARMBaseInstrInfo.cpp
+++ b/lib/Target/ARM/ARMBaseInstrInfo.cpp
@@ -1866,6 +1866,15 @@ void llvm::emitARMRegPlusImmediate(MachineBasicBlock &MBB,
   }
 }
 
+static bool isAnySubRegLive(unsigned Reg, const TargetRegisterInfo *TRI,
+                      MachineInstr *MI) {
+  for (MCSubRegIterator Subreg(Reg, TRI, /* IncludeSelf */ true);
+       Subreg.isValid(); ++Subreg)
+    if (MI->getParent()->computeRegisterLiveness(TRI, *Subreg, MI) !=
+        MachineBasicBlock::LQR_Dead)
+      return true;
+  return false;
+}
 bool llvm::tryFoldSPUpdateIntoPushPop(const ARMSubtarget &Subtarget,
                                       MachineFunction &MF, MachineInstr *MI,
                                       unsigned NumBytes) {
@@ -1920,7 +1929,6 @@ bool llvm::tryFoldSPUpdateIntoPushPop(const ARMSubtarget &Subtarget,
   for (int i = MI->getNumOperands() - 1; i >= RegListIdx; --i)
     RegList.push_back(MI->getOperand(i));
 
-  MachineBasicBlock *MBB = MI->getParent();
   const TargetRegisterInfo *TRI = MF.getRegInfo().getTargetRegisterInfo();
   const MCPhysReg *CSRegs = TRI->getCalleeSavedRegs(&MF);
 
@@ -1941,9 +1949,11 @@ bool llvm::tryFoldSPUpdateIntoPushPop(const ARMSubtarget &Subtarget,
     // registers live within the function we might clobber a return value
     // register; the other way a register can be live here is if it's
     // callee-saved.
+    // TODO: Currently, computeRegisterLiveness() does not report "live" if a
+    // sub reg is live. When computeRegisterLiveness() works for sub reg, it
+    // can replace isAnySubRegLive().
     if (isCalleeSavedRegister(CurReg, CSRegs) ||
-        MBB->computeRegisterLiveness(TRI, CurReg, MI) !=
-            MachineBasicBlock::LQR_Dead) {
+        isAnySubRegLive(CurReg, TRI, MI)) {
       // VFP pops don't allow holes in the register list, so any skip is fatal
       // for our transformation. GPR pops do, so we should just keep looking.
       if (IsVFPPushPop)
diff --git a/test/CodeGen/ARM/fold-stack-adjust.ll b/test/CodeGen/ARM/fold-stack-adjust.ll
index feac793f4f8..695a20b5976 100644
--- a/test/CodeGen/ARM/fold-stack-adjust.ll
+++ b/test/CodeGen/ARM/fold-stack-adjust.ll
@@ -1,6 +1,7 @@
 ; RUN: llc -mtriple=thumbv7-apple-none-macho < %s | FileCheck %s
 ; RUN: llc -mtriple=thumbv6m-apple-none-macho -disable-fp-elim < %s | FileCheck %s --check-prefix=CHECK-T1
 ; RUN: llc -mtriple=thumbv7-apple-darwin-ios -disable-fp-elim < %s | FileCheck %s --check-prefix=CHECK-IOS
+; RUN: llc -mtriple=thumbv7--linux-gnueabi -disable-fp-elim < %s | FileCheck %s --check-prefix=CHECK-LINUX
 
 
 declare void @bar(i8*)
@@ -185,3 +186,33 @@ define void @test_varsize(...) minsize {
   call void @bar(i8* %var)
   ret void
 }
+
+%"MyClass" = type { i8*, i32, i32, float, float, float, [2 x i8], i32, i32* }
+
+declare float @foo()
+
+declare void @bar3()
+
+declare %"MyClass"* @bar2(%"MyClass"* returned, i16*, i32, float, float, i32, i32, i1 zeroext, i1 zeroext, i32)
+
+define fastcc float @check_vfp_no_return_clobber2(i16* %r, i16* %chars, i32 %length, i1 zeroext %flag) minsize {
+entry:
+; CHECK-LINUX-LABEL: check_vfp_no_return_clobber2
+; CHECK-LINUX: vpush	{d0, d1, d2, d3, d4, d5, d6, d7, d8}
+; CHECK-NOT: sub sp,
+; ...
+; CHECK-LINUX: add sp
+; CHECK-LINUX: vpop {d8}
+  %run = alloca %"MyClass", align 4
+  %call = call %"MyClass"* @bar2(%"MyClass"* %run, i16* %chars, i32 %length, float 0.000000e+00, float 0.000000e+00, i32 1, i32 1, i1 zeroext false, i1 zeroext true, i32 3)
+  %call1 = call float @foo()
+  %cmp = icmp eq %"MyClass"* %run, null
+  br i1 %cmp, label %exit, label %if.then
+
+if.then:                                          ; preds = %entry
+  call void @bar3()
+  br label %exit
+
+exit:                                             ; preds = %if.then, %entry
+  ret float %call1
+}