Dynamic stack realignment use of sp register as source/dest register

in "bic sp, sp, #15" leads to unpredicatble behaviour in Thumb2 mode.
Emit the following code instead:
mov r4, sp
bic r4, r4, #15
mov sp, r4

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@90724 91177308-0d34-0410-b5e6-96231b3b80d8

3 files changed, 29 insertions, 6 deletions

diff --git a/lib/Target/ARM/ARMBaseRegisterInfo.cpp b/lib/Target/ARM/ARMBaseRegisterInfo.cpp
index ff11576030..9b5f79fb10 100644
--- a/lib/Target/ARM/ARMBaseRegisterInfo.cpp
+++ b/lib/Target/ARM/ARMBaseRegisterInfo.cpp
@@ -578,6 +578,13 @@ ARMBaseRegisterInfo::processFunctionBeforeCalleeSavedScan(MachineFunction &MF,
     MFI->calculateMaxStackAlignment();
   }
 
+  // Spill R4 if Thumb2 function requires stack realignment - it will be used as
+  // scratch register.
+  // FIXME: It will be better just to find spare register here.
+  if (needsStackRealignment(MF) &&
+      AFI->isThumb2Function())
+    MF.getRegInfo().setPhysRegUsed(ARM::R4);
+
   // Don't spill FP if the frame can be eliminated. This is determined
   // by scanning the callee-save registers to see if any is used.
   const unsigned *CSRegs = getCalleeSavedRegs();
@@ -1351,14 +1358,30 @@ emitPrologue(MachineFunction &MF) const {
 
   // If we need dynamic stack realignment, do it here.
   if (needsStackRealignment(MF)) {
-    unsigned Opc;
     unsigned MaxAlign = MFI->getMaxAlignment();
     assert (!AFI->isThumb1OnlyFunction());
-    Opc = AFI->isThumbFunction() ? ARM::t2BICri : ARM::BICri;
-
-    AddDefaultCC(AddDefaultPred(BuildMI(MBB, MBBI, dl, TII.get(Opc), ARM::SP)
+    if (!AFI->isThumbFunction()) {
+      // Emit bic sp, sp, MaxAlign
+      AddDefaultCC(AddDefaultPred(BuildMI(MBB, MBBI, dl,
+                                          TII.get(ARM::BICri), ARM::SP)
                                   .addReg(ARM::SP, RegState::Kill)
                                   .addImm(MaxAlign-1)));
+    } else {
+      // We cannot use sp as source/dest register here, thus we're emitting the
+      // following sequence:
+      // mov r4, sp
+      // bic r4, r4, MaxAlign
+      // mov sp, r4
+      // FIXME: It will be better just to find spare register here.
+      BuildMI(MBB, MBBI, dl, TII.get(ARM::tMOVtgpr2gpr), ARM::R4)
+        .addReg(ARM::SP, RegState::Kill);
+      AddDefaultCC(AddDefaultPred(BuildMI(MBB, MBBI, dl,
+                                          TII.get(ARM::t2BICri), ARM::R4)
+                                  .addReg(ARM::R4, RegState::Kill)
+                                  .addImm(MaxAlign-1)));
+      BuildMI(MBB, MBBI, dl, TII.get(ARM::tMOVtgpr2gpr), ARM::SP)
+        .addReg(ARM::R4, RegState::Kill);
+    }
   }
 }
 
diff --git a/test/CodeGen/Thumb2/large-stack.ll b/test/CodeGen/Thumb2/large-stack.ll
index 6f5996174a..da44cdea0f 100644
--- a/test/CodeGen/Thumb2/large-stack.ll
+++ b/test/CodeGen/Thumb2/large-stack.ll
@@ -18,7 +18,7 @@ define void @test2() {
 define i32 @test3() {
 ; CHECK: test3:
 ; CHECK: sub.w sp, sp, #805306368
-; CHECK: sub sp, #24
+; CHECK: sub sp, #20
     %retval = alloca i32, align 4
     %tmp = alloca i32, align 4
     %a = alloca [805306369 x i8], align 16
diff --git a/test/CodeGen/Thumb2/thumb2-spill-q.ll b/test/CodeGen/Thumb2/thumb2-spill-q.ll
index aef167b07f..2b087893fd 100644
--- a/test/CodeGen/Thumb2/thumb2-spill-q.ll
+++ b/test/CodeGen/Thumb2/thumb2-spill-q.ll
@@ -11,7 +11,7 @@ declare <4 x float> @llvm.arm.neon.vld1.v4f32(i8*) nounwind readonly
 
 define arm_apcscc void @aaa(%quuz* %this, i8* %block) {
 ; CHECK: aaa:
-; CHECK: bic sp, sp, #15
+; CHECK: bic r4, r4, #15
 ; CHECK: vst1.64 {{.*}}sp, :128
 ; CHECK: vld1.64 {{.*}}sp, :128
 entry: