Merging r263139:

------------------------------------------------------------------------
r263139 | michael.kuperstein | 2016-03-10 10:43:21 -0800 (Thu, 10 Mar 2016) | 12 lines

[X86] Correctly select registers to pop into for x86_64

When trying to replace an add to esp with pops, we need to choose dead
registers to pop into. Registers clobbered by the call and not imp-def'd
by it should be safe. Except that it's not enough to check the register
itself isn't defined, we also need to make sure no overlapping registers
are defined either.

This fixes PR26711.

Differential Revision: http://reviews.llvm.org/D18029

------------------------------------------------------------------------

git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_38@269657 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/X86/X86FrameLowering.cpp b/lib/Target/X86/X86FrameLowering.cpp
index 75894e6734cfbc8317a3fbaf6c897a795545ae6a..3348e56bc7760bf9f04e765bdc7dfdc10a1525e3 100644 (file)
--- a/lib/Target/X86/X86FrameLowering.cpp
+++ b/lib/Target/X86/X86FrameLowering.cpp
@@ -2440,7 +2440,9 @@ bool X86FrameLowering::adjustStackWithPops(MachineBasicBlock &MBB,
 
     bool IsDef = false;
     for (const MachineOperand &MO : Prev->implicit_operands()) {
-      if (MO.isReg() && MO.isDef() && MO.getReg() == Candidate) {
+      if (MO.isReg() && MO.isDef() &&
+          (TRI->isSubRegisterEq(MO.getReg(), Candidate) ||
+           TRI->isSuperRegister(MO.getReg(), Candidate))) {
         IsDef = true;
         break;
       }

diff --git a/test/CodeGen/X86/pop-stack-cleanup.ll b/test/CodeGen/X86/pop-stack-cleanup.ll
index bcf7594065f396dc6c8c2b0f5426e14dfce7af5e..1738ccff54eb462262d3570c062496f156d55622 100644 (file)
--- a/test/CodeGen/X86/pop-stack-cleanup.ll
+++ b/test/CodeGen/X86/pop-stack-cleanup.ll
@@ -7,6 +7,7 @@ declare i64 @param2_ret64(i32 %a, i32 %b)
 declare void @param2(i32 %a, i32 %b)
 declare void @param3(i32 %a, i32 %b, i32 %c)
 declare void @param8(i64, i64, i64, i64, i64, i64, i64, i64)
+declare i32 @param8_ret(i64, i64, i64, i64, i64, i64, i64, i64)
 
 
 define void @test() minsize nounwind {
@@ -74,3 +75,13 @@ define void @test_linux64(i32 %size) minsize nounwind {
   call void @param8(i64 1, i64 2, i64 3, i64 4, i64 5, i64 6, i64 7, i64 8)
   ret void
 }
+
+define i32 @test_linux64_i32(i32 %size) minsize nounwind {
+; LINUX64-LABEL: test_linux64_i32:
+; LINUX64: callq param8_ret
+; LINUX64-NOT: popq %rax
+; LINUX64: retq
+  %a = alloca i64, i32 %size, align 8
+  %r = call i32 @param8_ret(i64 1, i64 2, i64 3, i64 4, i64 5, i64 6, i64 7, i64 8)
+  ret i32 %r
+}