replica
/
hanchenye-llvm-project
2
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Fix PR4485. 

Avoid unnecessary duplication of operand 0 of X86::FpSET_ST0_80. This duplication would
cause one register to remain on the stack at the function return.

llvm-svn: 74534
This commit is contained in:
Rafael Espindola 2009-06-30 16:40:03 +00:00
parent 5f12b5386c
commit 317fd045e2
3 changed files with 22 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
llvm/lib/Target/X86/X86FloatingPoint.cpp
View File

@ -991,13 +991,16 @@ void FPS::handleSpecialFP(MachineBasicBlock::iterator &I) {
case X86::FpSET_ST0_32:
case X86::FpSET_ST0_64:
case X86::FpSET_ST0_80: {
unsigned RegOnTop = getStackEntry(0);
unsigned Op0 = getFPReg(MI->getOperand(0));
// FpSET_ST0_80 is generated by copyRegToReg for both function return
// and inline assembly with the "st" constrain. In the latter case,
// it is possible for ST(0) to be alive after this instruction.
if (!MI->killsRegister(X86::FP0 + RegOnTop)) {
// Duplicate ST0
if (!MI->killsRegister(X86::FP0 + Op0)) {
// Duplicate Op0
duplicateToTop(0, 7 /*temp register*/, I);
} else {
moveToTop(Op0, I);
}
--StackTop; // "Forget" we have something on the top of stack!
break;

2
llvm/test/CodeGen/X86/inline-asm-fpstack4.ll
View File

@ -1,4 +1,4 @@
; RUN: llvm-as < %s | llc -march=x86 > %t
; RUN: llvm-as < %s | llc -march=x86
; PR4484
declare x86_fp80 @ceil()

15
llvm/test/CodeGen/X86/inline-asm-fpstack5.ll Normal file
View File

@ -0,0 +1,15 @@
; RUN: llvm-as < %s | llc -march=x86
; PR4485
define void @test(x86_fp80* %a) {
entry:
%0 = load x86_fp80* %a, align 16
%1 = fmul x86_fp80 %0, 0xK4006B400000000000000
%2 = fmul x86_fp80 %1, 0xK4012F424000000000000
tail call void asm sideeffect "fistpl $0", "{st},~{st}"(x86_fp80 %2)
%3 = load x86_fp80* %a, align 16
%4 = fmul x86_fp80 %3, 0xK4006B400000000000000
%5 = fmul x86_fp80 %4, 0xK4012F424000000000000
tail call void asm sideeffect "fistpl $0", "{st},~{st}"(x86_fp80 %5)
ret void
}