6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2024-07-21 18:29:45 +00:00
Code Issues Projects Releases Wiki Activity

Fix PR9127 by reversing the operands even if they have more then one use.

Browse Source 
Reversing the operands allows us to fold, but doesn't force us to. Also, at
this point the DAG is still being optimized, so the check for hasOneUse is not
very precise.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@124773 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Rafael Espindola 2011-02-03 03:58:05 +00:00
parent b0cfa6cab8
commit f297c93191
2 changed files with 14 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
lib/Target/X86/X86ISelLowering.cpp
View File

@ -2712,8 +2712,8 @@ static unsigned TranslateX86CC(ISD::CondCode SetCCOpcode, bool isFP,
// First determine if it is required or is profitable to flip the operands.
// If LHS is a foldable load, but RHS is not, flip the condition.
if ((ISD::isNON_EXTLoad(LHS.getNode()) && LHS.hasOneUse()) &&
!(ISD::isNON_EXTLoad(RHS.getNode()) && RHS.hasOneUse())) {
if (ISD::isNON_EXTLoad(LHS.getNode()) &&
!ISD::isNON_EXTLoad(RHS.getNode())) {
SetCCOpcode = getSetCCSwappedOperands(SetCCOpcode);
std::swap(LHS, RHS);
}

12
test/CodeGen/X86/pr9127.ll Normal file
View File

@ -0,0 +1,12 @@
; RUN: llc < %s | FileCheck %s
define i8 @foobar(double %d, double* %x) {
entry:
%tmp2 = load double* %x, align 8
%cmp = fcmp oeq double %tmp2, %d
%conv3 = zext i1 %cmp to i8
ret i8 %conv3
}
; test that the load is folded.
; CHECK: ucomisd (%rdi), %xmm0