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

[AMDGPU] Fixed addReg() in SIOptimizeExecMaskingPreRA.cpp 

The second argument is flags, not subreg.

Differential Revision: https://reviews.llvm.org/D61031

llvm-svn: 359017
This commit is contained in:
Stanislav Mekhanoshin 2019-04-23 17:59:26 +00:00
parent db41fe166a
commit c464dddccb
2 changed files with 23 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
llvm/lib/Target/AMDGPU/SIOptimizeExecMaskingPreRA.cpp
View File

@ -246,7 +246,7 @@ static unsigned optimizeVcndVcmpPair(MachineBasicBlock &MBB,
MachineInstr *Andn2 = BuildMI(MBB, *And, And->getDebugLoc(),
TII->get(Andn2Opc), And->getOperand(0).getReg())
.addReg(ExecReg)
.addReg(CCReg, CC->getSubReg());
.addReg(CCReg, 0, CC->getSubReg());
And->eraseFromParent();
LIS->InsertMachineInstrInMaps(*Andn2);

22
llvm/test/CodeGen/AMDGPU/optimize-negated-cond-exec-masking.mir
View File

@ -463,3 +463,25 @@ body: |
bb.4:
S_ENDPGM 0
...
# GCN: name: negated_cond_subreg
# GCN: %0.sub0_sub1:sreg_128 = IMPLICIT_DEF
# GCN-NEXT: $vcc = S_ANDN2_B64 $exec, %0.sub0_sub1, implicit-def $scc
# GCN-NEXT: S_CBRANCH_VCCNZ %bb.2, implicit $vcc
---
name: negated_cond_subreg
body: |
bb.0:
%0.sub0_sub1:sreg_128 = IMPLICIT_DEF
%1:vgpr_32 = V_CNDMASK_B32_e64 0, 0, 0, 1, %0.sub0_sub1, implicit $exec
%2.sub0_sub1:sreg_128 = V_CMP_NE_U32_e64 %1, 1, implicit $exec
$vcc = S_AND_B64 $exec, killed %2.sub0_sub1:sreg_128, implicit-def dead $scc
S_CBRANCH_VCCNZ %bb.2, implicit killed $vcc
S_BRANCH %bb.1
bb.1:
S_BRANCH %bb.0
bb.2:
S_ENDPGM 0
...