RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-22 10:42:39 +01:00
Code Issues Projects Releases Wiki Activity

AMDGPU: Don't leave dead illegal VGPR->SGPR copies

Browse Source 
Normally DCE kills these, but at -O0 these get left behind
leaving suspicious looking illegal copies.

Replace with IMPLICIT_DEF to avoid iterator issues.

llvm-svn: 327842
This commit is contained in:
Matt Arsenault 2018-03-19 14:07:15 +00:00
parent 923a94b688
commit 942aacbda6
3 changed files with 46 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
lib/Target/AMDGPU/SIInstrInfo.cpp
View File

@ -3923,6 +3923,13 @@ void SIInstrInfo::moveToVALU(MachineInstr &TopInst) const {
MRI.replaceRegWith(DstReg, Inst.getOperand(1).getReg());
MRI.clearKillFlags(Inst.getOperand(1).getReg());
Inst.getOperand(0).setReg(DstReg);
// Make sure we don't leave around a dead VGPR->SGPR copy. Normally
// these are deleted later, but at -O0 it would leave a suspicious
// looking illegal copy of an undef register.
for (unsigned I = Inst.getNumOperands() - 1; I != 0; --I)
Inst.RemoveOperand(I);
Inst.setDesc(get(AMDGPU::IMPLICIT_DEF));
continue;
}

5
test/CodeGen/AMDGPU/control-flow-fastregalloc.ll
View File

@ -41,16 +41,17 @@
; GCN: {{^}}BB{{[0-9]+}}_1: ; %if
; GCN: s_mov_b32 m0, -1
; GCN: ds_read_b32 [[LOAD1:v[0-9]+]]
; GCN: s_waitcnt lgkmcnt(0)
; GCN: buffer_load_dword [[RELOAD_LOAD0:v[0-9]+]], off, s[0:3], s7 offset:[[LOAD0_OFFSET]] ; 4-byte Folded Reload
; GCN: s_waitcnt vmcnt(0) lgkmcnt(0)
; Spill val register
; GCN: v_add_i32_e32 [[VAL:v[0-9]+]], vcc, [[LOAD1]], [[RELOAD_LOAD0]]
; GCN: buffer_store_dword [[VAL]], off, s[0:3], s7 offset:[[VAL_OFFSET:[0-9]+]] ; 4-byte Folded Spill
; VMEM: [[ENDIF]]:
; Reload and restore exec mask
; VGPR: s_waitcnt lgkmcnt(0)
; VGPR: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_LO:[0-9]+]], [[SPILL_VGPR]], [[SAVEEXEC_LO_LANE]]
; VGPR: v_readlane_b32 s[[S_RELOAD_SAVEEXEC_HI:[0-9]+]], [[SPILL_VGPR]], [[SAVEEXEC_HI_LANE]]

38
test/CodeGen/AMDGPU/si-fix-sgpr-copies.mir
View File

@ -1,7 +1,6 @@
# RUN: llc -march=amdgcn -run-pass si-fix-sgpr-copies %s -o - | FileCheck %s -check-prefixes=GCN
--- |
define amdgpu_kernel void @phi_visit_order() { ret void }
---
name: phi_visit_order
tracksRegLiveness: true
@ -37,4 +36,39 @@ body: |
S_BRANCH %bb.1
...
---
# GCN-LABEL: name: dead_illegal_virtreg_copy
# GCN: %0:vgpr_32 = COPY $vgpr0
# GCN: %1:sreg_32_xm0 = IMPLICIT_DEF
# GCN: S_ENDPGM implicit %0
name: dead_illegal_virtreg_copy
tracksRegLiveness: true
body: |
bb.0:
liveins: $vgpr0
%0:vgpr_32 = COPY $vgpr0
%1:sreg_32_xm0 = COPY %0
S_ENDPGM implicit %1
...
---
# GCN-LABEL: name: dead_illegal_physreg_copy
# GCN %2:vgpr_32 = COPY $vgpr0
# GCN: %1:sreg_32_xm0 = IMPLICIT_DEF
# GCN: S_ENDPGM implicit %2
name: dead_illegal_physreg_copy
tracksRegLiveness: true
body: |
bb.0:
liveins: $vgpr0
%0:sreg_32_xm0 = COPY $vgpr0
%1:sreg_32_xm0 = COPY %0
S_ENDPGM implicit %1
...