Files
derek.holloway 7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
Flatten subprojects into a single base monorepo
2026-09-02 16:34:47 -07:00

324 lines
9.5 KiB
YAML

# RUN: llc -mtriple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -verify-machineinstrs -run-pass post-RA-hazard-rec -o - %s | FileCheck -check-prefixes=GCN,GFX10 %s
# RUN: llc -mtriple=amdgcn -mcpu=gfx1100 -mattr=+wavefrontsize64 -verify-machineinstrs -run-pass post-RA-hazard-rec -o - %s | FileCheck -check-prefixes=GCN,GFX11 %s
# GCN-LABEL: name: hazard_smem_war
# GCN: S_LOAD_DWORD_IMM
# GFX10-NEXT: $sgpr_null = S_MOV_B32 0
# GCN-NEXT: V_CMP_EQ_F32
---
name: hazard_smem_war
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $vgpr0, $vgpr1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_no_hazard
# GCN: S_LOAD_DWORD_IMM
# GCN-NEXT: S_ADD_U32
# GCN-NEXT: V_CMP_EQ_F32
---
name: hazard_smem_war_no_hazard
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $sgpr4, $sgpr5, $vgpr0, $vgpr1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
$sgpr3 = S_ADD_U32 $sgpr4, $sgpr5, implicit-def $scc
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_dependent_salu
# GCN: S_LOAD_DWORD_IMM
# GCN-NEXT: S_WAITCNT
# GCN-NEXT: S_ADD_U32
# GCN-NEXT: V_CMP_EQ_F32
---
name: hazard_smem_war_dependent_salu
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $sgpr4, $vgpr0, $vgpr1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
S_WAITCNT 0
$sgpr3 = S_ADD_U32 $sgpr2, $sgpr4, implicit-def $scc
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_independent_salu
# GCN: S_LOAD_DWORD_IMM
# GCN-NEXT: S_WAITCNT
# GCN-NEXT: S_ADD_U32
# GCN-NEXT: V_CMP_EQ_F32
---
name: hazard_smem_war_independent_salu
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $sgpr4, $sgpr5, $vgpr0, $vgpr1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
S_WAITCNT 0
$sgpr3 = S_ADD_U32 $sgpr5, $sgpr4, implicit-def $scc
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_only_smem
# GCN: S_LOAD_DWORD_IMM
# GCN-NEXT: S_LOAD_DWORD_IMM
# GFX10-NEXT: $sgpr_null = S_MOV_B32 0
# GCN-NEXT: V_CMP_EQ_F32
---
name: hazard_smem_war_only_smem
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $sgpr6, $sgpr7, $vgpr0, $vgpr1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
$sgpr5 = S_LOAD_DWORD_IMM $sgpr6_sgpr7, 0, 0
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_only_waitcnt_0
# GCN: S_LOAD_DWORD_IMM
# GCN-NEXT: S_WAITCNT
# GCN-NEXT: V_CMP_EQ_F32
---
name: hazard_smem_war_only_waitcnt_0
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $vgpr0, $vgpr1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
S_WAITCNT 0
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_only_vmcnt_0
# GCN: S_LOAD_DWORD_IMM
# GFX10-NEXT: S_WAITCNT .Vmcnt_0_Lgkmcnt_15
# GFX10-NEXT: $sgpr_null = S_MOV_B32 0
# GFX11-NEXT: S_WAITCNT .Vmcnt_3_Expcnt_0_Lgkmcnt_55
# GCN-NEXT: V_CMP_EQ_F32
---
name: hazard_smem_war_only_vmcnt_0
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $vgpr0, $vgpr1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
S_WAITCNT 3952
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_only_expcnt_0
# GCN: S_LOAD_DWORD_IMM
# GFX10-NEXT: S_WAITCNT .Expcnt_0_Lgkmcnt_15
# GFX10-NEXT: $sgpr_null = S_MOV_B32 0
# GFX11-NEXT: S_WAITCNT .Vmcnt_51_Lgkmcnt_48
# GCN-NEXT: V_CMP_EQ_F32
---
name: hazard_smem_war_only_expcnt_0
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $vgpr0, $vgpr1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
S_WAITCNT 53007
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_only_lgkmcnt_0
# GCN: S_LOAD_DWORD_IMM
# GFX10-NEXT: S_WAITCNT .Lgkmcnt_0
# GFX11-NEXT: S_WAITCNT .Vmcnt_48_Lgkmcnt_7
# GCN-NEXT: V_CMP_EQ_F32
---
name: hazard_smem_war_only_lgkmcnt_0
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $vgpr0, $vgpr1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
S_WAITCNT 49279
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_only_waitcnt_lgkmcnt_0
# GCN: S_LOAD_DWORD_IMM
# GCN-NEXT: S_WAITCNT_LGKMCNT
# GCN-NEXT: V_CMP_EQ_F32
---
name: hazard_smem_war_only_waitcnt_lgkmcnt_0
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $vgpr0, $vgpr1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
S_WAITCNT_LGKMCNT $sgpr_null, 0
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_only_waitcnt_lgkmcnt_1
# GCN: S_LOAD_DWORD_IMM
# GCN-NEXT: S_WAITCNT_LGKMCNT
# GFX10-NEXT: $sgpr_null = S_MOV_B32 0
# GCN-NEXT: V_CMP_EQ_F32
---
name: hazard_smem_war_only_waitcnt_lgkmcnt_1
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $vgpr0, $vgpr1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
S_WAITCNT_LGKMCNT $sgpr_null, 1
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_branch
# GCN: S_LOAD_DWORD_IMM
# GFX10: $sgpr_null = S_MOV_B32 0
# GCN: V_CMP_EQ_F32
---
name: hazard_smem_war_branch
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $sgpr4, $vgpr0, $vgpr1
successors: %bb.1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
S_BRANCH %bb.1
bb.1:
liveins: $sgpr0, $sgpr1, $sgpr2, $vgpr0, $vgpr1
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_cbranch
# GCN: S_AND_B64
# GCN: S_LOAD_DWORD_IMM
# GCN: S_CBRANCH_VCCZ
# GFX10-NOT: $sgpr_null = S_MOV_B32 0
# GCN: V_CMP_EQ_F32
# GCN: S_ENDPGM 0
# GFX10: $sgpr_null = S_MOV_B32 0
# GCN: V_CMP_EQ_F32
---
name: hazard_smem_war_cbranch
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $sgpr4, $sgpr5, $vgpr0, $vgpr1
successors: %bb.1, %bb.2
$vcc = S_AND_B64 $sgpr4_sgpr5, $sgpr4_sgpr5, implicit-def $scc
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
S_CBRANCH_VCCZ %bb.2, implicit killed $vcc
bb.1:
liveins: $sgpr0, $sgpr1, $sgpr2, $sgpr4, $sgpr5, $vgpr0, $vgpr1
$sgpr4_sgpr5 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
bb.2:
liveins: $sgpr0, $sgpr1, $sgpr2, $sgpr4, $sgpr5, $vgpr0, $vgpr1
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_cbranch_carry
# GCN: S_AND_B64
# GCN: S_LOAD_DWORD_IMM
# GCN: S_CBRANCH_VCCZ
# GFX10-NOT: $sgpr_null = S_MOV_B32 0
# GCN: V_CMP_EQ_F32
# GCN-NEXT: S_ENDPGM 0
# GFX10-NOT: $sgpr_null = S_MOV_B32 0
# GCN: V_CMP_EQ_F32
# GFX10: $sgpr_null = S_MOV_B32 0
# GCN: V_CMP_EQ_F32
---
name: hazard_smem_war_cbranch_carry
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $sgpr4, $sgpr5, $vgpr0, $vgpr1
successors: %bb.1, %bb.2
$vcc = S_AND_B64 $sgpr4_sgpr5, $sgpr4_sgpr5, implicit-def $scc
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
S_CBRANCH_VCCZ %bb.2, implicit killed $vcc
bb.1:
liveins: $sgpr0, $sgpr1, $sgpr2, $sgpr4, $sgpr5, $vgpr0, $vgpr1
$sgpr4_sgpr5 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
bb.2:
successors: %bb.3
liveins: $sgpr0, $sgpr1, $sgpr2, $sgpr4, $sgpr5, $vgpr0, $vgpr1
$sgpr4_sgpr5 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
bb.3:
liveins: $sgpr0, $sgpr1, $sgpr2, $sgpr4, $sgpr5, $vgpr0, $vgpr1
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_backedge
# GFX10: $sgpr_null = S_MOV_B32 0
# GCN: V_CMP_EQ_F32
# GCN: S_LOAD_DWORD_IMM
---
name: hazard_smem_war_backedge
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $sgpr2, $vgpr0, $vgpr1
successors: %bb.1
$sgpr0_sgpr1 = V_CMP_EQ_F32_e64 0, $vgpr0, 0, $vgpr1, 1, implicit $mode, implicit $exec
bb.1:
liveins: $sgpr0, $sgpr1, $sgpr2, $vgpr0, $vgpr1
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
S_BRANCH %bb.0
...
# GCN-LABEL: name: hazard_smem_war_impdef
# GCN: S_LOAD_DWORD_IMM
# GFX10: $sgpr_null = S_MOV_B32 0
# GCN-NEXT: V_CMP_EQ_F32
---
name: hazard_smem_war_impdef
body: |
bb.0:
liveins: $vcc, $vgpr0
$sgpr0 = S_LOAD_DWORD_IMM $vcc, 0, 0
V_CMP_EQ_F32_e32 $vgpr0, $vgpr0, implicit-def $vcc, implicit $mode, implicit $exec
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_readlane
# GCN: S_LOAD_DWORD_IMM
# GFX10: $sgpr_null = S_MOV_B32 0
# GCN-NEXT: V_READLANE_B32
---
name: hazard_smem_war_readlane
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $sgpr3, $vgpr0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
$sgpr0 = V_READLANE_B32 $vgpr0, $sgpr3
S_ENDPGM 0
...
# GCN-LABEL: name: hazard_smem_war_readfirstlane
# GCN: S_LOAD_DWORD_IMM
# GFX10: $sgpr_null = S_MOV_B32 0
# GCN-NEXT: V_READFIRSTLANE_B32
---
name: hazard_smem_war_readfirstlane
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $vgpr0
$sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
$sgpr0 = V_READFIRSTLANE_B32 $vgpr0, implicit $exec
S_ENDPGM 0
...