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
70 lines
3.2 KiB
YAML
70 lines
3.2 KiB
YAML
# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 6
|
|
# Test: fixVcmpxPermlaneHazards + S_SET_VGPR_MSB interaction on GFX1250
|
|
#
|
|
# Proves that VGPR lowering MUST run AFTER the hazard recognizer.
|
|
#
|
|
# fixVcmpxPermlaneHazards creates V_MOV_B32 Reg, Reg using permlane's src0.
|
|
# When src0 is a high VGPR (>= v256) but vdst is a low VGPR, the V_MOV_B32
|
|
# needs {dst=1, src0=1} MSB mode, while the permlane needs {dst=0, src0=1}.
|
|
#
|
|
# Current order (hazard-rec then vgpr-lowering): VGPR lowering sees the
|
|
# V_MOV_B32 and inserts S_SET_VGPR_MSB with correct {dst=1, src0=1}.
|
|
#
|
|
# Reversed order (vgpr-lowering then hazard-rec): The V_MOV_B32 inherits the
|
|
# permlane's {dst=0, src0=1}, causing it to write to v44 instead of v300.
|
|
|
|
# RUN: llc -mtriple=amdgcn -mcpu=gfx1250 -run-pass=post-RA-hazard-rec,amdgpu-lower-vgpr-encoding -o - %s | FileCheck %s
|
|
# RUN: llc -mtriple=amdgcn -mcpu=gfx1250 -start-before=post-RA-hazard-rec -o - %s | FileCheck -check-prefix=ASM %s
|
|
|
|
# vcmpx + permlane with high VGPR src0 ($vgpr300) and low VGPR vdst ($vgpr0).
|
|
# fixVcmpxPermlaneHazards inserts V_MOV_B32 $vgpr300, $vgpr300 before permlane.
|
|
#
|
|
# With correct order (hazard-rec then vgpr-lowering), VGPR lowering sees the
|
|
# V_MOV_B32 and inserts S_SET_VGPR_MSB 65 {dst=1, src0=1} before it,
|
|
# then S_SET_VGPR_MSB {dst=0, src0=1} before the permlane.
|
|
#
|
|
# If the order were reversed (vgpr-lowering then hazard-rec), VGPR lowering
|
|
# would only see the permlane and insert S_SET_VGPR_MSB 1 {dst=0, src0=1}.
|
|
# The V_MOV_B32 inserted later by the hazard recognizer would inherit dst=0,
|
|
# writing to v44 (300 & 0xFF) instead of v300.
|
|
#
|
|
# ASM-LABEL: vcmpx_permlane_high_vgpr_src0:
|
|
# ASM: v_cmpx_le_f32_e32 0, v0
|
|
# ASM: s_set_vgpr_msb 0x41
|
|
# ASM-NEXT: v_mov_b32_e32 v44 /*v300*/, v44 /*v300*/
|
|
# ASM-NEXT: s_set_vgpr_msb 0x4101
|
|
# ASM: v_permlane16_b32 v0, v44 /*v300*/, s1, s0
|
|
---
|
|
name: vcmpx_permlane_high_vgpr_src0
|
|
tracksRegLiveness: true
|
|
body: |
|
|
; CHECK-LABEL: name: vcmpx_permlane_high_vgpr_src0
|
|
; CHECK: bb.0:
|
|
; CHECK-NEXT: successors: %bb.1(0x80000000)
|
|
; CHECK-NEXT: liveins: $vgpr0
|
|
; CHECK-NEXT: {{ $}}
|
|
; CHECK-NEXT: $vgpr0 = V_MOV_B32_e32 0, implicit $exec
|
|
; CHECK-NEXT: V_CMPX_LE_F32_nosdst_e32 0, $vgpr0, implicit-def $exec, implicit $mode, implicit $exec
|
|
; CHECK-NEXT: S_BRANCH %bb.1
|
|
; CHECK-NEXT: {{ $}}
|
|
; CHECK-NEXT: bb.1:
|
|
; CHECK-NEXT: liveins: $vgpr0, $vgpr300, $sgpr0, $sgpr1
|
|
; CHECK-NEXT: {{ $}}
|
|
; CHECK-NEXT: S_SET_VGPR_MSB 65, implicit-def $mode
|
|
; CHECK-NEXT: $vgpr300 = V_MOV_B32_e32 killed $vgpr300, implicit $exec
|
|
; CHECK-NEXT: S_SET_VGPR_MSB 16641, implicit-def $mode
|
|
; CHECK-NEXT: $vgpr0 = V_PERMLANE16_B32_e64 0, killed $vgpr300, 0, killed $sgpr1, 0, killed $sgpr0, $vgpr0, 0, implicit $exec
|
|
; CHECK-NEXT: S_ENDPGM 0
|
|
bb.0:
|
|
successors: %bb.1
|
|
liveins: $vgpr0
|
|
$vgpr0 = V_MOV_B32_e32 0, implicit $exec
|
|
V_CMPX_LE_F32_nosdst_e32 0, $vgpr0, implicit-def $exec, implicit $mode, implicit $exec
|
|
S_BRANCH %bb.1
|
|
|
|
bb.1:
|
|
liveins: $vgpr0, $vgpr300, $sgpr0, $sgpr1
|
|
$vgpr0 = V_PERMLANE16_B32_e64 0, killed $vgpr300, 0, killed $sgpr1, 0, killed $sgpr0, $vgpr0, 0, implicit $exec
|
|
S_ENDPGM 0
|
|
...
|