Files
TheRock/compiler/amd-llvm/llvm/test/CodeGen/AArch64/licm-regclass-copy.mir
T
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

198 lines
6.4 KiB
YAML

# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
# RUN: llc -mtriple=aarch64 -run-pass=early-machinelicm -o - %s | FileCheck %s
# This test verifies that cross-bank copies (e.g., GPR to FPR, FPR to GPR)
# are hoisted out of loops by MachineLICM, as they are expensive on AArch64.
--- |
declare void @use_float(float)
declare void @use_int(i32)
define void @gpr_to_fpr_virtual_copy_hoisted() {
ret void
}
define void @gpr_to_fpr_physical_copy_hoisted() {
ret void
}
define void @fpr_to_gpr_virtual_copy_hoisted() {
ret void
}
...
---
name: gpr_to_fpr_virtual_copy_hoisted
tracksRegLiveness: true
body: |
; CHECK-LABEL: name: gpr_to_fpr_virtual_copy_hoisted
; CHECK: bb.0:
; CHECK-NEXT: successors: %bb.1(0x80000000)
; CHECK-NEXT: liveins: $w0, $w1
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: [[COPY:%[0-9]+]]:gpr32 = COPY $w0
; CHECK-NEXT: [[COPY1:%[0-9]+]]:gpr32 = COPY $w1
; CHECK-NEXT: [[COPY2:%[0-9]+]]:gpr32all = COPY $wzr
; CHECK-NEXT: [[COPY3:%[0-9]+]]:gpr32all = COPY [[COPY2]]
; CHECK-NEXT: [[COPY4:%[0-9]+]]:fpr32 = COPY [[COPY1]]
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: bb.1:
; CHECK-NEXT: successors: %bb.3(0x40000000), %bb.2(0x40000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: [[PHI:%[0-9]+]]:gpr32common = PHI [[COPY3]], %bb.0, %5, %bb.2
; CHECK-NEXT: [[SUBSWrr:%[0-9]+]]:gpr32 = SUBSWrr [[PHI]], [[COPY]], implicit-def $nzcv
; CHECK-NEXT: Bcc 1, %bb.3, implicit $nzcv
; CHECK-NEXT: B %bb.2
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: bb.2:
; CHECK-NEXT: successors: %bb.1(0x80000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: $s0 = COPY [[COPY4]]
; CHECK-NEXT: BL @use_float, implicit-def dead $lr, implicit $sp, implicit $s0, implicit-def $sp
; CHECK-NEXT: [[ADDWri:%[0-9]+]]:gpr32sp = ADDWri [[PHI]], 1, 0
; CHECK-NEXT: [[COPY5:%[0-9]+]]:gpr32all = COPY [[ADDWri]]
; CHECK-NEXT: B %bb.1
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: bb.3:
; CHECK-NEXT: RET_ReallyLR
bb.0:
liveins: $w0, $w1
%1:gpr32 = COPY $w0
%0:gpr32 = COPY $w1
%3:gpr32all = COPY $wzr
%2:gpr32all = COPY %3:gpr32all
bb.1:
%4:gpr32common = PHI %2:gpr32all, %bb.0, %5:gpr32all, %bb.2
%6:gpr32 = SUBSWrr %4:gpr32common, %1:gpr32, implicit-def $nzcv
Bcc 1, %bb.3, implicit $nzcv
B %bb.2
bb.2:
%7:fpr32 = COPY %0:gpr32
$s0 = COPY %7:fpr32
BL @use_float, implicit-def dead $lr, implicit $sp, implicit $s0, implicit-def $sp
%8:gpr32sp = ADDWri %4:gpr32common, 1, 0
%5:gpr32all = COPY %8:gpr32sp
B %bb.1
bb.3:
RET_ReallyLR
...
---
name: gpr_to_fpr_physical_copy_hoisted
tracksRegLiveness: true
body: |
; CHECK-LABEL: name: gpr_to_fpr_physical_copy_hoisted
; CHECK: bb.0:
; CHECK-NEXT: successors: %bb.1(0x80000000)
; CHECK-NEXT: liveins: $w0
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: [[COPY:%[0-9]+]]:gpr32 = COPY $w0
; CHECK-NEXT: [[COPY1:%[0-9]+]]:gpr32all = COPY $wzr
; CHECK-NEXT: [[COPY2:%[0-9]+]]:gpr32all = COPY [[COPY1]]
; CHECK-NEXT: [[COPY3:%[0-9]+]]:fpr32 = COPY $wzr
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: bb.1:
; CHECK-NEXT: successors: %bb.3(0x40000000), %bb.2(0x40000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: [[PHI:%[0-9]+]]:gpr32common = PHI [[COPY2]], %bb.0, %4, %bb.2
; CHECK-NEXT: [[SUBSWrr:%[0-9]+]]:gpr32 = SUBSWrr [[PHI]], [[COPY]], implicit-def $nzcv
; CHECK-NEXT: Bcc 1, %bb.3, implicit $nzcv
; CHECK-NEXT: B %bb.2
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: bb.2:
; CHECK-NEXT: successors: %bb.1(0x80000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: $s0 = COPY [[COPY3]]
; CHECK-NEXT: BL @use_float, implicit-def dead $lr, implicit $sp, implicit $s0, implicit-def $sp
; CHECK-NEXT: [[ADDWri:%[0-9]+]]:gpr32sp = ADDWri [[PHI]], 1, 0
; CHECK-NEXT: [[COPY4:%[0-9]+]]:gpr32all = COPY [[ADDWri]]
; CHECK-NEXT: B %bb.1
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: bb.3:
; CHECK-NEXT: RET_ReallyLR
bb.0:
liveins: $w0
%1:gpr32 = COPY $w0
%3:gpr32all = COPY $wzr
%2:gpr32all = COPY %3:gpr32all
bb.1:
%4:gpr32common = PHI %2:gpr32all, %bb.0, %5:gpr32all, %bb.2
%6:gpr32 = SUBSWrr %4:gpr32common, %1:gpr32, implicit-def $nzcv
Bcc 1, %bb.3, implicit $nzcv
B %bb.2
bb.2:
%7:fpr32 = COPY $wzr
$s0 = COPY %7:fpr32
BL @use_float, implicit-def dead $lr, implicit $sp, implicit $s0, implicit-def $sp
%8:gpr32sp = ADDWri %4:gpr32common, 1, 0
%5:gpr32all = COPY %8:gpr32sp
B %bb.1
bb.3:
RET_ReallyLR
...
---
name: fpr_to_gpr_virtual_copy_hoisted
tracksRegLiveness: true
body: |
; CHECK-LABEL: name: fpr_to_gpr_virtual_copy_hoisted
; CHECK: bb.0:
; CHECK-NEXT: successors: %bb.1(0x80000000)
; CHECK-NEXT: liveins: $w0, $s0
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: [[COPY:%[0-9]+]]:gpr32 = COPY $w0
; CHECK-NEXT: [[COPY1:%[0-9]+]]:fpr32 = COPY $s0
; CHECK-NEXT: [[COPY2:%[0-9]+]]:gpr32all = COPY $wzr
; CHECK-NEXT: [[COPY3:%[0-9]+]]:gpr32all = COPY [[COPY2]]
; CHECK-NEXT: [[COPY4:%[0-9]+]]:gpr32 = COPY [[COPY1]]
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: bb.1:
; CHECK-NEXT: successors: %bb.3(0x40000000), %bb.2(0x40000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: [[PHI:%[0-9]+]]:gpr32common = PHI [[COPY3]], %bb.0, %5, %bb.2
; CHECK-NEXT: [[SUBSWrr:%[0-9]+]]:gpr32 = SUBSWrr [[PHI]], [[COPY]], implicit-def $nzcv
; CHECK-NEXT: Bcc 1, %bb.3, implicit $nzcv
; CHECK-NEXT: B %bb.2
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: bb.2:
; CHECK-NEXT: successors: %bb.1(0x80000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: $w0 = COPY [[COPY4]]
; CHECK-NEXT: BL @use_int, implicit-def dead $lr, implicit $sp, implicit $w0, implicit-def $sp
; CHECK-NEXT: [[ADDWri:%[0-9]+]]:gpr32sp = ADDWri [[PHI]], 1, 0
; CHECK-NEXT: [[COPY5:%[0-9]+]]:gpr32all = COPY [[ADDWri]]
; CHECK-NEXT: B %bb.1
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: bb.3:
; CHECK-NEXT: RET_ReallyLR
bb.0:
liveins: $w0, $s0
%1:gpr32 = COPY $w0
%0:fpr32 = COPY $s0
%3:gpr32all = COPY $wzr
%2:gpr32all = COPY %3:gpr32all
bb.1:
%4:gpr32common = PHI %2:gpr32all, %bb.0, %5:gpr32all, %bb.2
%6:gpr32 = SUBSWrr %4:gpr32common, %1:gpr32, implicit-def $nzcv
Bcc 1, %bb.3, implicit $nzcv
B %bb.2
bb.2:
%7:gpr32 = COPY %0:fpr32
$w0 = COPY %7:gpr32
BL @use_int, implicit-def dead $lr, implicit $sp, implicit $w0, implicit-def $sp
%8:gpr32sp = ADDWri %4:gpr32common, 1, 0
%5:gpr32all = COPY %8:gpr32sp
B %bb.1
bb.3:
RET_ReallyLR
...