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
80 lines
3.1 KiB
ArmAsm
80 lines
3.1 KiB
ArmAsm
# This test checks that reordering blocks which contain short range
|
|
# conditional branches works even if the target is beyond 1KB and
|
|
# the branch condition may or may not be inverted.
|
|
|
|
# REQUIRES: system-linux, asserts
|
|
|
|
# RUN: %clang %cflags -march=armv9-a+cmpbr -Wl,-q %s -o %t -DNUM_NOPS=0 -DIMM=0
|
|
# RUN: link_fdata --no-lbr %s %t %t.fdata
|
|
# RUN: llvm-strip --strip-unneeded %t
|
|
# RUN: llvm-bolt %t -o %t.bolt --data %t.fdata --reorder-blocks=ext-tsp --compact-code-model
|
|
# RUN: llvm-objdump -d %t.bolt | FileCheck %s
|
|
|
|
# RUN: %clang %cflags -march=armv9-a+cmpbr -Wl,-q %s -o %t -DNUM_NOPS=256 -DIMM=0
|
|
# RUN: link_fdata --no-lbr %s %t %t.fdata
|
|
# RUN: llvm-strip --strip-unneeded %t
|
|
# RUN: llvm-bolt %t -o %t.bolt --data %t.fdata --reorder-blocks=ext-tsp --compact-code-model --keep-nops
|
|
# RUN: llvm-objdump -d %t.bolt | FileCheck %s --check-prefix=BEYOND-1KB
|
|
|
|
# RUN: %clang %cflags -march=armv9-a+cmpbr -Wl,-q %s -o %t -DNUM_NOPS=256 -DIMM=63
|
|
# RUN: link_fdata --no-lbr %s %t %t.fdata
|
|
# RUN: llvm-strip --strip-unneeded %t
|
|
# RUN: llvm-bolt %t -o %t.bolt --data %t.fdata --reorder-blocks=ext-tsp --compact-code-model --keep-nops
|
|
# RUN: llvm-objdump -d %t.bolt | FileCheck %s --check-prefix=BEYOND-1KB-IRREVERSIBLE
|
|
|
|
.globl reorder_blocks
|
|
.type reorder_blocks, %function
|
|
reorder_blocks:
|
|
.entry:
|
|
# FDATA: 1 reorder_blocks #.entry# 10
|
|
cbgt x0, #IMM, .cold_exit
|
|
.fall_through:
|
|
# FDATA: 1 reorder_blocks #.fall_through# 10
|
|
b .hot_exit
|
|
.cold_exit:
|
|
# FDATA: 1 reorder_blocks #.cold_exit# 1
|
|
mov x0, #1
|
|
ret
|
|
.hot_exit:
|
|
# FDATA: 1 reorder_blocks #.hot_exit# 10
|
|
.rept NUM_NOPS
|
|
nop
|
|
.endr
|
|
mov x0, #2
|
|
ret
|
|
|
|
## Force relocation mode.
|
|
.reloc 0, R_AARCH64_NONE
|
|
|
|
# CHECK: Disassembly of section .text:
|
|
|
|
# CHECK: <reorder_blocks>:
|
|
# CHECK-NEXT: {{.*}} cbgt x0, #0x0, 0x[[ADDR:[0-9a-f]+]] <{{.*}}>
|
|
# CHECK-NEXT: {{.*}} mov x0, #0x2 // =2
|
|
# CHECK-NEXT: {{.*}} ret
|
|
# CHECK-NEXT: [[ADDR]]: {{.*}} mov x0, #0x1 // =1
|
|
# CHECK-NEXT: {{.*}} ret
|
|
|
|
# BEYOND-1KB: Disassembly of section .text:
|
|
|
|
# BEYOND-1KB: <reorder_blocks>:
|
|
# BEYOND-1KB-NEXT: {{.*}} cblt x0, #0x1, 0x[[ADDR0:[0-9a-f]+]] <{{.*}}>
|
|
# BEYOND-1KB-NEXT: {{.*}} b 0x[[ADDR1:[0-9a-f]+]] <{{.*}}>
|
|
# BEYOND-1KB-NEXT: [[ADDR0]]: {{.*}} nop
|
|
# BEYOND-1KB: {{.*}} mov x0, #0x2 // =2
|
|
# BEYOND-1KB-NEXT: {{.*}} ret
|
|
# BEYOND-1KB-NEXT: [[ADDR1]]: {{.*}} mov x0, #0x1 // =1
|
|
# BEYOND-1KB-NEXT: {{.*}} ret
|
|
|
|
# BEYOND-1KB-IRREVERSIBLE: Disassembly of section .text:
|
|
|
|
# BEYOND-1KB-IRREVERSIBLE: <reorder_blocks>:
|
|
# BEYOND-1KB-IRREVERSIBLE-NEXT: {{.*}} cbgt x0, #0x3f, 0x[[ADDR0:[0-9a-f]+]] <{{.*}}>
|
|
# BEYOND-1KB-IRREVERSIBLE-NEXT: {{.*}} b 0x[[ADDR1:[0-9a-f]+]] <{{.*}}>
|
|
# BEYOND-1KB-IRREVERSIBLE-NEXT: [[ADDR0]]: {{.*}} b 0x[[ADDR2:[0-9a-f]+]] <{{.*}}>
|
|
# BEYOND-1KB-IRREVERSIBLE-NEXT: [[ADDR1]]: {{.*}} nop
|
|
# BEYOND-1KB-IRREVERSIBLE: {{.*}} mov x0, #0x2 // =2
|
|
# BEYOND-1KB-IRREVERSIBLE-NEXT: {{.*}} ret
|
|
# BEYOND-1KB-IRREVERSIBLE-NEXT: [[ADDR2]]: {{.*}} mov x0, #0x1 // =1
|
|
# BEYOND-1KB-IRREVERSIBLE-NEXT: {{.*}} ret
|