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
109 lines
4.4 KiB
LLVM
109 lines
4.4 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc -mtriple=aarch64 -mattr=+sme2 --pass-remarks-analysis=sme -o /dev/null < %s 2>&1 | FileCheck %s
|
|
|
|
declare void @private_za_callee()
|
|
declare void @private_za_callee_a()
|
|
declare void @private_za_callee_b()
|
|
declare void @private_za_callee_c()
|
|
|
|
declare void @shared_za_callee() "aarch64_inout_za"
|
|
declare void @shared_za_zt0_callee() "aarch64_inout_za" "aarch64_inout_zt0"
|
|
|
|
; Note: These remarks are more useful with source debug info (which gives line numbers for `<unknown>:0:0`).
|
|
|
|
define void @test_lazy_save_1_callee() nounwind "aarch64_inout_za" {
|
|
; CHECK: remark: <unknown>:0:0: lazy save of ZA emitted in 'test_lazy_save_1_callee'
|
|
; CHECK-NEXT: remark: <unknown>:0:0: call to 'private_za_callee' requires ZA save
|
|
call void @private_za_callee()
|
|
ret void
|
|
}
|
|
|
|
define void @test_lazy_save_2_callees() nounwind "aarch64_inout_za" {
|
|
; CHECK: remark: <unknown>:0:0: lazy save of ZA emitted in 'test_lazy_save_2_callees'
|
|
; CHECK-NEXT: remark: <unknown>:0:0: call to 'private_za_callee' requires ZA save
|
|
call void @private_za_callee()
|
|
call void @private_za_callee()
|
|
ret void
|
|
}
|
|
|
|
define float @test_lazy_save_expanded_intrinsic(float %a) nounwind "aarch64_inout_za" {
|
|
; CHECK: remark: <unknown>:0:0: lazy save of ZA emitted in 'test_lazy_save_expanded_intrinsic'
|
|
; CHECK-NEXT: remark: <unknown>:0:0: call to 'cosf' requires ZA save
|
|
%res = call float @llvm.cos.f32(float %a)
|
|
ret float %res
|
|
}
|
|
|
|
define void @test_lazy_save_multiple_paths(i1 %a) "aarch64_inout_za" {
|
|
; CHECK: remark: <unknown>:0:0: lazy save of ZA emitted in 'test_lazy_save_multiple_paths'
|
|
; CHECK-NEXT: remark: <unknown>:0:0: call to 'private_za_callee_b' requires ZA save
|
|
; CHECK-NEXT: remark: <unknown>:0:0: call to 'private_za_callee_a' requires ZA save
|
|
entry:
|
|
br i1 %a, label %if.end, label %if.else
|
|
|
|
if.else:
|
|
call void @private_za_callee_a()
|
|
br label %if.end
|
|
|
|
if.end:
|
|
call void @private_za_callee_b()
|
|
; The new lowering won't report this call as the save is already needed due to
|
|
; the call to `private_za_callee_b()`.
|
|
call void @private_za_callee_c()
|
|
|
|
ret void
|
|
}
|
|
|
|
define void @test_lazy_save_with_zt0() "aarch64_inout_za" "aarch64_inout_zt0"
|
|
{
|
|
; CHECK: remark: <unknown>:0:0: spill of ZT0 emitted in 'test_lazy_save_with_zt0'
|
|
; CHECK-NEXT: remark: <unknown>:0:0: call to 'shared_za_callee' requires ZT0 save
|
|
; CHECK-NEXT: remark: <unknown>:0:0: lazy save of ZA emitted in 'test_lazy_save_with_zt0'
|
|
; CHECK-NEXT: remark: <unknown>:0:0: call to 'private_za_callee' requires ZA save
|
|
call void @shared_za_callee() ; Save ZT0 (remark ZT0 spill)
|
|
call void @private_za_callee() ; Save ZA (remark ZA save)
|
|
ret void
|
|
}
|
|
|
|
define void @test_lazy_save_with_zt0_reload() "aarch64_inout_za" "aarch64_inout_zt0"
|
|
{
|
|
; CHECK: remark: <unknown>:0:0: spill of ZT0 emitted in 'test_lazy_save_with_zt0_reload'
|
|
; CHECK-NEXT: remark: <unknown>:0:0: call to 'shared_za_callee' requires ZT0 save
|
|
; CHECK-NEXT: remark: <unknown>:0:0: spill of ZT0 emitted in 'test_lazy_save_with_zt0_reload'
|
|
; CHECK-NEXT: remark: <unknown>:0:0: lazy save of ZA emitted in 'test_lazy_save_with_zt0_reload'
|
|
; CHECK-NEXT: remark: <unknown>:0:0: call to 'private_za_callee' requires ZA save
|
|
call void @shared_za_callee() ; Save ZT0 (remark ZT0 spill)
|
|
call void @shared_za_zt0_callee() ; Reload ZT0
|
|
call void @private_za_callee() ; Save ZA, ZT0 (remark ZT0 spill and ZA save)
|
|
ret void
|
|
}
|
|
|
|
define void @test_za_merge_paths(i1 %a) "aarch64_za_state_agnostic" {
|
|
; CHECK: remark: <unknown>:0:0: full save of ZA emitted in 'test_za_merge_paths'
|
|
; CHECK-NEXT: remark: <unknown>:0:0: call to 'private_za_callee_b' requires ZA save
|
|
; CHECK-NEXT: remark: <unknown>:0:0: call to 'private_za_callee_a' requires ZA save
|
|
entry:
|
|
br i1 %a, label %if.end, label %if.else
|
|
|
|
if.else:
|
|
call void @private_za_callee_a()
|
|
br label %exit
|
|
|
|
if.end:
|
|
call void @private_za_callee_b()
|
|
br label %exit
|
|
|
|
exit:
|
|
; The remarks won't report this call as the save is already needed due to
|
|
; the call to `private_za_callee_*()` calls on both paths to this BB.
|
|
call void @private_za_callee_c()
|
|
|
|
ret void
|
|
}
|
|
|
|
define void @test_lazy_save_function_ptr_callee(ptr %private_za_callee) nounwind "aarch64_inout_za" {
|
|
; CHECK: remark: <unknown>:0:0: lazy save of ZA emitted in 'test_lazy_save_function_ptr_callee'
|
|
; CHECK-NEXT: remark: <unknown>:0:0: call requires ZA save
|
|
call void %private_za_callee()
|
|
ret void
|
|
}
|