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
36 lines
1.9 KiB
LLVM
36 lines
1.9 KiB
LLVM
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s
|
|
|
|
@e = dso_local global [8 x i32] [i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8], align 16
|
|
@d = dso_local global [8 x i32] [i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1], align 16
|
|
|
|
; The global 'e' has 16 byte alignment, so make sure we don't generate an
|
|
; aligned 32-byte load instruction when we combine the load+insert sequence.
|
|
|
|
define dso_local i32 @subb() nounwind ssp {
|
|
; CHECK-LABEL: subb:
|
|
; CHECK: vmovups e(%rip), %ymm
|
|
entry:
|
|
%0 = load i32, ptr getelementptr inbounds ([8 x i32], ptr @e, i64 0, i64 7), align 4
|
|
%1 = load i32, ptr getelementptr inbounds ([8 x i32], ptr @e, i64 0, i64 6), align 8
|
|
%2 = load i32, ptr getelementptr inbounds ([8 x i32], ptr @e, i64 0, i64 5), align 4
|
|
%3 = load i32, ptr getelementptr inbounds ([8 x i32], ptr @e, i64 0, i64 4), align 16
|
|
%4 = load i32, ptr getelementptr inbounds ([8 x i32], ptr @e, i64 0, i64 3), align 4
|
|
%5 = load i32, ptr getelementptr inbounds ([8 x i32], ptr @e, i64 0, i64 2), align 8
|
|
%6 = load i32, ptr getelementptr inbounds ([8 x i32], ptr @e, i64 0, i64 1), align 4
|
|
%7 = load i32, ptr @e, align 16
|
|
%vecinit.i = insertelement <8 x i32> undef, i32 %7, i32 0
|
|
%vecinit1.i = insertelement <8 x i32> %vecinit.i, i32 %6, i32 1
|
|
%vecinit2.i = insertelement <8 x i32> %vecinit1.i, i32 %5, i32 2
|
|
%vecinit3.i = insertelement <8 x i32> %vecinit2.i, i32 %4, i32 3
|
|
%vecinit4.i = insertelement <8 x i32> %vecinit3.i, i32 %3, i32 4
|
|
%vecinit5.i = insertelement <8 x i32> %vecinit4.i, i32 %2, i32 5
|
|
%vecinit6.i = insertelement <8 x i32> %vecinit5.i, i32 %1, i32 6
|
|
%vecinit7.i = insertelement <8 x i32> %vecinit6.i, i32 %0, i32 7
|
|
%8 = bitcast <8 x i32> %vecinit7.i to <32 x i8>
|
|
tail call void @llvm.x86.avx.storeu.dq.256(ptr @d, <32 x i8> %8)
|
|
ret i32 0
|
|
}
|
|
|
|
declare void @llvm.x86.avx.storeu.dq.256(ptr, <32 x i8>) nounwind
|
|
|