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
89 lines
2.8 KiB
LLVM
89 lines
2.8 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc < %s -mtriple=x86_64-- -mcpu=x86-64 | FileCheck %s --check-prefixes=SSE2
|
|
; RUN: llc < %s -mtriple=x86_64-- -mcpu=x86-64-v2 | FileCheck %s --check-prefixes=SSE4
|
|
; RUN: llc < %s -mtriple=x86_64-- -mcpu=x86-64-v3 | FileCheck %s --check-prefixes=AVX2
|
|
; RUN: llc < %s -mtriple=x86_64-- -mcpu=x86-64-v4 | FileCheck %s --check-prefixes=AVX512
|
|
|
|
define void @pr166534(ptr %pa, ptr %pb, ptr %pc, ptr %pd) {
|
|
; SSE2-LABEL: pr166534:
|
|
; SSE2: # %bb.0: # %entry
|
|
; SSE2-NEXT: movdqu (%rdi), %xmm0
|
|
; SSE2-NEXT: movdqu (%rsi), %xmm1
|
|
; SSE2-NEXT: pcmpeqb %xmm0, %xmm1
|
|
; SSE2-NEXT: pmovmskb %xmm1, %esi
|
|
; SSE2-NEXT: xorl %eax, %eax
|
|
; SSE2-NEXT: cmpl $65535, %esi # imm = 0xFFFF
|
|
; SSE2-NEXT: sete %al
|
|
; SSE2-NEXT: orq %rax, (%rdx)
|
|
; SSE2-NEXT: cmpl $65535, %esi # imm = 0xFFFF
|
|
; SSE2-NEXT: jne .LBB0_2
|
|
; SSE2-NEXT: # %bb.1: # %if.then
|
|
; SSE2-NEXT: orq %rax, (%rcx)
|
|
; SSE2-NEXT: .LBB0_2: # %if.end
|
|
; SSE2-NEXT: retq
|
|
;
|
|
; SSE4-LABEL: pr166534:
|
|
; SSE4: # %bb.0: # %entry
|
|
; SSE4-NEXT: movdqu (%rdi), %xmm0
|
|
; SSE4-NEXT: movdqu (%rsi), %xmm1
|
|
; SSE4-NEXT: pxor %xmm0, %xmm1
|
|
; SSE4-NEXT: xorl %eax, %eax
|
|
; SSE4-NEXT: ptest %xmm1, %xmm1
|
|
; SSE4-NEXT: sete %al
|
|
; SSE4-NEXT: orq %rax, (%rdx)
|
|
; SSE4-NEXT: ptest %xmm1, %xmm1
|
|
; SSE4-NEXT: jne .LBB0_2
|
|
; SSE4-NEXT: # %bb.1: # %if.then
|
|
; SSE4-NEXT: orq %rax, (%rcx)
|
|
; SSE4-NEXT: .LBB0_2: # %if.end
|
|
; SSE4-NEXT: retq
|
|
;
|
|
; AVX2-LABEL: pr166534:
|
|
; AVX2: # %bb.0: # %entry
|
|
; AVX2-NEXT: vmovdqu (%rdi), %xmm0
|
|
; AVX2-NEXT: vpxor (%rsi), %xmm0, %xmm0
|
|
; AVX2-NEXT: xorl %eax, %eax
|
|
; AVX2-NEXT: vptest %xmm0, %xmm0
|
|
; AVX2-NEXT: sete %al
|
|
; AVX2-NEXT: orq %rax, (%rdx)
|
|
; AVX2-NEXT: vptest %xmm0, %xmm0
|
|
; AVX2-NEXT: jne .LBB0_2
|
|
; AVX2-NEXT: # %bb.1: # %if.then
|
|
; AVX2-NEXT: orq %rax, (%rcx)
|
|
; AVX2-NEXT: .LBB0_2: # %if.end
|
|
; AVX2-NEXT: retq
|
|
;
|
|
; AVX512-LABEL: pr166534:
|
|
; AVX512: # %bb.0: # %entry
|
|
; AVX512-NEXT: vmovdqu (%rdi), %xmm0
|
|
; AVX512-NEXT: vpxor (%rsi), %xmm0, %xmm0
|
|
; AVX512-NEXT: xorl %eax, %eax
|
|
; AVX512-NEXT: vptest %xmm0, %xmm0
|
|
; AVX512-NEXT: sete %al
|
|
; AVX512-NEXT: orq %rax, (%rdx)
|
|
; AVX512-NEXT: vptest %xmm0, %xmm0
|
|
; AVX512-NEXT: jne .LBB0_2
|
|
; AVX512-NEXT: # %bb.1: # %if.then
|
|
; AVX512-NEXT: orq %rax, (%rcx)
|
|
; AVX512-NEXT: .LBB0_2: # %if.end
|
|
; AVX512-NEXT: retq
|
|
entry:
|
|
%a = load i128, ptr %pa, align 8
|
|
%b = load i128, ptr %pb, align 8
|
|
%cmp = icmp eq i128 %a, %b
|
|
%conv1 = zext i1 %cmp to i128
|
|
%c = load i128, ptr %pc, align 8
|
|
%or = or i128 %c, %conv1
|
|
store i128 %or, ptr %pc, align 8
|
|
br i1 %cmp, label %if.then, label %if.end
|
|
|
|
if.then:
|
|
%d = load i128, ptr %pd, align 8
|
|
%or7 = or i128 %d, %conv1
|
|
store i128 %or7, ptr %pd, align 8
|
|
br label %if.end
|
|
|
|
if.end:
|
|
ret void
|
|
}
|