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
192 lines
6.4 KiB
LLVM
192 lines
6.4 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc < %s -mtriple=i686-unknown -mattr=+avx512cd | FileCheck %s --check-prefixes=CHECK,X86
|
|
; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+avx512cd | FileCheck %s --check-prefixes=CHECK,X64
|
|
|
|
declare <16 x i32> @llvm.x86.avx512.mask.conflict.d.512(<16 x i32>, <16 x i32>, i16) nounwind readonly
|
|
|
|
define <16 x i32> @test_conflict_d(<16 x i32> %a) {
|
|
; CHECK-LABEL: test_conflict_d:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vpconflictd %zmm0, %zmm0
|
|
; CHECK-NEXT: ret{{[l|q]}}
|
|
%res = call <16 x i32> @llvm.x86.avx512.mask.conflict.d.512(<16 x i32> %a, <16 x i32> undef, i16 -1)
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
define <16 x i32> @test_mask_conflict_d(<16 x i32> %a, <16 x i32> %b, i16 %mask) {
|
|
; X86-LABEL: test_mask_conflict_d:
|
|
; X86: # %bb.0:
|
|
; X86-NEXT: kmovw {{[0-9]+}}(%esp), %k1
|
|
; X86-NEXT: vpconflictd %zmm0, %zmm1 {%k1}
|
|
; X86-NEXT: vmovdqa64 %zmm1, %zmm0
|
|
; X86-NEXT: retl
|
|
;
|
|
; X64-LABEL: test_mask_conflict_d:
|
|
; X64: # %bb.0:
|
|
; X64-NEXT: kmovw %edi, %k1
|
|
; X64-NEXT: vpconflictd %zmm0, %zmm1 {%k1}
|
|
; X64-NEXT: vmovdqa64 %zmm1, %zmm0
|
|
; X64-NEXT: retq
|
|
%res = call <16 x i32> @llvm.x86.avx512.mask.conflict.d.512(<16 x i32> %a, <16 x i32> %b, i16 %mask)
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
define <16 x i32> @test_maskz_conflict_d(<16 x i32> %a, i16 %mask) {
|
|
; X86-LABEL: test_maskz_conflict_d:
|
|
; X86: # %bb.0:
|
|
; X86-NEXT: kmovw {{[0-9]+}}(%esp), %k1
|
|
; X86-NEXT: vpconflictd %zmm0, %zmm0 {%k1} {z}
|
|
; X86-NEXT: retl
|
|
;
|
|
; X64-LABEL: test_maskz_conflict_d:
|
|
; X64: # %bb.0:
|
|
; X64-NEXT: kmovw %edi, %k1
|
|
; X64-NEXT: vpconflictd %zmm0, %zmm0 {%k1} {z}
|
|
; X64-NEXT: retq
|
|
%res = call <16 x i32> @llvm.x86.avx512.mask.conflict.d.512(<16 x i32> %a, <16 x i32> zeroinitializer, i16 %mask)
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
declare <8 x i64> @llvm.x86.avx512.mask.conflict.q.512(<8 x i64>, <8 x i64>, i8) nounwind readonly
|
|
|
|
define <8 x i64> @test_conflict_q(<8 x i64> %a) {
|
|
; CHECK-LABEL: test_conflict_q:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vpconflictq %zmm0, %zmm0
|
|
; CHECK-NEXT: ret{{[l|q]}}
|
|
%res = call <8 x i64> @llvm.x86.avx512.mask.conflict.q.512(<8 x i64> %a, <8 x i64> undef, i8 -1)
|
|
ret <8 x i64> %res
|
|
}
|
|
|
|
define <8 x i64> @test_mask_conflict_q(<8 x i64> %a, <8 x i64> %b, i8 %mask) {
|
|
; X86-LABEL: test_mask_conflict_q:
|
|
; X86: # %bb.0:
|
|
; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax
|
|
; X86-NEXT: kmovw %eax, %k1
|
|
; X86-NEXT: vpconflictq %zmm0, %zmm1 {%k1}
|
|
; X86-NEXT: vmovdqa64 %zmm1, %zmm0
|
|
; X86-NEXT: retl
|
|
;
|
|
; X64-LABEL: test_mask_conflict_q:
|
|
; X64: # %bb.0:
|
|
; X64-NEXT: kmovw %edi, %k1
|
|
; X64-NEXT: vpconflictq %zmm0, %zmm1 {%k1}
|
|
; X64-NEXT: vmovdqa64 %zmm1, %zmm0
|
|
; X64-NEXT: retq
|
|
%res = call <8 x i64> @llvm.x86.avx512.mask.conflict.q.512(<8 x i64> %a, <8 x i64> %b, i8 %mask)
|
|
ret <8 x i64> %res
|
|
}
|
|
|
|
define <8 x i64> @test_maskz_conflict_q(<8 x i64> %a, i8 %mask) {
|
|
; X86-LABEL: test_maskz_conflict_q:
|
|
; X86: # %bb.0:
|
|
; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax
|
|
; X86-NEXT: kmovw %eax, %k1
|
|
; X86-NEXT: vpconflictq %zmm0, %zmm0 {%k1} {z}
|
|
; X86-NEXT: retl
|
|
;
|
|
; X64-LABEL: test_maskz_conflict_q:
|
|
; X64: # %bb.0:
|
|
; X64-NEXT: kmovw %edi, %k1
|
|
; X64-NEXT: vpconflictq %zmm0, %zmm0 {%k1} {z}
|
|
; X64-NEXT: retq
|
|
%res = call <8 x i64> @llvm.x86.avx512.mask.conflict.q.512(<8 x i64> %a, <8 x i64> zeroinitializer, i8 %mask)
|
|
ret <8 x i64> %res
|
|
}
|
|
|
|
define <16 x i32> @test_lzcnt_d(<16 x i32> %a) {
|
|
; CHECK-LABEL: test_lzcnt_d:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vplzcntd %zmm0, %zmm0
|
|
; CHECK-NEXT: ret{{[l|q]}}
|
|
%res = call <16 x i32> @llvm.x86.avx512.mask.lzcnt.d.512(<16 x i32> %a, <16 x i32> zeroinitializer, i16 -1)
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
declare <16 x i32> @llvm.x86.avx512.mask.lzcnt.d.512(<16 x i32>, <16 x i32>, i16) nounwind readonly
|
|
|
|
define <8 x i64> @test_lzcnt_q(<8 x i64> %a) {
|
|
; CHECK-LABEL: test_lzcnt_q:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vplzcntq %zmm0, %zmm0
|
|
; CHECK-NEXT: ret{{[l|q]}}
|
|
%res = call <8 x i64> @llvm.x86.avx512.mask.lzcnt.q.512(<8 x i64> %a, <8 x i64> zeroinitializer, i8 -1)
|
|
ret <8 x i64> %res
|
|
}
|
|
|
|
declare <8 x i64> @llvm.x86.avx512.mask.lzcnt.q.512(<8 x i64>, <8 x i64>, i8) nounwind readonly
|
|
|
|
|
|
define <16 x i32> @test_mask_lzcnt_d(<16 x i32> %a, <16 x i32> %b, i16 %mask) {
|
|
; X86-LABEL: test_mask_lzcnt_d:
|
|
; X86: # %bb.0:
|
|
; X86-NEXT: kmovw {{[0-9]+}}(%esp), %k1
|
|
; X86-NEXT: vplzcntd %zmm0, %zmm1 {%k1}
|
|
; X86-NEXT: vmovdqa64 %zmm1, %zmm0
|
|
; X86-NEXT: retl
|
|
;
|
|
; X64-LABEL: test_mask_lzcnt_d:
|
|
; X64: # %bb.0:
|
|
; X64-NEXT: kmovw %edi, %k1
|
|
; X64-NEXT: vplzcntd %zmm0, %zmm1 {%k1}
|
|
; X64-NEXT: vmovdqa64 %zmm1, %zmm0
|
|
; X64-NEXT: retq
|
|
%res = call <16 x i32> @llvm.x86.avx512.mask.lzcnt.d.512(<16 x i32> %a, <16 x i32> %b, i16 %mask)
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
define <8 x i64> @test_mask_lzcnt_q(<8 x i64> %a, <8 x i64> %b, i8 %mask) {
|
|
; X86-LABEL: test_mask_lzcnt_q:
|
|
; X86: # %bb.0:
|
|
; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax
|
|
; X86-NEXT: kmovw %eax, %k1
|
|
; X86-NEXT: vplzcntq %zmm0, %zmm1 {%k1}
|
|
; X86-NEXT: vmovdqa64 %zmm1, %zmm0
|
|
; X86-NEXT: retl
|
|
;
|
|
; X64-LABEL: test_mask_lzcnt_q:
|
|
; X64: # %bb.0:
|
|
; X64-NEXT: kmovw %edi, %k1
|
|
; X64-NEXT: vplzcntq %zmm0, %zmm1 {%k1}
|
|
; X64-NEXT: vmovdqa64 %zmm1, %zmm0
|
|
; X64-NEXT: retq
|
|
%res = call <8 x i64> @llvm.x86.avx512.mask.lzcnt.q.512(<8 x i64> %a, <8 x i64> %b, i8 %mask)
|
|
ret <8 x i64> %res
|
|
}
|
|
|
|
define <16 x i32> @test_x86_vbroadcastmw_512(i16 %a0) {
|
|
; X86-LABEL: test_x86_vbroadcastmw_512:
|
|
; X86: # %bb.0:
|
|
; X86-NEXT: movzwl {{[0-9]+}}(%esp), %eax
|
|
; X86-NEXT: vpbroadcastd %eax, %zmm0
|
|
; X86-NEXT: retl
|
|
;
|
|
; X64-LABEL: test_x86_vbroadcastmw_512:
|
|
; X64: # %bb.0:
|
|
; X64-NEXT: movzwl %di, %eax
|
|
; X64-NEXT: vpbroadcastd %eax, %zmm0
|
|
; X64-NEXT: retq
|
|
%res = call <16 x i32> @llvm.x86.avx512.broadcastmw.512(i16 %a0)
|
|
ret <16 x i32> %res
|
|
}
|
|
declare <16 x i32> @llvm.x86.avx512.broadcastmw.512(i16)
|
|
|
|
define <8 x i64> @test_x86_broadcastmb_512(i8 %a0) {
|
|
; X86-LABEL: test_x86_broadcastmb_512:
|
|
; X86: # %bb.0:
|
|
; X86-NEXT: movzbl {{[0-9]+}}(%esp), %eax
|
|
; X86-NEXT: vmovd %eax, %xmm0
|
|
; X86-NEXT: vpbroadcastq %xmm0, %zmm0
|
|
; X86-NEXT: retl
|
|
;
|
|
; X64-LABEL: test_x86_broadcastmb_512:
|
|
; X64: # %bb.0:
|
|
; X64-NEXT: movzbl %dil, %eax
|
|
; X64-NEXT: vpbroadcastq %rax, %zmm0
|
|
; X64-NEXT: retq
|
|
%res = call <8 x i64> @llvm.x86.avx512.broadcastmb.512(i8 %a0)
|
|
ret <8 x i64> %res
|
|
}
|
|
declare <8 x i64> @llvm.x86.avx512.broadcastmb.512(i8)
|
|
|