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
110 lines
4.6 KiB
LLVM
110 lines
4.6 KiB
LLVM
; RUN: llc -mtriple=arm-eabi -mattr=+neon %s -o - -verify-machineinstrs | FileCheck %s
|
|
|
|
%struct.__neon_int8x8x2_t = type { <8 x i8>, <8 x i8> }
|
|
%struct.__neon_int8x8x3_t = type { <8 x i8>, <8 x i8>, <8 x i8> }
|
|
%struct.__neon_int8x8x4_t = type { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }
|
|
|
|
define <8 x i8> @vtbl1(ptr %A, ptr %B) nounwind {
|
|
;CHECK-LABEL: vtbl1:
|
|
;CHECK: vtbl.8
|
|
%tmp1 = load <8 x i8>, ptr %A
|
|
%tmp2 = load <8 x i8>, ptr %B
|
|
%tmp3 = call <8 x i8> @llvm.arm.neon.vtbl1(<8 x i8> %tmp1, <8 x i8> %tmp2)
|
|
ret <8 x i8> %tmp3
|
|
}
|
|
|
|
define <8 x i8> @vtbl2(ptr %A, ptr %B) nounwind {
|
|
;CHECK-LABEL: vtbl2:
|
|
;CHECK: vtbl.8
|
|
%tmp1 = load <8 x i8>, ptr %A
|
|
%tmp2 = load %struct.__neon_int8x8x2_t, ptr %B
|
|
%tmp3 = extractvalue %struct.__neon_int8x8x2_t %tmp2, 0
|
|
%tmp4 = extractvalue %struct.__neon_int8x8x2_t %tmp2, 1
|
|
%tmp5 = call <8 x i8> @llvm.arm.neon.vtbl2(<8 x i8> %tmp1, <8 x i8> %tmp3, <8 x i8> %tmp4)
|
|
ret <8 x i8> %tmp5
|
|
}
|
|
|
|
define <8 x i8> @vtbl3(ptr %A, ptr %B) nounwind {
|
|
;CHECK-LABEL: vtbl3:
|
|
;CHECK: vtbl.8
|
|
%tmp1 = load <8 x i8>, ptr %A
|
|
%tmp2 = load %struct.__neon_int8x8x3_t, ptr %B
|
|
%tmp3 = extractvalue %struct.__neon_int8x8x3_t %tmp2, 0
|
|
%tmp4 = extractvalue %struct.__neon_int8x8x3_t %tmp2, 1
|
|
%tmp5 = extractvalue %struct.__neon_int8x8x3_t %tmp2, 2
|
|
%tmp6 = call <8 x i8> @llvm.arm.neon.vtbl3(<8 x i8> %tmp1, <8 x i8> %tmp3, <8 x i8> %tmp4, <8 x i8> %tmp5)
|
|
ret <8 x i8> %tmp6
|
|
}
|
|
|
|
define <8 x i8> @vtbl4(ptr %A, ptr %B) nounwind {
|
|
;CHECK-LABEL: vtbl4:
|
|
;CHECK: vtbl.8
|
|
%tmp1 = load <8 x i8>, ptr %A
|
|
%tmp2 = load %struct.__neon_int8x8x4_t, ptr %B
|
|
%tmp3 = extractvalue %struct.__neon_int8x8x4_t %tmp2, 0
|
|
%tmp4 = extractvalue %struct.__neon_int8x8x4_t %tmp2, 1
|
|
%tmp5 = extractvalue %struct.__neon_int8x8x4_t %tmp2, 2
|
|
%tmp6 = extractvalue %struct.__neon_int8x8x4_t %tmp2, 3
|
|
%tmp7 = call <8 x i8> @llvm.arm.neon.vtbl4(<8 x i8> %tmp1, <8 x i8> %tmp3, <8 x i8> %tmp4, <8 x i8> %tmp5, <8 x i8> %tmp6)
|
|
ret <8 x i8> %tmp7
|
|
}
|
|
|
|
define <8 x i8> @vtbx1(ptr %A, ptr %B, ptr %C) nounwind {
|
|
;CHECK-LABEL: vtbx1:
|
|
;CHECK: vtbx.8
|
|
%tmp1 = load <8 x i8>, ptr %A
|
|
%tmp2 = load <8 x i8>, ptr %B
|
|
%tmp3 = load <8 x i8>, ptr %C
|
|
%tmp4 = call <8 x i8> @llvm.arm.neon.vtbx1(<8 x i8> %tmp1, <8 x i8> %tmp2, <8 x i8> %tmp3)
|
|
ret <8 x i8> %tmp4
|
|
}
|
|
|
|
define <8 x i8> @vtbx2(ptr %A, ptr %B, ptr %C) nounwind {
|
|
;CHECK-LABEL: vtbx2:
|
|
;CHECK: vtbx.8
|
|
%tmp1 = load <8 x i8>, ptr %A
|
|
%tmp2 = load %struct.__neon_int8x8x2_t, ptr %B
|
|
%tmp3 = extractvalue %struct.__neon_int8x8x2_t %tmp2, 0
|
|
%tmp4 = extractvalue %struct.__neon_int8x8x2_t %tmp2, 1
|
|
%tmp5 = load <8 x i8>, ptr %C
|
|
%tmp6 = call <8 x i8> @llvm.arm.neon.vtbx2(<8 x i8> %tmp1, <8 x i8> %tmp3, <8 x i8> %tmp4, <8 x i8> %tmp5)
|
|
ret <8 x i8> %tmp6
|
|
}
|
|
|
|
define <8 x i8> @vtbx3(ptr %A, ptr %B, ptr %C) nounwind {
|
|
;CHECK-LABEL: vtbx3:
|
|
;CHECK: vtbx.8
|
|
%tmp1 = load <8 x i8>, ptr %A
|
|
%tmp2 = load %struct.__neon_int8x8x3_t, ptr %B
|
|
%tmp3 = extractvalue %struct.__neon_int8x8x3_t %tmp2, 0
|
|
%tmp4 = extractvalue %struct.__neon_int8x8x3_t %tmp2, 1
|
|
%tmp5 = extractvalue %struct.__neon_int8x8x3_t %tmp2, 2
|
|
%tmp6 = load <8 x i8>, ptr %C
|
|
%tmp7 = call <8 x i8> @llvm.arm.neon.vtbx3(<8 x i8> %tmp1, <8 x i8> %tmp3, <8 x i8> %tmp4, <8 x i8> %tmp5, <8 x i8> %tmp6)
|
|
ret <8 x i8> %tmp7
|
|
}
|
|
|
|
define <8 x i8> @vtbx4(ptr %A, ptr %B, ptr %C) nounwind {
|
|
;CHECK-LABEL: vtbx4:
|
|
;CHECK: vtbx.8
|
|
%tmp1 = load <8 x i8>, ptr %A
|
|
%tmp2 = load %struct.__neon_int8x8x4_t, ptr %B
|
|
%tmp3 = extractvalue %struct.__neon_int8x8x4_t %tmp2, 0
|
|
%tmp4 = extractvalue %struct.__neon_int8x8x4_t %tmp2, 1
|
|
%tmp5 = extractvalue %struct.__neon_int8x8x4_t %tmp2, 2
|
|
%tmp6 = extractvalue %struct.__neon_int8x8x4_t %tmp2, 3
|
|
%tmp7 = load <8 x i8>, ptr %C
|
|
%tmp8 = call <8 x i8> @llvm.arm.neon.vtbx4(<8 x i8> %tmp1, <8 x i8> %tmp3, <8 x i8> %tmp4, <8 x i8> %tmp5, <8 x i8> %tmp6, <8 x i8> %tmp7)
|
|
ret <8 x i8> %tmp8
|
|
}
|
|
|
|
declare <8 x i8> @llvm.arm.neon.vtbl1(<8 x i8>, <8 x i8>) nounwind readnone
|
|
declare <8 x i8> @llvm.arm.neon.vtbl2(<8 x i8>, <8 x i8>, <8 x i8>) nounwind readnone
|
|
declare <8 x i8> @llvm.arm.neon.vtbl3(<8 x i8>, <8 x i8>, <8 x i8>, <8 x i8>) nounwind readnone
|
|
declare <8 x i8> @llvm.arm.neon.vtbl4(<8 x i8>, <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8>) nounwind readnone
|
|
|
|
declare <8 x i8> @llvm.arm.neon.vtbx1(<8 x i8>, <8 x i8>, <8 x i8>) nounwind readnone
|
|
declare <8 x i8> @llvm.arm.neon.vtbx2(<8 x i8>, <8 x i8>, <8 x i8>, <8 x i8>) nounwind readnone
|
|
declare <8 x i8> @llvm.arm.neon.vtbx3(<8 x i8>, <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8>) nounwind readnone
|
|
declare <8 x i8> @llvm.arm.neon.vtbx4(<8 x i8>, <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8>) nounwind readnone
|