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
96 lines
3.7 KiB
LLVM
96 lines
3.7 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 3
|
|
; RUN: llc < %s -mtriple=aarch64 | FileCheck %s --check-prefix=CHECKLE
|
|
; RUN: llc < %s -mtriple=aarch64_be | FileCheck %s --check-prefix=CHECKBE
|
|
|
|
define <4 x i16> @test_reconstructshuffle(<16 x i8> %a, <16 x i8> %b) nounwind {
|
|
; CHECKLE-LABEL: test_reconstructshuffle:
|
|
; CHECKLE: // %bb.0:
|
|
; CHECKLE-NEXT: mov b2, v0.b[3]
|
|
; CHECKLE-NEXT: mov d1, v1.d[1]
|
|
; CHECKLE-NEXT: mov v2.b[2], v0.b[2]
|
|
; CHECKLE-NEXT: mov v2.b[4], v0.b[1]
|
|
; CHECKLE-NEXT: mov v2.b[6], v0.b[0]
|
|
; CHECKLE-NEXT: zip2 v0.8b, v1.8b, v0.8b
|
|
; CHECKLE-NEXT: add v0.4h, v2.4h, v0.4h
|
|
; CHECKLE-NEXT: bic v0.4h, #255, lsl #8
|
|
; CHECKLE-NEXT: ret
|
|
;
|
|
; CHECKBE-LABEL: test_reconstructshuffle:
|
|
; CHECKBE: // %bb.0:
|
|
; CHECKBE-NEXT: rev64 v0.16b, v0.16b
|
|
; CHECKBE-NEXT: rev64 v1.16b, v1.16b
|
|
; CHECKBE-NEXT: ext v0.16b, v0.16b, v0.16b, #8
|
|
; CHECKBE-NEXT: ext v1.16b, v1.16b, v1.16b, #8
|
|
; CHECKBE-NEXT: mov b2, v0.b[3]
|
|
; CHECKBE-NEXT: mov d1, v1.d[1]
|
|
; CHECKBE-NEXT: mov v2.b[2], v0.b[2]
|
|
; CHECKBE-NEXT: mov v2.b[4], v0.b[1]
|
|
; CHECKBE-NEXT: mov v2.b[6], v0.b[0]
|
|
; CHECKBE-NEXT: zip2 v0.8b, v1.8b, v0.8b
|
|
; CHECKBE-NEXT: add v0.4h, v2.4h, v0.4h
|
|
; CHECKBE-NEXT: bic v0.4h, #255, lsl #8
|
|
; CHECKBE-NEXT: rev64 v0.4h, v0.4h
|
|
; CHECKBE-NEXT: ret
|
|
%tmp1 = shufflevector <16 x i8> %a, <16 x i8> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
|
|
%tmp2 = shufflevector <16 x i8> %b, <16 x i8> poison, <4 x i32> <i32 12, i32 13, i32 14, i32 15>
|
|
%tmp3 = add <4 x i8> %tmp1, %tmp2
|
|
%tmp4 = zext <4 x i8> %tmp3 to <4 x i16>
|
|
ret <4 x i16> %tmp4
|
|
}
|
|
|
|
; a shufflevector of even elements can become just xtn on little-endian, but not on big-endian.
|
|
define <8 x i8> @xtn_shuffle_even_v16i8(<16 x i8> %a) {
|
|
; CHECKLE-LABEL: xtn_shuffle_even_v16i8:
|
|
; CHECKLE: // %bb.0:
|
|
; CHECKLE-NEXT: xtn v0.8b, v0.8h
|
|
; CHECKLE-NEXT: ret
|
|
;
|
|
; CHECKBE-LABEL: xtn_shuffle_even_v16i8:
|
|
; CHECKBE: // %bb.0:
|
|
; CHECKBE-NEXT: rev64 v0.16b, v0.16b
|
|
; CHECKBE-NEXT: ext v0.16b, v0.16b, v0.16b, #8
|
|
; CHECKBE-NEXT: uzp1 v0.16b, v0.16b, v0.16b
|
|
; CHECKBE-NEXT: rev64 v0.8b, v0.8b
|
|
; CHECKBE-NEXT: ret
|
|
%r = shufflevector <16 x i8> %a, <16 x i8> poison, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
|
|
ret <8 x i8> %r
|
|
}
|
|
|
|
; a shufflevector of even elements can become just xtn on little-endian, but not on big-endian.
|
|
define <4 x i16> @xtn_shuffle_even_v8i16(<8 x i16> %a) {
|
|
; CHECKLE-LABEL: xtn_shuffle_even_v8i16:
|
|
; CHECKLE: // %bb.0:
|
|
; CHECKLE-NEXT: xtn v0.4h, v0.4s
|
|
; CHECKLE-NEXT: ret
|
|
;
|
|
; CHECKBE-LABEL: xtn_shuffle_even_v8i16:
|
|
; CHECKBE: // %bb.0:
|
|
; CHECKBE-NEXT: rev64 v0.8h, v0.8h
|
|
; CHECKBE-NEXT: ext v0.16b, v0.16b, v0.16b, #8
|
|
; CHECKBE-NEXT: uzp1 v0.8h, v0.8h, v0.8h
|
|
; CHECKBE-NEXT: rev64 v0.4h, v0.4h
|
|
; CHECKBE-NEXT: ret
|
|
%r = shufflevector <8 x i16> %a, <8 x i16> poison, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
|
|
ret <4 x i16> %r
|
|
}
|
|
|
|
; a shufflevector of even elements can become just xtn on little-endian, but not on big-endian.
|
|
define <4 x i32> @xtn_shuffle_uzp1_poison_v4i32(<4 x i32> %a) {
|
|
; CHECKLE-LABEL: xtn_shuffle_uzp1_poison_v4i32:
|
|
; CHECKLE: // %bb.0:
|
|
; CHECKLE-NEXT: xtn v0.2s, v0.2d
|
|
; CHECKLE-NEXT: ret
|
|
;
|
|
; CHECKBE-LABEL: xtn_shuffle_uzp1_poison_v4i32:
|
|
; CHECKBE: // %bb.0:
|
|
; CHECKBE-NEXT: rev64 v0.4s, v0.4s
|
|
; CHECKBE-NEXT: ext v0.16b, v0.16b, v0.16b, #8
|
|
; CHECKBE-NEXT: uzp1 v0.4s, v0.4s, v0.4s
|
|
; CHECKBE-NEXT: rev64 v0.4s, v0.4s
|
|
; CHECKBE-NEXT: ext v0.16b, v0.16b, v0.16b, #8
|
|
; CHECKBE-NEXT: ret
|
|
%r = shufflevector <4 x i32> %a, <4 x i32> poison, <4 x i32> <i32 0, i32 2, i32 poison, i32 poison>
|
|
ret <4 x i32> %r
|
|
|
|
}
|