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
164 lines
5.7 KiB
LLVM
164 lines
5.7 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 6
|
|
; RUN: llc -mtriple=aarch64 %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-SD
|
|
; RUN: llc -mtriple=aarch64 -global-isel %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-GI
|
|
|
|
define <16 x i8> @vqneg_test16(<16 x i8> %A) nounwind {
|
|
; CHECK-LABEL: vqneg_test16:
|
|
; CHECK: // %bb.0: // %entry
|
|
; CHECK-NEXT: movi v1.16b, #128
|
|
; CHECK-NEXT: movi v2.16b, #127
|
|
; CHECK-NEXT: cmeq v1.16b, v0.16b, v1.16b
|
|
; CHECK-NEXT: neg v0.16b, v0.16b
|
|
; CHECK-NEXT: bit v0.16b, v2.16b, v1.16b
|
|
; CHECK-NEXT: ret
|
|
entry:
|
|
%0 = icmp eq <16 x i8> %A, <i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128>
|
|
%1 = sub nsw <16 x i8> zeroinitializer, %A
|
|
%2 = select <16 x i1> %0, <16 x i8> <i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127>, <16 x i8> %1
|
|
ret <16 x i8> %2
|
|
}
|
|
|
|
define <8 x i16> @vqneg_test8(<8 x i16> %A) nounwind {
|
|
; CHECK-SD-LABEL: vqneg_test8:
|
|
; CHECK-SD: // %bb.0: // %entry
|
|
; CHECK-SD-NEXT: movi v1.8h, #128, lsl #8
|
|
; CHECK-SD-NEXT: neg v2.8h, v0.8h
|
|
; CHECK-SD-NEXT: cmeq v0.8h, v0.8h, v1.8h
|
|
; CHECK-SD-NEXT: bic v1.16b, v2.16b, v0.16b
|
|
; CHECK-SD-NEXT: bic v0.8h, #128, lsl #8
|
|
; CHECK-SD-NEXT: orr v0.16b, v0.16b, v1.16b
|
|
; CHECK-SD-NEXT: ret
|
|
;
|
|
; CHECK-GI-LABEL: vqneg_test8:
|
|
; CHECK-GI: // %bb.0: // %entry
|
|
; CHECK-GI-NEXT: movi v1.8h, #128, lsl #8
|
|
; CHECK-GI-NEXT: mvni v2.8h, #128, lsl #8
|
|
; CHECK-GI-NEXT: cmeq v1.8h, v0.8h, v1.8h
|
|
; CHECK-GI-NEXT: neg v0.8h, v0.8h
|
|
; CHECK-GI-NEXT: bit v0.16b, v2.16b, v1.16b
|
|
; CHECK-GI-NEXT: ret
|
|
entry:
|
|
%0 = icmp eq <8 x i16> %A, <i16 -32768, i16 -32768, i16 -32768, i16 -32768, i16 -32768, i16 -32768, i16 -32768, i16 -32768>
|
|
%1 = sub nsw <8 x i16> zeroinitializer, %A
|
|
%2 = select <8 x i1> %0, <8 x i16> <i16 32767, i16 32767, i16 32767, i16 32767, i16 32767, i16 32767, i16 32767, i16 32767>, <8 x i16> %1
|
|
ret <8 x i16> %2
|
|
}
|
|
|
|
define <4 x i32> @vqneg_test4(<4 x i32> %A) nounwind {
|
|
; CHECK-SD-LABEL: vqneg_test4:
|
|
; CHECK-SD: // %bb.0: // %entry
|
|
; CHECK-SD-NEXT: movi v1.4s, #128, lsl #24
|
|
; CHECK-SD-NEXT: neg v2.4s, v0.4s
|
|
; CHECK-SD-NEXT: cmeq v0.4s, v0.4s, v1.4s
|
|
; CHECK-SD-NEXT: bic v1.16b, v2.16b, v0.16b
|
|
; CHECK-SD-NEXT: bic v0.4s, #128, lsl #24
|
|
; CHECK-SD-NEXT: orr v0.16b, v0.16b, v1.16b
|
|
; CHECK-SD-NEXT: ret
|
|
;
|
|
; CHECK-GI-LABEL: vqneg_test4:
|
|
; CHECK-GI: // %bb.0: // %entry
|
|
; CHECK-GI-NEXT: movi v1.4s, #128, lsl #24
|
|
; CHECK-GI-NEXT: mvni v2.4s, #128, lsl #24
|
|
; CHECK-GI-NEXT: cmeq v1.4s, v0.4s, v1.4s
|
|
; CHECK-GI-NEXT: neg v0.4s, v0.4s
|
|
; CHECK-GI-NEXT: bit v0.16b, v2.16b, v1.16b
|
|
; CHECK-GI-NEXT: ret
|
|
entry:
|
|
%0 = icmp eq <4 x i32> %A, <i32 -2147483648, i32 -2147483648, i32 -2147483648, i32 -2147483648>
|
|
%1 = sub nsw <4 x i32> zeroinitializer, %A
|
|
%2 = select <4 x i1> %0, <4 x i32> <i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647>, <4 x i32> %1
|
|
ret <4 x i32> %2
|
|
}
|
|
|
|
define <16 x i8> @vqneg_sat_v16i8(<16 x i8> %A) {
|
|
; CHECK-LABEL: vqneg_sat_v16i8:
|
|
; CHECK: // %bb.0: // %entry
|
|
; CHECK-NEXT: sqneg v0.16b, v0.16b
|
|
; CHECK-NEXT: ret
|
|
entry:
|
|
%0 = tail call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> zeroinitializer, <16 x i8> %A)
|
|
ret <16 x i8> %0
|
|
}
|
|
|
|
define <8 x i16> @vqneg_sat_v8i16(<8 x i16> %A) {
|
|
; CHECK-LABEL: vqneg_sat_v8i16:
|
|
; CHECK: // %bb.0: // %entry
|
|
; CHECK-NEXT: sqneg v0.8h, v0.8h
|
|
; CHECK-NEXT: ret
|
|
entry:
|
|
%0 = tail call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> zeroinitializer, <8 x i16> %A)
|
|
ret <8 x i16> %0
|
|
}
|
|
|
|
define <4 x i32> @vqneg_sat_v4i32(<4 x i32> %A) {
|
|
; CHECK-LABEL: vqneg_sat_v4i32:
|
|
; CHECK: // %bb.0: // %entry
|
|
; CHECK-NEXT: sqneg v0.4s, v0.4s
|
|
; CHECK-NEXT: ret
|
|
entry:
|
|
%0 = tail call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> zeroinitializer, <4 x i32> %A)
|
|
ret <4 x i32> %0
|
|
}
|
|
|
|
define <2 x i64> @vqneg_sat_v2i64(<2 x i64> %A) {
|
|
; CHECK-LABEL: vqneg_sat_v2i64:
|
|
; CHECK: // %bb.0: // %entry
|
|
; CHECK-NEXT: sqneg v0.2d, v0.2d
|
|
; CHECK-NEXT: ret
|
|
entry:
|
|
%0 = tail call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> zeroinitializer, <2 x i64> %A)
|
|
ret <2 x i64> %0
|
|
}
|
|
|
|
define <8 x i8> @vqneg_sat_v8i8(<8 x i8> %A) {
|
|
; CHECK-LABEL: vqneg_sat_v8i8:
|
|
; CHECK: // %bb.0: // %entry
|
|
; CHECK-NEXT: sqneg v0.8b, v0.8b
|
|
; CHECK-NEXT: ret
|
|
entry:
|
|
%0 = tail call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> zeroinitializer, <8 x i8> %A)
|
|
ret <8 x i8> %0
|
|
}
|
|
|
|
define <4 x i16> @vqneg_sat_v4i16(<4 x i16> %A) {
|
|
; CHECK-LABEL: vqneg_sat_v4i16:
|
|
; CHECK: // %bb.0: // %entry
|
|
; CHECK-NEXT: sqneg v0.4h, v0.4h
|
|
; CHECK-NEXT: ret
|
|
entry:
|
|
%0 = tail call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> zeroinitializer, <4 x i16> %A)
|
|
ret <4 x i16> %0
|
|
}
|
|
|
|
define <2 x i32> @vqneg_sat_v2i32(<2 x i32> %A) {
|
|
; CHECK-LABEL: vqneg_sat_v2i32:
|
|
; CHECK: // %bb.0: // %entry
|
|
; CHECK-NEXT: sqneg v0.2s, v0.2s
|
|
; CHECK-NEXT: ret
|
|
entry:
|
|
%0 = tail call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> zeroinitializer, <2 x i32> %A)
|
|
ret <2 x i32> %0
|
|
}
|
|
|
|
define <1 x i64> @vqneg_sat_v1i64(<1 x i64> %A) {
|
|
; CHECK-SD-LABEL: vqneg_sat_v1i64:
|
|
; CHECK-SD: // %bb.0: // %entry
|
|
; CHECK-SD-NEXT: sqneg d0, d0
|
|
; CHECK-SD-NEXT: ret
|
|
;
|
|
; CHECK-GI-LABEL: vqneg_sat_v1i64:
|
|
; CHECK-GI: // %bb.0: // %entry
|
|
; CHECK-GI-NEXT: fmov x9, d0
|
|
; CHECK-GI-NEXT: mov x8, #-9223372036854775808 // =0x8000000000000000
|
|
; CHECK-GI-NEXT: negs x9, x9
|
|
; CHECK-GI-NEXT: cset w10, vs
|
|
; CHECK-GI-NEXT: add x8, x8, x9, asr #63
|
|
; CHECK-GI-NEXT: tst w10, #0x1
|
|
; CHECK-GI-NEXT: csel x8, x8, x9, ne
|
|
; CHECK-GI-NEXT: fmov d0, x8
|
|
; CHECK-GI-NEXT: ret
|
|
entry:
|
|
%0 = tail call <1 x i64> @llvm.ssub.sat.v1i64(<1 x i64> zeroinitializer, <1 x i64> %A)
|
|
ret <1 x i64> %0
|
|
}
|