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
301 lines
9.0 KiB
LLVM
301 lines
9.0 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc -mtriple=riscv32 -mattr=+v -verify-machineinstrs < %s | FileCheck %s
|
|
; RUN: llc -mtriple=riscv64 -mattr=+v -verify-machineinstrs < %s | FileCheck %s
|
|
; RUN: llc -mtriple=riscv32 -mattr=+v,+experimental-zvabd -verify-machineinstrs < %s \
|
|
; RUN: | FileCheck %s --check-prefixes=ZVABD
|
|
; RUN: llc -mtriple=riscv64 -mattr=+v,+experimental-zvabd -verify-machineinstrs < %s \
|
|
; RUN: | FileCheck %s --check-prefixes=ZVABD
|
|
|
|
define void @abs_v16i8(ptr %x) {
|
|
; CHECK-LABEL: abs_v16i8:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
|
|
; CHECK-NEXT: vle8.v v8, (a0)
|
|
; CHECK-NEXT: vrsub.vi v9, v8, 0
|
|
; CHECK-NEXT: vmax.vv v8, v8, v9
|
|
; CHECK-NEXT: vse8.v v8, (a0)
|
|
; CHECK-NEXT: ret
|
|
;
|
|
; ZVABD-LABEL: abs_v16i8:
|
|
; ZVABD: # %bb.0:
|
|
; ZVABD-NEXT: vsetivli zero, 16, e8, m1, ta, ma
|
|
; ZVABD-NEXT: vle8.v v8, (a0)
|
|
; ZVABD-NEXT: vabs.v v8, v8
|
|
; ZVABD-NEXT: vse8.v v8, (a0)
|
|
; ZVABD-NEXT: ret
|
|
%a = load <16 x i8>, ptr %x
|
|
%b = call <16 x i8> @llvm.abs.v16i8(<16 x i8> %a, i1 false)
|
|
store <16 x i8> %b, ptr %x
|
|
ret void
|
|
}
|
|
|
|
define void @abs_v8i16(ptr %x) {
|
|
; CHECK-LABEL: abs_v8i16:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma
|
|
; CHECK-NEXT: vle16.v v8, (a0)
|
|
; CHECK-NEXT: vrsub.vi v9, v8, 0
|
|
; CHECK-NEXT: vmax.vv v8, v8, v9
|
|
; CHECK-NEXT: vse16.v v8, (a0)
|
|
; CHECK-NEXT: ret
|
|
;
|
|
; ZVABD-LABEL: abs_v8i16:
|
|
; ZVABD: # %bb.0:
|
|
; ZVABD-NEXT: vsetivli zero, 8, e16, m1, ta, ma
|
|
; ZVABD-NEXT: vle16.v v8, (a0)
|
|
; ZVABD-NEXT: vabs.v v8, v8
|
|
; ZVABD-NEXT: vse16.v v8, (a0)
|
|
; ZVABD-NEXT: ret
|
|
%a = load <8 x i16>, ptr %x
|
|
%b = call <8 x i16> @llvm.abs.v8i16(<8 x i16> %a, i1 false)
|
|
store <8 x i16> %b, ptr %x
|
|
ret void
|
|
}
|
|
|
|
define void @abs_v6i16(ptr %x) {
|
|
; CHECK-LABEL: abs_v6i16:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vsetivli zero, 6, e16, m1, ta, ma
|
|
; CHECK-NEXT: vle16.v v8, (a0)
|
|
; CHECK-NEXT: vrsub.vi v9, v8, 0
|
|
; CHECK-NEXT: vmax.vv v8, v8, v9
|
|
; CHECK-NEXT: vse16.v v8, (a0)
|
|
; CHECK-NEXT: ret
|
|
;
|
|
; ZVABD-LABEL: abs_v6i16:
|
|
; ZVABD: # %bb.0:
|
|
; ZVABD-NEXT: vsetivli zero, 6, e16, m1, ta, ma
|
|
; ZVABD-NEXT: vle16.v v8, (a0)
|
|
; ZVABD-NEXT: vabs.v v8, v8
|
|
; ZVABD-NEXT: vse16.v v8, (a0)
|
|
; ZVABD-NEXT: ret
|
|
%a = load <6 x i16>, ptr %x
|
|
%b = call <6 x i16> @llvm.abs.v6i16(<6 x i16> %a, i1 false)
|
|
store <6 x i16> %b, ptr %x
|
|
ret void
|
|
}
|
|
|
|
define void @abs_v4i32(ptr %x) {
|
|
; CHECK-LABEL: abs_v4i32:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
|
|
; CHECK-NEXT: vle32.v v8, (a0)
|
|
; CHECK-NEXT: vrsub.vi v9, v8, 0
|
|
; CHECK-NEXT: vmax.vv v8, v8, v9
|
|
; CHECK-NEXT: vse32.v v8, (a0)
|
|
; CHECK-NEXT: ret
|
|
;
|
|
; ZVABD-LABEL: abs_v4i32:
|
|
; ZVABD: # %bb.0:
|
|
; ZVABD-NEXT: vsetivli zero, 4, e32, m1, ta, ma
|
|
; ZVABD-NEXT: vle32.v v8, (a0)
|
|
; ZVABD-NEXT: vabs.v v8, v8
|
|
; ZVABD-NEXT: vse32.v v8, (a0)
|
|
; ZVABD-NEXT: ret
|
|
%a = load <4 x i32>, ptr %x
|
|
%b = call <4 x i32> @llvm.abs.v4i32(<4 x i32> %a, i1 false)
|
|
store <4 x i32> %b, ptr %x
|
|
ret void
|
|
}
|
|
|
|
define void @abs_v2i64(ptr %x) {
|
|
; CHECK-LABEL: abs_v2i64:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vsetivli zero, 2, e64, m1, ta, ma
|
|
; CHECK-NEXT: vle64.v v8, (a0)
|
|
; CHECK-NEXT: vrsub.vi v9, v8, 0
|
|
; CHECK-NEXT: vmax.vv v8, v8, v9
|
|
; CHECK-NEXT: vse64.v v8, (a0)
|
|
; CHECK-NEXT: ret
|
|
;
|
|
; ZVABD-LABEL: abs_v2i64:
|
|
; ZVABD: # %bb.0:
|
|
; ZVABD-NEXT: vsetivli zero, 2, e64, m1, ta, ma
|
|
; ZVABD-NEXT: vle64.v v8, (a0)
|
|
; ZVABD-NEXT: vabs.v v8, v8
|
|
; ZVABD-NEXT: vse64.v v8, (a0)
|
|
; ZVABD-NEXT: ret
|
|
%a = load <2 x i64>, ptr %x
|
|
%b = call <2 x i64> @llvm.abs.v2i64(<2 x i64> %a, i1 false)
|
|
store <2 x i64> %b, ptr %x
|
|
ret void
|
|
}
|
|
|
|
define void @abs_v32i8(ptr %x) {
|
|
; CHECK-LABEL: abs_v32i8:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: li a1, 32
|
|
; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
|
|
; CHECK-NEXT: vle8.v v8, (a0)
|
|
; CHECK-NEXT: vrsub.vi v10, v8, 0
|
|
; CHECK-NEXT: vmax.vv v8, v8, v10
|
|
; CHECK-NEXT: vse8.v v8, (a0)
|
|
; CHECK-NEXT: ret
|
|
;
|
|
; ZVABD-LABEL: abs_v32i8:
|
|
; ZVABD: # %bb.0:
|
|
; ZVABD-NEXT: li a1, 32
|
|
; ZVABD-NEXT: vsetvli zero, a1, e8, m2, ta, ma
|
|
; ZVABD-NEXT: vle8.v v8, (a0)
|
|
; ZVABD-NEXT: vabs.v v8, v8
|
|
; ZVABD-NEXT: vse8.v v8, (a0)
|
|
; ZVABD-NEXT: ret
|
|
%a = load <32 x i8>, ptr %x
|
|
%b = call <32 x i8> @llvm.abs.v32i8(<32 x i8> %a, i1 false)
|
|
store <32 x i8> %b, ptr %x
|
|
ret void
|
|
}
|
|
|
|
define void @abs_v16i16(ptr %x) {
|
|
; CHECK-LABEL: abs_v16i16:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vsetivli zero, 16, e16, m2, ta, ma
|
|
; CHECK-NEXT: vle16.v v8, (a0)
|
|
; CHECK-NEXT: vrsub.vi v10, v8, 0
|
|
; CHECK-NEXT: vmax.vv v8, v8, v10
|
|
; CHECK-NEXT: vse16.v v8, (a0)
|
|
; CHECK-NEXT: ret
|
|
;
|
|
; ZVABD-LABEL: abs_v16i16:
|
|
; ZVABD: # %bb.0:
|
|
; ZVABD-NEXT: vsetivli zero, 16, e16, m2, ta, ma
|
|
; ZVABD-NEXT: vle16.v v8, (a0)
|
|
; ZVABD-NEXT: vabs.v v8, v8
|
|
; ZVABD-NEXT: vse16.v v8, (a0)
|
|
; ZVABD-NEXT: ret
|
|
%a = load <16 x i16>, ptr %x
|
|
%b = call <16 x i16> @llvm.abs.v16i16(<16 x i16> %a, i1 false)
|
|
store <16 x i16> %b, ptr %x
|
|
ret void
|
|
}
|
|
|
|
define void @abs_v8i32(ptr %x) {
|
|
; CHECK-LABEL: abs_v8i32:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
|
|
; CHECK-NEXT: vle32.v v8, (a0)
|
|
; CHECK-NEXT: vrsub.vi v10, v8, 0
|
|
; CHECK-NEXT: vmax.vv v8, v8, v10
|
|
; CHECK-NEXT: vse32.v v8, (a0)
|
|
; CHECK-NEXT: ret
|
|
;
|
|
; ZVABD-LABEL: abs_v8i32:
|
|
; ZVABD: # %bb.0:
|
|
; ZVABD-NEXT: vsetivli zero, 8, e32, m2, ta, ma
|
|
; ZVABD-NEXT: vle32.v v8, (a0)
|
|
; ZVABD-NEXT: vabs.v v8, v8
|
|
; ZVABD-NEXT: vse32.v v8, (a0)
|
|
; ZVABD-NEXT: ret
|
|
%a = load <8 x i32>, ptr %x
|
|
%b = call <8 x i32> @llvm.abs.v8i32(<8 x i32> %a, i1 false)
|
|
store <8 x i32> %b, ptr %x
|
|
ret void
|
|
}
|
|
|
|
define void @abs_v4i64(ptr %x) {
|
|
; CHECK-LABEL: abs_v4i64:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, ma
|
|
; CHECK-NEXT: vle64.v v8, (a0)
|
|
; CHECK-NEXT: vrsub.vi v10, v8, 0
|
|
; CHECK-NEXT: vmax.vv v8, v8, v10
|
|
; CHECK-NEXT: vse64.v v8, (a0)
|
|
; CHECK-NEXT: ret
|
|
;
|
|
; ZVABD-LABEL: abs_v4i64:
|
|
; ZVABD: # %bb.0:
|
|
; ZVABD-NEXT: vsetivli zero, 4, e64, m2, ta, ma
|
|
; ZVABD-NEXT: vle64.v v8, (a0)
|
|
; ZVABD-NEXT: vabs.v v8, v8
|
|
; ZVABD-NEXT: vse64.v v8, (a0)
|
|
; ZVABD-NEXT: ret
|
|
%a = load <4 x i64>, ptr %x
|
|
%b = call <4 x i64> @llvm.abs.v4i64(<4 x i64> %a, i1 false)
|
|
store <4 x i64> %b, ptr %x
|
|
ret void
|
|
}
|
|
|
|
define void @abs_v4i64_of_sext_v4i8(ptr %x) {
|
|
; CHECK-LABEL: abs_v4i64_of_sext_v4i8:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
|
|
; CHECK-NEXT: vle8.v v8, (a0)
|
|
; CHECK-NEXT: vrsub.vi v9, v8, 0
|
|
; CHECK-NEXT: vmax.vv v10, v8, v9
|
|
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
|
|
; CHECK-NEXT: vzext.vf8 v8, v10
|
|
; CHECK-NEXT: vse64.v v8, (a0)
|
|
; CHECK-NEXT: ret
|
|
;
|
|
; ZVABD-LABEL: abs_v4i64_of_sext_v4i8:
|
|
; ZVABD: # %bb.0:
|
|
; ZVABD-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
|
|
; ZVABD-NEXT: vle8.v v8, (a0)
|
|
; ZVABD-NEXT: vabs.v v10, v8
|
|
; ZVABD-NEXT: vsetvli zero, zero, e64, m2, ta, ma
|
|
; ZVABD-NEXT: vzext.vf8 v8, v10
|
|
; ZVABD-NEXT: vse64.v v8, (a0)
|
|
; ZVABD-NEXT: ret
|
|
%a = load <4 x i8>, ptr %x
|
|
%a.ext = sext <4 x i8> %a to <4 x i64>
|
|
%b = call <4 x i64> @llvm.abs.v4i64(<4 x i64> %a.ext, i1 false)
|
|
store <4 x i64> %b, ptr %x
|
|
ret void
|
|
}
|
|
|
|
define void @abs_v4i64_of_sext_v4i16(ptr %x) {
|
|
; CHECK-LABEL: abs_v4i64_of_sext_v4i16:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
|
|
; CHECK-NEXT: vle16.v v8, (a0)
|
|
; CHECK-NEXT: vrsub.vi v9, v8, 0
|
|
; CHECK-NEXT: vmax.vv v10, v8, v9
|
|
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
|
|
; CHECK-NEXT: vzext.vf4 v8, v10
|
|
; CHECK-NEXT: vse64.v v8, (a0)
|
|
; CHECK-NEXT: ret
|
|
;
|
|
; ZVABD-LABEL: abs_v4i64_of_sext_v4i16:
|
|
; ZVABD: # %bb.0:
|
|
; ZVABD-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
|
|
; ZVABD-NEXT: vle16.v v8, (a0)
|
|
; ZVABD-NEXT: vabs.v v10, v8
|
|
; ZVABD-NEXT: vsetvli zero, zero, e64, m2, ta, ma
|
|
; ZVABD-NEXT: vzext.vf4 v8, v10
|
|
; ZVABD-NEXT: vse64.v v8, (a0)
|
|
; ZVABD-NEXT: ret
|
|
%a = load <4 x i16>, ptr %x
|
|
%a.ext = sext <4 x i16> %a to <4 x i64>
|
|
%b = call <4 x i64> @llvm.abs.v4i64(<4 x i64> %a.ext, i1 false)
|
|
store <4 x i64> %b, ptr %x
|
|
ret void
|
|
}
|
|
|
|
define void @abs_v4i64_of_sext_v4i32(ptr %x) {
|
|
; CHECK-LABEL: abs_v4i64_of_sext_v4i32:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
|
|
; CHECK-NEXT: vle32.v v8, (a0)
|
|
; CHECK-NEXT: vrsub.vi v9, v8, 0
|
|
; CHECK-NEXT: vmax.vv v10, v8, v9
|
|
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
|
|
; CHECK-NEXT: vzext.vf2 v8, v10
|
|
; CHECK-NEXT: vse64.v v8, (a0)
|
|
; CHECK-NEXT: ret
|
|
;
|
|
; ZVABD-LABEL: abs_v4i64_of_sext_v4i32:
|
|
; ZVABD: # %bb.0:
|
|
; ZVABD-NEXT: vsetivli zero, 4, e32, m1, ta, ma
|
|
; ZVABD-NEXT: vle32.v v8, (a0)
|
|
; ZVABD-NEXT: vabs.v v10, v8
|
|
; ZVABD-NEXT: vsetvli zero, zero, e64, m2, ta, ma
|
|
; ZVABD-NEXT: vzext.vf2 v8, v10
|
|
; ZVABD-NEXT: vse64.v v8, (a0)
|
|
; ZVABD-NEXT: ret
|
|
%a = load <4 x i32>, ptr %x
|
|
%a.ext = sext <4 x i32> %a to <4 x i64>
|
|
%b = call <4 x i64> @llvm.abs.v4i64(<4 x i64> %a.ext, i1 false)
|
|
store <4 x i64> %b, ptr %x
|
|
ret void
|
|
}
|