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
160 lines
5.9 KiB
LLVM
160 lines
5.9 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc < %s -mtriple=x86_64-- -mattr=sse2 | FileCheck %s --check-prefix=SSE
|
|
; RUN: llc < %s -mtriple=x86_64-- -mattr=avx | FileCheck %s --check-prefix=AVX
|
|
|
|
define <2 x double> @splat_fdiv_v2f64(<2 x double> %x, double %y) {
|
|
; SSE-LABEL: splat_fdiv_v2f64:
|
|
; SSE: # %bb.0:
|
|
; SSE-NEXT: movsd {{.*#+}} xmm2 = [1.0E+0,0.0E+0]
|
|
; SSE-NEXT: divsd %xmm1, %xmm2
|
|
; SSE-NEXT: unpcklpd {{.*#+}} xmm2 = xmm2[0,0]
|
|
; SSE-NEXT: mulpd %xmm2, %xmm0
|
|
; SSE-NEXT: retq
|
|
;
|
|
; AVX-LABEL: splat_fdiv_v2f64:
|
|
; AVX: # %bb.0:
|
|
; AVX-NEXT: vmovsd {{.*#+}} xmm2 = [1.0E+0,0.0E+0]
|
|
; AVX-NEXT: vdivsd %xmm1, %xmm2, %xmm1
|
|
; AVX-NEXT: vmovddup {{.*#+}} xmm1 = xmm1[0,0]
|
|
; AVX-NEXT: vmulpd %xmm1, %xmm0, %xmm0
|
|
; AVX-NEXT: retq
|
|
%vy = insertelement <2 x double> undef, double %y, i32 0
|
|
%splaty = shufflevector <2 x double> %vy, <2 x double> undef, <2 x i32> zeroinitializer
|
|
%r = fdiv fast <2 x double> %x, %splaty
|
|
ret <2 x double> %r
|
|
}
|
|
|
|
define <4 x double> @splat_fdiv_v4f64(<4 x double> %x, double %y) {
|
|
; SSE-LABEL: splat_fdiv_v4f64:
|
|
; SSE: # %bb.0:
|
|
; SSE-NEXT: movsd {{.*#+}} xmm3 = [1.0E+0,0.0E+0]
|
|
; SSE-NEXT: divsd %xmm2, %xmm3
|
|
; SSE-NEXT: unpcklpd {{.*#+}} xmm3 = xmm3[0,0]
|
|
; SSE-NEXT: mulpd %xmm3, %xmm0
|
|
; SSE-NEXT: mulpd %xmm3, %xmm1
|
|
; SSE-NEXT: retq
|
|
;
|
|
; AVX-LABEL: splat_fdiv_v4f64:
|
|
; AVX: # %bb.0:
|
|
; AVX-NEXT: vmovsd {{.*#+}} xmm2 = [1.0E+0,0.0E+0]
|
|
; AVX-NEXT: vdivsd %xmm1, %xmm2, %xmm1
|
|
; AVX-NEXT: vmovddup {{.*#+}} xmm1 = xmm1[0,0]
|
|
; AVX-NEXT: vinsertf128 $1, %xmm1, %ymm1, %ymm1
|
|
; AVX-NEXT: vmulpd %ymm1, %ymm0, %ymm0
|
|
; AVX-NEXT: retq
|
|
%vy = insertelement <4 x double> undef, double %y, i32 0
|
|
%splaty = shufflevector <4 x double> %vy, <4 x double> undef, <4 x i32> zeroinitializer
|
|
%r = fdiv arcp <4 x double> %x, %splaty
|
|
ret <4 x double> %r
|
|
}
|
|
|
|
define <4 x float> @splat_fdiv_v4f32(<4 x float> %x, float %y) {
|
|
; SSE-LABEL: splat_fdiv_v4f32:
|
|
; SSE: # %bb.0:
|
|
; SSE-NEXT: movss {{.*#+}} xmm2 = [1.0E+0,0.0E+0,0.0E+0,0.0E+0]
|
|
; SSE-NEXT: divss %xmm1, %xmm2
|
|
; SSE-NEXT: shufps {{.*#+}} xmm2 = xmm2[0,0,0,0]
|
|
; SSE-NEXT: mulps %xmm2, %xmm0
|
|
; SSE-NEXT: retq
|
|
;
|
|
; AVX-LABEL: splat_fdiv_v4f32:
|
|
; AVX: # %bb.0:
|
|
; AVX-NEXT: vmovss {{.*#+}} xmm2 = [1.0E+0,0.0E+0,0.0E+0,0.0E+0]
|
|
; AVX-NEXT: vdivss %xmm1, %xmm2, %xmm1
|
|
; AVX-NEXT: vshufps {{.*#+}} xmm1 = xmm1[0,0,0,0]
|
|
; AVX-NEXT: vmulps %xmm1, %xmm0, %xmm0
|
|
; AVX-NEXT: retq
|
|
%vy = insertelement <4 x float> undef, float %y, i32 0
|
|
%splaty = shufflevector <4 x float> %vy, <4 x float> undef, <4 x i32> zeroinitializer
|
|
%r = fdiv arcp reassoc <4 x float> %x, %splaty
|
|
ret <4 x float> %r
|
|
}
|
|
|
|
define <8 x float> @splat_fdiv_v8f32(<8 x float> %x, float %y) {
|
|
; SSE-LABEL: splat_fdiv_v8f32:
|
|
; SSE: # %bb.0:
|
|
; SSE-NEXT: movss {{.*#+}} xmm3 = [1.0E+0,0.0E+0,0.0E+0,0.0E+0]
|
|
; SSE-NEXT: divss %xmm2, %xmm3
|
|
; SSE-NEXT: shufps {{.*#+}} xmm3 = xmm3[0,0,0,0]
|
|
; SSE-NEXT: mulps %xmm3, %xmm0
|
|
; SSE-NEXT: mulps %xmm3, %xmm1
|
|
; SSE-NEXT: retq
|
|
;
|
|
; AVX-LABEL: splat_fdiv_v8f32:
|
|
; AVX: # %bb.0:
|
|
; AVX-NEXT: vmovss {{.*#+}} xmm2 = [1.0E+0,0.0E+0,0.0E+0,0.0E+0]
|
|
; AVX-NEXT: vdivss %xmm1, %xmm2, %xmm1
|
|
; AVX-NEXT: vshufps {{.*#+}} xmm1 = xmm1[0,0,0,0]
|
|
; AVX-NEXT: vinsertf128 $1, %xmm1, %ymm1, %ymm1
|
|
; AVX-NEXT: vmulps %ymm1, %ymm0, %ymm0
|
|
; AVX-NEXT: retq
|
|
%vy = insertelement <8 x float> undef, float %y, i32 0
|
|
%splaty = shufflevector <8 x float> %vy, <8 x float> undef, <8 x i32> zeroinitializer
|
|
%r = fdiv fast <8 x float> %x, %splaty
|
|
ret <8 x float> %r
|
|
}
|
|
|
|
define <4 x float> @splat_fdiv_v4f32_estimate(<4 x float> %x, float %y) #0 {
|
|
; SSE-LABEL: splat_fdiv_v4f32_estimate:
|
|
; SSE: # %bb.0:
|
|
; SSE-NEXT: rcpss %xmm1, %xmm2
|
|
; SSE-NEXT: mulss %xmm2, %xmm1
|
|
; SSE-NEXT: movss {{.*#+}} xmm3 = [1.0E+0,0.0E+0,0.0E+0,0.0E+0]
|
|
; SSE-NEXT: subss %xmm1, %xmm3
|
|
; SSE-NEXT: mulss %xmm2, %xmm3
|
|
; SSE-NEXT: addss %xmm2, %xmm3
|
|
; SSE-NEXT: shufps {{.*#+}} xmm3 = xmm3[0,0,0,0]
|
|
; SSE-NEXT: mulps %xmm3, %xmm0
|
|
; SSE-NEXT: retq
|
|
;
|
|
; AVX-LABEL: splat_fdiv_v4f32_estimate:
|
|
; AVX: # %bb.0:
|
|
; AVX-NEXT: vrcpss %xmm1, %xmm1, %xmm2
|
|
; AVX-NEXT: vmulss %xmm2, %xmm1, %xmm1
|
|
; AVX-NEXT: vmovss {{.*#+}} xmm3 = [1.0E+0,0.0E+0,0.0E+0,0.0E+0]
|
|
; AVX-NEXT: vsubss %xmm1, %xmm3, %xmm1
|
|
; AVX-NEXT: vmulss %xmm1, %xmm2, %xmm1
|
|
; AVX-NEXT: vaddss %xmm1, %xmm2, %xmm1
|
|
; AVX-NEXT: vshufps {{.*#+}} xmm1 = xmm1[0,0,0,0]
|
|
; AVX-NEXT: vmulps %xmm1, %xmm0, %xmm0
|
|
; AVX-NEXT: retq
|
|
%vy = insertelement <4 x float> undef, float %y, i32 0
|
|
%splaty = shufflevector <4 x float> %vy, <4 x float> undef, <4 x i32> zeroinitializer
|
|
%r = fdiv arcp reassoc ninf <4 x float> %x, %splaty
|
|
ret <4 x float> %r
|
|
}
|
|
|
|
define <8 x float> @splat_fdiv_v8f32_estimate(<8 x float> %x, float %y) #0 {
|
|
; SSE-LABEL: splat_fdiv_v8f32_estimate:
|
|
; SSE: # %bb.0:
|
|
; SSE-NEXT: rcpss %xmm2, %xmm3
|
|
; SSE-NEXT: mulss %xmm3, %xmm2
|
|
; SSE-NEXT: movss {{.*#+}} xmm4 = [1.0E+0,0.0E+0,0.0E+0,0.0E+0]
|
|
; SSE-NEXT: subss %xmm2, %xmm4
|
|
; SSE-NEXT: mulss %xmm3, %xmm4
|
|
; SSE-NEXT: addss %xmm3, %xmm4
|
|
; SSE-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,0,0,0]
|
|
; SSE-NEXT: mulps %xmm4, %xmm0
|
|
; SSE-NEXT: mulps %xmm4, %xmm1
|
|
; SSE-NEXT: retq
|
|
;
|
|
; AVX-LABEL: splat_fdiv_v8f32_estimate:
|
|
; AVX: # %bb.0:
|
|
; AVX-NEXT: vrcpss %xmm1, %xmm1, %xmm2
|
|
; AVX-NEXT: vmulss %xmm2, %xmm1, %xmm1
|
|
; AVX-NEXT: vmovss {{.*#+}} xmm3 = [1.0E+0,0.0E+0,0.0E+0,0.0E+0]
|
|
; AVX-NEXT: vsubss %xmm1, %xmm3, %xmm1
|
|
; AVX-NEXT: vmulss %xmm1, %xmm2, %xmm1
|
|
; AVX-NEXT: vaddss %xmm1, %xmm2, %xmm1
|
|
; AVX-NEXT: vshufps {{.*#+}} xmm1 = xmm1[0,0,0,0]
|
|
; AVX-NEXT: vinsertf128 $1, %xmm1, %ymm1, %ymm1
|
|
; AVX-NEXT: vmulps %ymm1, %ymm0, %ymm0
|
|
; AVX-NEXT: retq
|
|
%vy = insertelement <8 x float> undef, float %y, i32 0
|
|
%splaty = shufflevector <8 x float> %vy, <8 x float> undef, <8 x i32> zeroinitializer
|
|
%r = fdiv fast <8 x float> %x, %splaty
|
|
ret <8 x float> %r
|
|
}
|
|
|
|
attributes #0 = { "reciprocal-estimates"="divf,vec-divf" }
|