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
54 lines
2.3 KiB
LLVM
54 lines
2.3 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc < %s -mtriple=ve -mattr=+vpu | FileCheck %s
|
|
|
|
declare <512 x float> @llvm.vp.fmul.v512f32(<512 x float>, <512 x float>, <512 x i1>, i32)
|
|
|
|
define fastcc <512 x float> @test_vp_fmul_v512f32_vv(<512 x float> %i0, <512 x float> %i1, <512 x i1> %m, i32 %n) {
|
|
; CHECK-LABEL: test_vp_fmul_v512f32_vv:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: adds.w.sx %s0, 1, %s0
|
|
; CHECK-NEXT: and %s0, %s0, (32)0
|
|
; CHECK-NEXT: srl %s0, %s0, 1
|
|
; CHECK-NEXT: lvl %s0
|
|
; CHECK-NEXT: pvfmul %v0, %v0, %v1, %vm2
|
|
; CHECK-NEXT: b.l.t (, %s10)
|
|
%r0 = call <512 x float> @llvm.vp.fmul.v512f32(<512 x float> %i0, <512 x float> %i1, <512 x i1> %m, i32 %n)
|
|
ret <512 x float> %r0
|
|
}
|
|
|
|
define fastcc <512 x float> @test_vp_fmul_v512f32_rv(float %s0, <512 x float> %i1, <512 x i1> %m, i32 %n) {
|
|
; CHECK-LABEL: test_vp_fmul_v512f32_rv:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: and %s2, %s0, (32)1
|
|
; CHECK-NEXT: srl %s0, %s0, 32
|
|
; CHECK-NEXT: or %s0, %s0, %s2
|
|
; CHECK-NEXT: adds.w.sx %s1, 1, %s1
|
|
; CHECK-NEXT: and %s1, %s1, (32)0
|
|
; CHECK-NEXT: srl %s1, %s1, 1
|
|
; CHECK-NEXT: lvl %s1
|
|
; CHECK-NEXT: pvfmul %v0, %s0, %v0, %vm2
|
|
; CHECK-NEXT: b.l.t (, %s10)
|
|
%xins = insertelement <512 x float> undef, float %s0, i32 0
|
|
%i0 = shufflevector <512 x float> %xins, <512 x float> undef, <512 x i32> zeroinitializer
|
|
%r0 = call <512 x float> @llvm.vp.fmul.v512f32(<512 x float> %i0, <512 x float> %i1, <512 x i1> %m, i32 %n)
|
|
ret <512 x float> %r0
|
|
}
|
|
|
|
define fastcc <512 x float> @test_vp_fmul_v512f32_vr(<512 x float> %i0, float %s1, <512 x i1> %m, i32 %n) {
|
|
; CHECK-LABEL: test_vp_fmul_v512f32_vr:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: and %s2, %s0, (32)1
|
|
; CHECK-NEXT: srl %s0, %s0, 32
|
|
; CHECK-NEXT: or %s0, %s0, %s2
|
|
; CHECK-NEXT: adds.w.sx %s1, 1, %s1
|
|
; CHECK-NEXT: and %s1, %s1, (32)0
|
|
; CHECK-NEXT: srl %s1, %s1, 1
|
|
; CHECK-NEXT: lvl %s1
|
|
; CHECK-NEXT: pvfmul %v0, %s0, %v0, %vm2
|
|
; CHECK-NEXT: b.l.t (, %s10)
|
|
%yins = insertelement <512 x float> undef, float %s1, i32 0
|
|
%i1 = shufflevector <512 x float> %yins, <512 x float> undef, <512 x i32> zeroinitializer
|
|
%r0 = call <512 x float> @llvm.vp.fmul.v512f32(<512 x float> %i0, <512 x float> %i1, <512 x i1> %m, i32 %n)
|
|
ret <512 x float> %r0
|
|
}
|