derek/TheRock
1
0 Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity
Files
main
TheRock/compiler/amd-llvm/llvm/test/Transforms/InstCombine/AMDGPU/mfma-scale.ll
T
derek.holloway 7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
Details
CodeQL / Analyze (python) (push) Canceled after 0s
Details
pre-commit / pre-commit (push) Canceled after 0s
Details
Multi-Arch CI / setup (push) Canceled after 0s
Details
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Details
Setup (Multi-Arch) / setup (push) Canceled after 0s
Details
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Details
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Details
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Details
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Details
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Details
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
Details
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Details
Multi-Arch CI / CI Summary (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Details
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Details
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Details
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Details
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Details
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Details
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
Details
Flatten subprojects into a single base monorepo
2026-09-02 16:34:47 -07:00

313 lines
23 KiB
LLVM
Raw Permalink Blame History

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5
; RUN: opt -S -mtriple=amdgcn-amd-amdhsa -passes=instcombine < %s | FileCheck %s
; --------------------------------------------------------------------
; Incorrect signature for format cases (IR vector too large) 16x16x128
; --------------------------------------------------------------------
define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp8__v8i32_fp6(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp8__v8i32_fp6(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <4 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[RESULT:%.*]] = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32> [[ARG0]], <6 x i32> [[TMP1]], <4 x float> [[ARG2]], i32 0, i32 2, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <4 x float> [[RESULT]]
;
%result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2,
i32 0, ; cbsz
i32 2, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <4 x float> %result
}
define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp6__v8i32_fp8(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp6__v8i32_fp8(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <4 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG0]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[RESULT:%.*]] = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v8i32(<6 x i32> [[TMP1]], <8 x i32> [[ARG1]], <4 x float> [[ARG2]], i32 2, i32 0, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <4 x float> [[RESULT]]
;
%result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2,
i32 2, ; cbsz
i32 0, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <4 x float> %result
}
define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp6__v8i32_fp6(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp6__v8i32_fp6(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <4 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG0]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[TMP2:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[RESULT:%.*]] = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v6i32(<6 x i32> [[TMP1]], <6 x i32> [[TMP2]], <4 x float> [[ARG2]], i32 2, i32 2, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <4 x float> [[RESULT]]
;
%result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2,
i32 2, ; cbsz
i32 2, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <4 x float> %result
}
define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp6__v8i32_fp6__0_scale(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) {
; CHECK-LABEL: define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp6__v8i32_fp6__0_scale(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <4 x float> [[ARG2:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG0]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[TMP2:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[RESULT:%.*]] = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v6i32(<6 x i32> [[TMP1]], <6 x i32> [[TMP2]], <4 x float> [[ARG2]], i32 2, i32 2, i32 0, i32 0, i32 0, i32 0)
; CHECK-NEXT: ret <4 x float> [[RESULT]]
;
%result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2,
i32 2, ; cbsz
i32 2, ; blgp
i32 0, i32 0, i32 0, i32 0)
ret <4 x float> %result
}
define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp8__v8i32_fp4(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp8__v8i32_fp4(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <4 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[RESULT:%.*]] = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v4i32(<8 x i32> [[ARG0]], <4 x i32> [[TMP1]], <4 x float> [[ARG2]], i32 0, i32 4, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <4 x float> [[RESULT]]
;
%result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2,
i32 0, ; cbsz
i32 4, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <4 x float> %result
}
define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp4__v8i32_fp8(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp4__v8i32_fp8(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <4 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG0]], <8 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[RESULT:%.*]] = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v8i32(<4 x i32> [[TMP1]], <8 x i32> [[ARG1]], <4 x float> [[ARG2]], i32 4, i32 0, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <4 x float> [[RESULT]]
;
%result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2,
i32 4, ; cbsz
i32 0, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <4 x float> %result
}
define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp8__v6i32_fp4(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp8__v6i32_fp4(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <6 x i32> [[ARG1:%.*]], <4 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <6 x i32> [[ARG1]], <6 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[RESULT:%.*]] = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v4i32(<8 x i32> [[ARG0]], <4 x i32> [[TMP1]], <4 x float> [[ARG2]], i32 0, i32 4, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <4 x float> [[RESULT]]
;
%result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2,
i32 0, ; cbsz
i32 4, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <4 x float> %result
}
define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v6i32_fp4__v8i32_fp8(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v6i32_fp4__v8i32_fp8(
; CHECK-SAME: <6 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <4 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <6 x i32> [[ARG0]], <6 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[RESULT:%.*]] = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v8i32(<4 x i32> [[TMP1]], <8 x i32> [[ARG1]], <4 x float> [[ARG2]], i32 4, i32 0, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <4 x float> [[RESULT]]
;
%result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2,
i32 4, ; cbsz
i32 0, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <4 x float> %result
}
define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp4__v8i32_fp4(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp4__v8i32_fp4(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <4 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG0]], <8 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[TMP2:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[RESULT:%.*]] = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v4i32(<4 x i32> [[TMP1]], <4 x i32> [[TMP2]], <4 x float> [[ARG2]], i32 4, i32 4, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <4 x float> [[RESULT]]
;
%result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2,
i32 4, ; cbsz
i32 4, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <4 x float> %result
}
define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp4__v8i32_fp4__0_scale(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) {
; CHECK-LABEL: define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp4__v8i32_fp4__0_scale(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <4 x float> [[ARG2:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG0]], <8 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[TMP2:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[RESULT:%.*]] = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v4i32(<4 x i32> [[TMP1]], <4 x i32> [[TMP2]], <4 x float> [[ARG2]], i32 4, i32 4, i32 0, i32 0, i32 0, i32 0)
; CHECK-NEXT: ret <4 x float> [[RESULT]]
;
%result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2,
i32 4, ; cbsz
i32 4, ; blgp
i32 0, i32 0, i32 0, i32 0)
ret <4 x float> %result
}
define <4 x float> @test_flags_shrink_src0(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <4 x float> @test_flags_shrink_src0(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <4 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[RESULT:%.*]] = call nnan nsz <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32> [[ARG0]], <6 x i32> [[TMP1]], <4 x float> [[ARG2]], i32 0, i32 2, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <4 x float> [[RESULT]]
;
%result = call nnan nsz <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2,
i32 0, ; cbsz
i32 2, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <4 x float> %result
}
; --------------------------------------------------------------------
; Incorrect signature for format cases (IR vector too large) 32x32x64
; --------------------------------------------------------------------
define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v8i32_fp6(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v8i32_fp6(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <16 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[RESULT:%.*]] = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v6i32(<8 x i32> [[ARG0]], <6 x i32> [[TMP1]], <16 x float> [[ARG2]], i32 0, i32 2, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <16 x float> [[RESULT]]
;
%result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2,
i32 0, ; cbsz
i32 2, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <16 x float> %result
}
define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp8(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp8(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <16 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG0]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[RESULT:%.*]] = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v8i32(<6 x i32> [[TMP1]], <8 x i32> [[ARG1]], <16 x float> [[ARG2]], i32 2, i32 0, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <16 x float> [[RESULT]]
;
%result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2,
i32 2, ; cbsz
i32 0, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <16 x float> %result
}
define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp6(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp6(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <16 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG0]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[TMP2:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[RESULT:%.*]] = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v6i32(<6 x i32> [[TMP1]], <6 x i32> [[TMP2]], <16 x float> [[ARG2]], i32 2, i32 2, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <16 x float> [[RESULT]]
;
%result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2,
i32 2, ; cbsz
i32 2, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <16 x float> %result
}
define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp6__0_scale(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) {
; CHECK-LABEL: define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp6__0_scale(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <16 x float> [[ARG2:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG0]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[TMP2:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <6 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5>
; CHECK-NEXT: [[RESULT:%.*]] = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v6i32(<6 x i32> [[TMP1]], <6 x i32> [[TMP2]], <16 x float> [[ARG2]], i32 2, i32 2, i32 0, i32 0, i32 0, i32 0)
; CHECK-NEXT: ret <16 x float> [[RESULT]]
;
%result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2,
i32 2, ; cbsz
i32 2, ; blgp
i32 0, i32 0, i32 0, i32 0)
ret <16 x float> %result
}
define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v8i32_fp4(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v8i32_fp4(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <16 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[RESULT:%.*]] = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v4i32(<8 x i32> [[ARG0]], <4 x i32> [[TMP1]], <16 x float> [[ARG2]], i32 0, i32 4, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <16 x float> [[RESULT]]
;
%result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2,
i32 0, ; cbsz
i32 4, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <16 x float> %result
}
define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp8(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp8(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <16 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG0]], <8 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[RESULT:%.*]] = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v8i32(<4 x i32> [[TMP1]], <8 x i32> [[ARG1]], <16 x float> [[ARG2]], i32 4, i32 0, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <16 x float> [[RESULT]]
;
%result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2,
i32 4, ; cbsz
i32 0, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <16 x float> %result
}
define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v6i32_fp4(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v6i32_fp4(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <6 x i32> [[ARG1:%.*]], <16 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <6 x i32> [[ARG1]], <6 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[RESULT:%.*]] = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v4i32(<8 x i32> [[ARG0]], <4 x i32> [[TMP1]], <16 x float> [[ARG2]], i32 0, i32 4, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <16 x float> [[RESULT]]
;
%result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2,
i32 0, ; cbsz
i32 4, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <16 x float> %result
}
define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v6i32_fp4__v8i32_fp8(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v6i32_fp4__v8i32_fp8(
; CHECK-SAME: <6 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <16 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <6 x i32> [[ARG0]], <6 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[RESULT:%.*]] = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v8i32(<4 x i32> [[TMP1]], <8 x i32> [[ARG1]], <16 x float> [[ARG2]], i32 4, i32 0, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <16 x float> [[RESULT]]
;
%result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2,
i32 4, ; cbsz
i32 0, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <16 x float> %result
}
define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp4(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) {
; CHECK-LABEL: define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp4(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <16 x float> [[ARG2:%.*]], i32 [[SCALE0:%.*]], i32 [[SCALE1:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG0]], <8 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[TMP2:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[RESULT:%.*]] = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v4i32(<4 x i32> [[TMP1]], <4 x i32> [[TMP2]], <16 x float> [[ARG2]], i32 4, i32 4, i32 0, i32 [[SCALE0]], i32 0, i32 [[SCALE1]])
; CHECK-NEXT: ret <16 x float> [[RESULT]]
;
%result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2,
i32 4, ; cbsz
i32 4, ; blgp
i32 0, i32 %scale0, i32 0, i32 %scale1)
ret <16 x float> %result
}
define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp4__0_scale(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) {
; CHECK-LABEL: define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp4__0_scale(
; CHECK-SAME: <8 x i32> [[ARG0:%.*]], <8 x i32> [[ARG1:%.*]], <16 x float> [[ARG2:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[ARG0]], <8 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[TMP2:%.*]] = shufflevector <8 x i32> [[ARG1]], <8 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; CHECK-NEXT: [[RESULT:%.*]] = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v4i32(<4 x i32> [[TMP1]], <4 x i32> [[TMP2]], <16 x float> [[ARG2]], i32 4, i32 4, i32 0, i32 0, i32 0, i32 0)
; CHECK-NEXT: ret <16 x float> [[RESULT]]
;
%result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2,
i32 4, ; cbsz
i32 4, ; blgp
i32 0, i32 0, i32 0, i32 0)
ret <16 x float> %result
}
Reference in new issue View Git Blame Copy Permalink