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
64 lines
4.6 KiB
MLIR
64 lines
4.6 KiB
MLIR
// RUN: mlir-opt -lower-vector-multi-reduction="lowering-strategy=inner-reduction" -split-input-file %s | FileCheck %s --check-prefixes=ALL,INNER-REDUCTION
|
|
// RUN: mlir-opt -lower-vector-multi-reduction="lowering-strategy=inner-parallel" -split-input-file %s | FileCheck %s --check-prefixes=ALL,INNER-PARALLEL
|
|
// RUN: mlir-opt -lower-vector-multi-reduction -split-input-file %s | FileCheck %s --check-prefixes=ALL,INNER-PARALLEL
|
|
|
|
func.func @vector_multi_reduction(%arg0: vector<2x4xf32>, %acc: vector<2xf32>) -> vector<2xf32> {
|
|
%0 = vector.multi_reduction <mul>, %arg0, %acc [1] : vector<2x4xf32> to vector<2xf32>
|
|
return %0 : vector<2xf32>
|
|
}
|
|
// ALL-LABEL: func @vector_multi_reduction
|
|
// ALL-SAME: %[[INPUT:.+]]: vector<2x4xf32>, %[[ACC:.*]]: vector<2xf32>)
|
|
// INNER-REDUCTION-DAG: %[[RESULT_VEC_0:.+]] = arith.constant dense<{{.*}}> : vector<2xf32>
|
|
// INNER-REDUCTION: %[[V0:.+]] = vector.extract %[[INPUT]][0]
|
|
// INNER-REDUCTION: %[[ACC0:.+]] = vector.extract %[[ACC]][0]
|
|
// INNER-REDUCTION: %[[RV0:.+]] = vector.reduction <mul>, %[[V0]], %[[ACC0]] : vector<4xf32> into f32
|
|
// INNER-REDUCTION: %[[RESULT_VEC_1:.+]] = vector.insert %[[RV0:.+]], %[[RESULT_VEC_0]] [0] : f32 into vector<2xf32>
|
|
// INNER-REDUCTION: %[[V1:.+]] = vector.extract %[[INPUT]][1]
|
|
// INNER-REDUCTION: %[[ACC1:.+]] = vector.extract %[[ACC]][1]
|
|
// INNER-REDUCTION: %[[RV1:.+]] = vector.reduction <mul>, %[[V1]], %[[ACC1]] : vector<4xf32> into f32
|
|
// INNER-REDUCTION: %[[RESULT_VEC:.+]] = vector.insert %[[RV1:.+]], %[[RESULT_VEC_1]] [1] : f32 into vector<2xf32>
|
|
// INNER-REDUCTION: return %[[RESULT_VEC]]
|
|
|
|
// INNER-PARALLEL: %[[TRANSPOSED:.+]] = vector.transpose %[[INPUT]], [1, 0] : vector<2x4xf32> to vector<4x2xf32>
|
|
// INNER-PARALLEL: %[[V0:.+]] = vector.extract %[[TRANSPOSED]][0] : vector<2xf32> from vector<4x2xf32>
|
|
// INNER-PARALLEL: %[[RV0:.+]] = arith.mulf %[[V0]], %[[ACC]] : vector<2xf32>
|
|
// INNER-PARALLEL: %[[V1:.+]] = vector.extract %[[TRANSPOSED]][1] : vector<2xf32> from vector<4x2xf32>
|
|
// INNER-PARALLEL: %[[RV01:.+]] = arith.mulf %[[V1]], %[[RV0]] : vector<2xf32>
|
|
// INNER-PARALLEL: %[[V2:.+]] = vector.extract %[[TRANSPOSED]][2] : vector<2xf32> from vector<4x2xf32>
|
|
// INNER-PARALLEL: %[[RV012:.+]] = arith.mulf %[[V2]], %[[RV01]] : vector<2xf32>
|
|
// INNER-PARALLEL: %[[V3:.+]] = vector.extract %[[TRANSPOSED]][3] : vector<2xf32> from vector<4x2xf32>
|
|
// INNER-PARALLEL: %[[RESULT_VEC:.+]] = arith.mulf %[[V3]], %[[RV012]] : vector<2xf32>
|
|
// INNER-PARALLEL: return %[[RESULT_VEC]] : vector<2xf32>
|
|
|
|
// -----
|
|
|
|
func.func @vector_multi_reduction_parallel_middle(%arg0: vector<3x4x5xf32>, %acc: vector<4xf32>) -> vector<4xf32> {
|
|
%0 = vector.multi_reduction <add>, %arg0, %acc [0, 2] : vector<3x4x5xf32> to vector<4xf32>
|
|
return %0 : vector<4xf32>
|
|
}
|
|
|
|
// ALL-LABEL: func @vector_multi_reduction_parallel_middle
|
|
// ALL-SAME: %[[INPUT:.+]]: vector<3x4x5xf32>, %[[ACC:.+]]: vector<4xf32>
|
|
// INNER-REDUCTION: vector.transpose %[[INPUT]], [1, 0, 2] : vector<3x4x5xf32> to vector<4x3x5xf32>
|
|
// INNER-PARALLEL: vector.transpose %[[INPUT]], [0, 2, 1] : vector<3x4x5xf32> to vector<3x5x4xf32>
|
|
|
|
// -----
|
|
|
|
func.func @vector_multi_reduction_masked(%arg0: vector<2x4xf32>, %acc: vector<2xf32>, %mask: vector<2x4xi1>) -> vector<2xf32> {
|
|
%0 = vector.mask %mask { vector.multi_reduction <mul>, %arg0, %acc [1] : vector<2x4xf32> to vector<2xf32> } : vector<2x4xi1> -> vector<2xf32>
|
|
return %0 : vector<2xf32>
|
|
}
|
|
|
|
// ALL-LABEL: func @vector_multi_reduction_masked
|
|
// ALL-SAME: %[[INPUT:.+]]: vector<2x4xf32>, %[[ACC:.+]]: vector<2xf32>, %[[MASK:.+]]: vector<2x4xi1>
|
|
// INNER-REDUCTION: %[[INNERVEC:.+]] = vector.extract %[[INPUT]][0] : vector<4xf32> from vector<2x4xf32>
|
|
// INNER-REDUCTION: %[[INNERACC:.+]] = vector.extract %[[ACC]][0] : f32 from vector<2xf32>
|
|
// INNER-REDUCTION: %[[INNERMASK:.+]] = vector.extract %[[MASK]][0] : vector<4xi1> from vector<2x4xi1>
|
|
// INNER-REDUCTION: vector.mask %[[INNERMASK]] { vector.reduction <mul>, %[[INNERVEC]], %[[INNERACC]] : vector<4xf32> into f32 } : vector<4xi1> -> f32
|
|
// INNER-PARALLEL: %[[TPMASK:.+]] = vector.transpose %[[MASK]], [1, 0] : vector<2x4xi1> to vector<4x2xi1>
|
|
// INNER-PARALLEL: %[[TPINPUT:.+]] = vector.transpose %[[INPUT]], [1, 0] : vector<2x4xf32> to vector<4x2xf32>
|
|
// INNER-PARALLEL: %[[INNERVEC:.+]] = vector.extract %[[TPINPUT]][0] : vector<2xf32> from vector<4x2xf32>
|
|
// INNER-PARALLEL: %[[INNERMASK:.+]] = vector.extract %[[TPMASK]][0] : vector<2xi1> from vector<4x2xi1>
|
|
// INNER-PARALLEL: %[[REDUCED:.+]] = arith.mulf %[[INNERVEC]], %[[ACC]] : vector<2xf32>
|
|
// INNER-PARALLEL: arith.select %[[INNERMASK]], %[[REDUCED]], %[[ACC]] : vector<2xi1>, vector<2xf32>
|