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
69 lines
2.4 KiB
MLIR
69 lines
2.4 KiB
MLIR
// RUN: mlir-opt %s \
|
|
// RUN: | mlir-opt -gpu-lower-to-nvvm-pipeline="cubin-format=%gpu_compilation_format allow-pattern-rollback=0" \
|
|
// RUN: | mlir-runner \
|
|
// RUN: --shared-libs=%mlir_cuda_runtime \
|
|
// RUN: --shared-libs=%mlir_runner_utils \
|
|
// RUN: --entry-point-result=void \
|
|
// RUN: | FileCheck %s
|
|
|
|
func.func @main() {
|
|
%data = memref.alloc() : memref<2x6xi32>
|
|
%sum = memref.alloc() : memref<2xi32>
|
|
%cst0 = arith.constant 0 : i32
|
|
%cst1 = arith.constant 1 : i32
|
|
%cst2 = arith.constant 2 : i32
|
|
%cst4 = arith.constant 4 : i32
|
|
%cst8 = arith.constant 8 : i32
|
|
%cst16 = arith.constant 16 : i32
|
|
|
|
%cst3 = arith.constant 3 : i32
|
|
%cst6 = arith.constant 6 : i32
|
|
%cst7 = arith.constant 7 : i32
|
|
%cst10 = arith.constant 10 : i32
|
|
%cst11 = arith.constant 11 : i32
|
|
|
|
%c0 = arith.constant 0 : index
|
|
%c1 = arith.constant 1 : index
|
|
%c2 = arith.constant 2 : index
|
|
%c3 = arith.constant 3 : index
|
|
%c4 = arith.constant 4 : index
|
|
%c5 = arith.constant 5 : index
|
|
%c6 = arith.constant 6 : index
|
|
|
|
%cast_data = memref.cast %data : memref<2x6xi32> to memref<*xi32>
|
|
gpu.host_register %cast_data : memref<*xi32>
|
|
%cast_sum = memref.cast %sum : memref<2xi32> to memref<*xi32>
|
|
gpu.host_register %cast_sum : memref<*xi32>
|
|
|
|
memref.store %cst0, %data[%c0, %c0] : memref<2x6xi32>
|
|
memref.store %cst1, %data[%c0, %c1] : memref<2x6xi32>
|
|
memref.store %cst2, %data[%c0, %c2] : memref<2x6xi32>
|
|
memref.store %cst4, %data[%c0, %c3] : memref<2x6xi32>
|
|
memref.store %cst8, %data[%c0, %c4] : memref<2x6xi32>
|
|
memref.store %cst16, %data[%c0, %c5] : memref<2x6xi32>
|
|
|
|
memref.store %cst2, %data[%c1, %c0] : memref<2x6xi32>
|
|
memref.store %cst3, %data[%c1, %c1] : memref<2x6xi32>
|
|
memref.store %cst6, %data[%c1, %c2] : memref<2x6xi32>
|
|
memref.store %cst7, %data[%c1, %c3] : memref<2x6xi32>
|
|
memref.store %cst10, %data[%c1, %c4] : memref<2x6xi32>
|
|
memref.store %cst11, %data[%c1, %c5] : memref<2x6xi32>
|
|
|
|
// MAX
|
|
gpu.launch blocks(%bx, %by, %bz) in (%grid_x = %c2, %grid_y = %c1, %grid_z = %c1)
|
|
threads(%tx, %ty, %tz) in (%block_x = %c6, %block_y = %c1, %block_z = %c1) {
|
|
%val = memref.load %data[%bx, %tx] : memref<2x6xi32>
|
|
%reduced = gpu.all_reduce maxsi %val uniform {} : (i32) -> (i32)
|
|
memref.store %reduced, %sum[%bx] : memref<2xi32>
|
|
gpu.terminator
|
|
}
|
|
|
|
call @printMemrefI32(%cast_sum) : (memref<*xi32>) -> ()
|
|
// CHECK: [16, 11]
|
|
|
|
return
|
|
}
|
|
|
|
func.func private @printMemrefI32(memref<*xi32>)
|
|
|