Skip to content

[mlir][gpu] Use alloc OP's host_shared in cuda runtime #99035

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Jul 17, 2024
Merged

[mlir][gpu] Use alloc OP's host_shared in cuda runtime #99035

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 10 additions & 3 deletions mlir/lib/ExecutionEngine/CudaRuntimeWrappers.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -237,11 +237,18 @@ extern "C" MLIR_CUDA_WRAPPERS_EXPORT void mgpuEventRecord(CUevent event,
}

extern "C" MLIR_CUDA_WRAPPERS_EXPORT void *
mgpuMemAlloc(uint64_t sizeBytes, CUstream /*stream*/, bool /*isHostShared*/) {
mgpuMemAlloc(uint64_t sizeBytes, CUstream stream, bool isHostShared) {
ScopedContext scopedContext;
CUdeviceptr ptr = 0;
if (sizeBytes != 0)
CUDA_REPORT_IF_ERROR(cuMemAlloc(&ptr, sizeBytes));
if (sizeBytes == 0)
return reinterpret_cast<void *>(ptr);

if (isHostShared) {
CUDA_REPORT_IF_ERROR(
cuMemAllocManaged(&ptr, sizeBytes, CU_MEM_ATTACH_GLOBAL));
return reinterpret_cast<void *>(ptr);
}
CUDA_REPORT_IF_ERROR(cuMemAlloc(&ptr, sizeBytes));
return reinterpret_cast<void *>(ptr);
}

Expand Down
27 changes: 27 additions & 0 deletions mlir/test/Integration/GPU/CUDA/alloc-host-shared.mlir
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,27 @@
// RUN: mlir-opt %s \
// RUN: | mlir-opt -gpu-lower-to-nvvm-pipeline="cubin-format=%gpu_compilation_format" \
// RUN: | mlir-cpu-runner \
// RUN: --shared-libs=%mlir_cuda_runtime \
// RUN: --shared-libs=%mlir_runner_utils \
// RUN: --entry-point-result=void \
// RUN: | FileCheck %s

// CHECK: 2000
module attributes {gpu.container_module} {
func.func @main() {
%c1 = arith.constant 1 : index
%c0 = arith.constant 0 : index
%c1000_i32 = arith.constant 1000 : i32
%memref = gpu.alloc host_shared () : memref<1xi32>
memref.store %c1000_i32, %memref[%c1] : memref<1xi32>
gpu.launch blocks(%arg0, %arg1, %arg2) in (%arg6 = %c1, %arg7 = %c1, %arg8 = %c1) threads(%arg3, %arg4, %arg5) in (%arg9 = %c1, %arg10 = %c1, %arg11 = %c1) {
%1 = memref.load %memref[%c1] : memref<1xi32>
%2 = arith.addi %1, %1 : i32
memref.store %2, %memref[%c1] : memref<1xi32>
gpu.terminator
}
%0 = memref.load %memref[%c1] : memref<1xi32>
vector.print %0 : i32
return
}
}
Loading