OSSForks/tensorflow
1
0
Fork 0
mirror of https://github.com/zebrajr/tensorflow.git synced 2025-12-06 12:20:11 +01:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

[XLA:GPU] follow up fix after pr#32919

Browse Source 
Change `gpu_version` parameter to const reference in `IntelGpuCompiler`.

This aligns the parameter type in `OptimizeHloConvolutionCanonicalization` with the base class signature.

PiperOrigin-RevId: 825083863
This commit is contained in:
Ilya Tikhonovskiy 2025-10-28 09:50:10 -07:00 committed by TensorFlower Gardener
parent d75ad2c4ff
commit 14db6f6317
2 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
third_party/xla/xla/service/gpu/intel_gpu_compiler.cc vendored
View File

@ -29,7 +29,7 @@ IntelGpuCompiler::IntelGpuCompiler()
spir::DataLayout()) {}
absl::Status IntelGpuCompiler::OptimizeHloConvolutionCanonicalization(
HloModule* hlo_module, se::GpuComputeCapability gpu_version,
HloModule* hlo_module, const se::GpuComputeCapability& gpu_version,
se::dnn::VersionInfo dnn_version,
const se::SemanticVersion& toolkit_version) {
// Note: this is a stub.

2
third_party/xla/xla/service/gpu/intel_gpu_compiler.h vendored
View File

@ -35,7 +35,7 @@ class IntelGpuCompiler : public GpuCompiler {
IntelGpuCompiler();
absl::Status OptimizeHloConvolutionCanonicalization(
HloModule* hlo_module, se::GpuComputeCapability gpu_version,
HloModule* hlo_module, const se::GpuComputeCapability& gpu_version,
se::dnn::VersionInfo dnn_version,
const se::SemanticVersion& toolkit_version) override;