diff options
Diffstat (limited to 'tensorflow/compiler/xla/service/gpu/gemm_thunk.cc')
| -rw-r--r-- | tensorflow/compiler/xla/service/gpu/gemm_thunk.cc | 4 |
1 files changed, 3 insertions, 1 deletions
diff --git a/tensorflow/compiler/xla/service/gpu/gemm_thunk.cc b/tensorflow/compiler/xla/service/gpu/gemm_thunk.cc index 79fca43d02..dbc7754e25 100644 --- a/tensorflow/compiler/xla/service/gpu/gemm_thunk.cc +++ b/tensorflow/compiler/xla/service/gpu/gemm_thunk.cc @@ -252,7 +252,8 @@ GemmThunk::GemmThunk(const BufferAllocation::Slice& lhs_buffer, alpha_(alpha) {} Status GemmThunk::ExecuteOnStream(const BufferAllocations& buffer_allocations, - se::Stream* stream) { + se::Stream* stream, + HloExecutionProfiler* profiler) { VLOG(2) << "Executing a GemmThunk"; se::DeviceMemoryBase lhs_data = @@ -352,6 +353,7 @@ Status GemmThunk::ExecuteOnStream(const BufferAllocations& buffer_allocations, alpha_, stream); }; + auto op_profiler = profiler->MakeScopedInstructionProfiler(hlo_instruction()); bool launch_ok; if (LayoutUtil::Minor(output_shape_.layout(), 0) == 0) { launch_ok = launch( |