diff options
Diffstat (limited to 'tensorflow/compiler/xla/service/gpu/ir_emitter_unnested.cc')
-rw-r--r-- | tensorflow/compiler/xla/service/gpu/ir_emitter_unnested.cc | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/tensorflow/compiler/xla/service/gpu/ir_emitter_unnested.cc b/tensorflow/compiler/xla/service/gpu/ir_emitter_unnested.cc index f7ddee7b61..b8bdd78da8 100644 --- a/tensorflow/compiler/xla/service/gpu/ir_emitter_unnested.cc +++ b/tensorflow/compiler/xla/service/gpu/ir_emitter_unnested.cc @@ -999,7 +999,7 @@ Status IrEmitterUnnested::EmitRowReduction( // for (shuffle_distance = 16; shuffle_distance > 0; shuffle_distance /= 2) // partial_result = Reducer( // partial_result, - // __shfl_down(partial_result, shuffle_distance)); + // __shfl_down_sync(CUDA_WARP_ALL, partial_result, shuffle_distance)); // if (lane_id == 0) // AtomicReducer(&output[y], partial_result); // } |