diff options
| author | Giorgio Arena <giorgio.arena@arm.com> | 2019-08-05 17:15:40 +0100 |
|---|---|---|
| committer | Giorgio Arena <giorgio.arena@arm.com> | 2019-08-06 07:57:32 +0000 |
| commit | 4746326ecb075dcfa123aaa8b38de5ec3e534b60 (patch) | |
| tree | f0d9027f0d4ea6c72f9f06d6e8f09fe72939e326 /src/core/CL/kernels/CLBatchNormalizationLayerKernel.cpp | |
| parent | d820db6fc479f7daef6788377cb765369fcddc22 (diff) | |
| download | ComputeLibrary-4746326ecb075dcfa123aaa8b38de5ec3e534b60.tar.gz | |
COMPMID-2336 Fix valgrind error for BatchNormalizationLayer on NEON with NHWC layout
Change-Id: I9ed2d0647ae3c33bce6290acfdac356ffffcb709
Signed-off-by: Giorgio Arena <giorgio.arena@arm.com>
Reviewed-on: https://review.mlplatform.org/c/1697
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/core/CL/kernels/CLBatchNormalizationLayerKernel.cpp')
| -rw-r--r-- | src/core/CL/kernels/CLBatchNormalizationLayerKernel.cpp | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/src/core/CL/kernels/CLBatchNormalizationLayerKernel.cpp b/src/core/CL/kernels/CLBatchNormalizationLayerKernel.cpp index f9b975392c..ae646a79b8 100644 --- a/src/core/CL/kernels/CLBatchNormalizationLayerKernel.cpp +++ b/src/core/CL/kernels/CLBatchNormalizationLayerKernel.cpp @@ -107,6 +107,7 @@ std::pair<Status, Window> validate_and_configure_window(ITensorInfo *input, ITen window_changed = update_window_and_padding(win, input_access); } + // Mean, var, gamma and beta get parallelized for the NHWC case as they follow the channel dimension, which is along the first axis if(input->data_layout() == DataLayout::NHWC) { AccessWindowHorizontal mean_access(mean, 0, num_elems_processed_per_iteration); |