diff options
author | Frank Lei <frank.lei@arm.com> | 2018-02-01 14:47:14 +0800 |
---|---|---|
committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:47:40 +0000 |
commit | 4406fd6cc4abded564d3791324e1f48bdfd34273 (patch) | |
tree | 22fe402fe9ac7ca338df49e9eccd6eb1587ae875 /src/core/GLES_COMPUTE/kernels/GCBatchNormalizationLayerKernel.cpp | |
parent | 898d399a0f62c15612a52df4bff5018e783214e4 (diff) | |
download | ComputeLibrary-4406fd6cc4abded564d3791324e1f48bdfd34273.tar.gz |
APPBROWSER-391: Fix GLES COMPUTE alignment issues
APPBROWSER-402: Performance optimization for squeezenet/xray model
Change-Id: If31b186b99a6d6087164019fe94d3ac9279e3204
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/119526
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Diffstat (limited to 'src/core/GLES_COMPUTE/kernels/GCBatchNormalizationLayerKernel.cpp')
-rw-r--r-- | src/core/GLES_COMPUTE/kernels/GCBatchNormalizationLayerKernel.cpp | 11 |
1 files changed, 8 insertions, 3 deletions
diff --git a/src/core/GLES_COMPUTE/kernels/GCBatchNormalizationLayerKernel.cpp b/src/core/GLES_COMPUTE/kernels/GCBatchNormalizationLayerKernel.cpp index a41b62fbab..cd93f6997e 100644 --- a/src/core/GLES_COMPUTE/kernels/GCBatchNormalizationLayerKernel.cpp +++ b/src/core/GLES_COMPUTE/kernels/GCBatchNormalizationLayerKernel.cpp @@ -119,7 +119,10 @@ void GCBatchNormalizationLayerKernel::run(const Window &window) _kernel.use(); - Window slice = window.first_slice_window_3D(); + _output->set_needs_shifting(true); + + Window slice = window.first_slice_window_3D(); + Window slice_in = window.first_slice_window_3D(); Window vector_slice = window.first_slice_window_1D(); vector_slice.set(Window::DimX, Window::Dimension(0, 0, 0)); @@ -130,14 +133,16 @@ void GCBatchNormalizationLayerKernel::run(const Window &window) add_1D_tensor_argument(idx, _beta, 5, vector_slice); add_1D_tensor_argument(idx, _gamma, 6, vector_slice); + slice.shift(Window::DimX, -(_output->info()->padding()).left); + do { idx = 0; - add_3D_tensor_argument(idx, _input, 1, slice); + add_3D_tensor_argument(idx, _input, 1, slice_in); add_3D_tensor_argument(idx, _output, 2, slice); _kernel.update_shader_params(); enqueue(*this, slice); } - while(window.slide_window_slice_3D(slice)); + while(window.slide_window_slice_3D(slice) && window.slide_window_slice_3D(slice_in)); } |