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/runtime/GLES_COMPUTE/functions/GCDepthwiseConvolutionLayer.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/runtime/GLES_COMPUTE/functions/GCDepthwiseConvolutionLayer.cpp')
| -rw-r--r-- | src/runtime/GLES_COMPUTE/functions/GCDepthwiseConvolutionLayer.cpp | 18 |
1 files changed, 17 insertions, 1 deletions
diff --git a/src/runtime/GLES_COMPUTE/functions/GCDepthwiseConvolutionLayer.cpp b/src/runtime/GLES_COMPUTE/functions/GCDepthwiseConvolutionLayer.cpp index ef65989f40..9cba37110b 100644 --- a/src/runtime/GLES_COMPUTE/functions/GCDepthwiseConvolutionLayer.cpp +++ b/src/runtime/GLES_COMPUTE/functions/GCDepthwiseConvolutionLayer.cpp @@ -1,5 +1,5 @@ /* - * Copyright (c) 2017 ARM Limited. + * Copyright (c) 2017-2018 ARM Limited. * * SPDX-License-Identifier: MIT * @@ -30,6 +30,11 @@ using namespace arm_compute; +GCDepthwiseConvolutionLayer3x3::GCDepthwiseConvolutionLayer3x3() + : _kernel(nullptr), _border_handler(), _shift_handler() +{ +} + void GCDepthwiseConvolutionLayer3x3::configure(IGCTensor *input, const IGCTensor *weights, const IGCTensor *biases, IGCTensor *output, const PadStrideInfo &conv_info) { auto k = arm_compute::support::cpp14::make_unique<GCDepthwiseConvolutionLayer3x3Kernel>(); @@ -38,4 +43,15 @@ void GCDepthwiseConvolutionLayer3x3::configure(IGCTensor *input, const IGCTensor // Configure border handler _border_handler.configure(input, _kernel->border_size(), BorderMode::CONSTANT, PixelValue(0)); + + _shift_handler.configure(input); +} + +void GCDepthwiseConvolutionLayer3x3::run() +{ + GCScheduler::get().dispatch(_shift_handler, false); + GCScheduler::get().memory_barrier(); + GCScheduler::get().dispatch(_border_handler, false); + GCScheduler::get().memory_barrier(); + GCScheduler::get().dispatch(*_kernel); } |