diff options
author | Matthew Bentham <matthew.bentham@arm.com> | 2020-03-05 23:37:48 +0000 |
---|---|---|
committer | Michele Di Giorgio <michele.digiorgio@arm.com> | 2020-03-10 18:41:04 +0000 |
commit | 758b5ba3e6d22509d4deab3d8b0b9c2f03418130 (patch) | |
tree | c16734e12b105819a919e8015a4d281dffd8817c /src/core/GLES_COMPUTE/kernels/GCDirectConvolutionLayerKernel.cpp | |
parent | 6b3865ad038d60a126fe1f90df815a480527a29f (diff) | |
download | ComputeLibrary-758b5ba3e6d22509d4deab3d8b0b9c2f03418130.tar.gz |
COMPMID-3069: Improve build time by splitting up ToolchainSupport.h
Split out the parts of ToolchainSupport coming from <memory> and
the parts coming from <string> into their own new header files.
This accounts for 99% of uses of ToolchainSupport, which means that
expensive header files such as arm_neon.h don't need to be included
everywhere.
Knocks about 10% of compilation time off kernel files.
Signed-off-by: Matthew Bentham <matthew.bentham@arm.com>
Change-Id: I2ae718fe766b5ff28608812b0f686f30eeac1b21
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/2852
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Diffstat (limited to 'src/core/GLES_COMPUTE/kernels/GCDirectConvolutionLayerKernel.cpp')
-rw-r--r-- | src/core/GLES_COMPUTE/kernels/GCDirectConvolutionLayerKernel.cpp | 22 |
1 files changed, 11 insertions, 11 deletions
diff --git a/src/core/GLES_COMPUTE/kernels/GCDirectConvolutionLayerKernel.cpp b/src/core/GLES_COMPUTE/kernels/GCDirectConvolutionLayerKernel.cpp index ecff233382..f3e47d9ae9 100644 --- a/src/core/GLES_COMPUTE/kernels/GCDirectConvolutionLayerKernel.cpp +++ b/src/core/GLES_COMPUTE/kernels/GCDirectConvolutionLayerKernel.cpp @@ -1,5 +1,5 @@ /* - * Copyright (c) 2017-2018 ARM Limited. + * Copyright (c) 2017-2020 ARM Limited. * * SPDX-License-Identifier: MIT * @@ -33,7 +33,7 @@ #include "arm_compute/core/ITensor.h" #include "arm_compute/core/Types.h" #include "arm_compute/core/Validate.h" -#include "support/ToolchainSupport.h" +#include "support/StringSupport.h" using namespace arm_compute; @@ -44,7 +44,7 @@ GCDirectConvolutionLayerKernel<kernel_size>::GCDirectConvolutionLayerKernel() } template <unsigned int kernel_size> -BorderSize GCDirectConvolutionLayerKernel<kernel_size>::border_size() const +BorderSize GCDirectConvolutionLayerKernel<kernel_size>::border_size() const { return _border_size; } @@ -70,8 +70,8 @@ void GCDirectConvolutionLayerKernel<kernel_size>::configure(const IGCTensor *inp } // Get convolved dimensions - unsigned int owidth = 0; - unsigned int oheight = 0; + unsigned int owidth = 0; + unsigned int oheight = 0; std::tie(owidth, oheight) = scaled_dimensions(input->info()->dimension(0), input->info()->dimension(1), kernel_size, kernel_size, conv_info); TensorShape output_shape = input->info()->tensor_shape(); @@ -238,20 +238,20 @@ void GCDirectConvolutionLayerKernel<kernel_size>::configure(const IGCTensor *inp num_elems_written_per_iteration_x = 4; #elif defined(PROCESS_4X_2Y_1Z) options.emplace("#define PROCESS_4X_2Y_1Z"); - num_elems_read_per_iteration_x = 4; - num_elems_read_per_iteration_y = 2; + num_elems_read_per_iteration_x = 4; + num_elems_read_per_iteration_y = 2; num_elems_written_per_iteration_x = 4; num_elems_written_per_iteration_y = 2; #elif defined(PROCESS_4X_3Y_1Z) options.emplace("#define PROCESS_4X_3Y_1Z"); - num_elems_read_per_iteration_x = 4; - num_elems_read_per_iteration_y = 3; + num_elems_read_per_iteration_x = 4; + num_elems_read_per_iteration_y = 3; num_elems_written_per_iteration_x = 4; num_elems_written_per_iteration_y = 3; #elif defined(PROCESS_4X_4Y_1Z) options.emplace("#define PROCESS_4X_4Y_1Z"); - num_elems_read_per_iteration_x = 4; - num_elems_read_per_iteration_y = 4; + num_elems_read_per_iteration_x = 4; + num_elems_read_per_iteration_y = 4; num_elems_written_per_iteration_x = 4; num_elems_written_per_iteration_y = 4; #elif defined(PROCESS_4X_2Y_2Z) |