diff options
author | Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> | 2023-09-27 17:46:17 +0100 |
---|---|---|
committer | felixjohnny.thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> | 2023-09-28 12:08:05 +0000 |
commit | afd38f0c617d6f89b2b4532c6c44f116617e2b6f (patch) | |
tree | 03bc7d5a762099989b16a656fa8d397b490ed70e /src/cpu/kernels/CpuGemmInterleave4x4Kernel.cpp | |
parent | bdcb4c148ee2fdeaaddf4cf1e57bbb0de02bb894 (diff) | |
download | ComputeLibrary-afd38f0c617d6f89b2b4532c6c44f116617e2b6f.tar.gz |
Apply clang-format on repository
Code is formatted as per a revised clang format configuration
file(not part of this delivery). Version 14.0.6 is used.
Exclusion List:
- files with .cl extension
- files that are not strictly C/C++ (e.g. Android.bp, Sconscript ...)
And the following directories
- compute_kernel_writer/validation/
- tests/
- include/
- src/core/NEON/kernels/convolution/
- src/core/NEON/kernels/arm_gemm/
- src/core/NEON/kernels/arm_conv/
- data/
There will be a follow up for formatting of .cl files and the
files under tests/ and compute_kernel_writer/validation/.
Signed-off-by: Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
Change-Id: Ib7eb1fcf4e7537b9feaefcfc15098a804a3fde0a
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/10391
Benchmark: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Gunes Bayir <gunes.bayir@arm.com>
Diffstat (limited to 'src/cpu/kernels/CpuGemmInterleave4x4Kernel.cpp')
-rw-r--r-- | src/cpu/kernels/CpuGemmInterleave4x4Kernel.cpp | 56 |
1 files changed, 32 insertions, 24 deletions
diff --git a/src/cpu/kernels/CpuGemmInterleave4x4Kernel.cpp b/src/cpu/kernels/CpuGemmInterleave4x4Kernel.cpp index 9fbf2d54c6..db433c99a8 100644 --- a/src/cpu/kernels/CpuGemmInterleave4x4Kernel.cpp +++ b/src/cpu/kernels/CpuGemmInterleave4x4Kernel.cpp @@ -24,9 +24,10 @@ #include "src/cpu/kernels/CpuGemmInterleave4x4Kernel.h" #include "arm_compute/core/ITensor.h" +#include "arm_compute/core/utils/misc/ShapeCalculator.h" #include "arm_compute/core/Validate.h" #include "arm_compute/core/Window.h" -#include "arm_compute/core/utils/misc/ShapeCalculator.h" + #include "src/core/helpers/AutoConfiguration.h" #include "src/core/helpers/WindowHelpers.h" @@ -60,7 +61,7 @@ Status CpuGemmInterleave4x4Kernel::validate(const ITensorInfo *src, const ITenso //Note: ARM_COMPUTE_RETURN_ERROR_ON_CPU_F16_UNSUPPORTED(src) is not needed here as this kernel doesn't use CPU FP16 instructions. ARM_COMPUTE_RETURN_ERROR_ON(src->data_type() == DataType::UNKNOWN); - if(dst->total_size() != 0) + if (dst->total_size() != 0) { const TensorShape dst_shape = compute_interleaved_shape(*src); ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DIMENSIONS(dst->tensor_shape(), dst_shape); @@ -111,35 +112,42 @@ void CpuGemmInterleave4x4Kernel::run_op(ITensorPack &tensors, const Window &wind Iterator in(src, win); Iterator out(dst, win_out); - execute_window_loop(win, [&](const Coordinates & id) - { - if(id.y() + 4 <= static_cast<int>(in_height)) + execute_window_loop( + win, + [&](const Coordinates &id) { - for(size_t x = window_start_x; x < window_end_x; ++x) + if (id.y() + 4 <= static_cast<int>(in_height)) { - std::memcpy(out.ptr() + (x * 4 + 0) * element_size, (in.ptr() + 0 * in_stride) + x * element_size, element_size); - std::memcpy(out.ptr() + (x * 4 + 1) * element_size, (in.ptr() + 1 * in_stride) + x * element_size, element_size); - std::memcpy(out.ptr() + (x * 4 + 2) * element_size, (in.ptr() + 2 * in_stride) + x * element_size, element_size); - std::memcpy(out.ptr() + (x * 4 + 3) * element_size, (in.ptr() + 3 * in_stride) + x * element_size, element_size); - } - } - else - { - for(size_t x = window_start_x; x < window_end_x; ++x) - { - size_t y = 0; - for(; y < partial_y; ++y) + for (size_t x = window_start_x; x < window_end_x; ++x) { - std::memcpy(out.ptr() + (x * 4 + y) * element_size, (in.ptr() + y * in_stride) + x * element_size, element_size); + std::memcpy(out.ptr() + (x * 4 + 0) * element_size, (in.ptr() + 0 * in_stride) + x * element_size, + element_size); + std::memcpy(out.ptr() + (x * 4 + 1) * element_size, (in.ptr() + 1 * in_stride) + x * element_size, + element_size); + std::memcpy(out.ptr() + (x * 4 + 2) * element_size, (in.ptr() + 2 * in_stride) + x * element_size, + element_size); + std::memcpy(out.ptr() + (x * 4 + 3) * element_size, (in.ptr() + 3 * in_stride) + x * element_size, + element_size); } - for(; y < 4; ++y) + } + else + { + for (size_t x = window_start_x; x < window_end_x; ++x) { - std::memset(out.ptr() + (x * 4 + y) * element_size, 0, element_size); + size_t y = 0; + for (; y < partial_y; ++y) + { + std::memcpy(out.ptr() + (x * 4 + y) * element_size, + (in.ptr() + y * in_stride) + x * element_size, element_size); + } + for (; y < 4; ++y) + { + std::memset(out.ptr() + (x * 4 + y) * element_size, 0, element_size); + } } } - } - }, - in, out); + }, + in, out); } const char *CpuGemmInterleave4x4Kernel::name() const |