diff options
author | ramelg01 <ramy.elgammal@arm.com> | 2022-04-07 02:42:52 +0100 |
---|---|---|
committer | Ramy Elgammal <ramy.elgammal@arm.com> | 2022-04-26 15:51:22 +0000 |
commit | 8a164884dddf769643cf3b9f7f94e43cb4f3c20b (patch) | |
tree | 35958dd48b6df1a851c880dad2b2ce285671b611 /src/core/NEON/kernels/assembly | |
parent | c827e99fc46521f43719b0c2d1b6f05d66abf68c (diff) | |
download | ComputeLibrary-8a164884dddf769643cf3b9f7f94e43cb4f3c20b.tar.gz |
Update Neon™ depthwise kernel
- Reduce duplication and simplify overall structure.
- Improve multi-threaded performance by sharing more data
in lower-level caches.
Partially Resolves: COMPMID-5054
Signed-off-by: Ramy Elgammal <ramy.elgammal@arm.com>
Change-Id: Iac747f39b21c540122fa75218762631c4d787911
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/7449
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Andrew Mundy
Reviewed-by: Sheri Zhang <sheri.zhang@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/core/NEON/kernels/assembly')
-rw-r--r-- | src/core/NEON/kernels/assembly/depthwise.hpp | 42 |
1 files changed, 38 insertions, 4 deletions
diff --git a/src/core/NEON/kernels/assembly/depthwise.hpp b/src/core/NEON/kernels/assembly/depthwise.hpp index 9262ea05a4..3998dfbc9a 100644 --- a/src/core/NEON/kernels/assembly/depthwise.hpp +++ b/src/core/NEON/kernels/assembly/depthwise.hpp @@ -59,6 +59,8 @@ struct DepthwiseArgs const DepthwiseConfig *config; + bool fast_mode = false; + DepthwiseArgs( const CPUInfo *cpu_info, unsigned int kernel_rows, unsigned int kernel_cols, @@ -83,15 +85,18 @@ private: protected: const DepthwiseArgs m_args; // Copy of arguments + public: std::string name() const { return _name; } + void set_name(const std::string &n) { _name = n; } + DepthwiseCommon(const DepthwiseArgs &args) : m_args(args) {}; DepthwiseCommon(DepthwiseCommon &) = delete; @@ -103,7 +108,7 @@ public: void *const output, void *const working_space, const unsigned int thread_id, - const unsigned int n_threads) const override + const unsigned int n_threads) const override final { const size_t ld_input_col = m_args.input_channels; const size_t ld_input_row = ld_input_col * m_args.input_cols; @@ -130,7 +135,7 @@ public: size_t ld_output_batch, void *const working_space, const unsigned int thread_id, - const unsigned int n_threads) const override + const unsigned int n_threads) const override final { execute( m_args.n_batches, m_args.input_rows, m_args.input_cols, @@ -142,7 +147,36 @@ public: working_space, thread_id, n_threads); } - virtual void execute( + void execute( + unsigned int batches, + unsigned int input_height, + unsigned int input_width, + unsigned int channels, + const PaddingValues &padding, + const void *input, + size_t ld_input_col, + size_t ld_input_row, + size_t ld_input_batch, + const void *parameters, + unsigned int output_height, + unsigned int output_width, + void *output, + size_t ld_output_col, + size_t ld_output_row, + size_t ld_output_batch, + void *working_space, + unsigned int thread_id, + unsigned int n_threads) const override final + { + this->execute_internal( + batches, input_height, input_width, channels, padding, input, + ld_input_col, ld_input_row, ld_input_batch, parameters, output_height, + output_width, output, ld_output_col, ld_output_row, ld_output_batch, + working_space, thread_id, n_threads); + } + +protected: + virtual void execute_internal( unsigned int batches, unsigned int input_height, unsigned int input_width, @@ -161,7 +195,7 @@ public: size_t ld_output_batch, void *working_space, unsigned int thread_id, - unsigned int n_threads) const override = 0; + unsigned int n_threads) const = 0; }; template <typename TInput, typename TWeight = TInput, typename TOutput = TInput> |