diff options
author | giuros01 <giuseppe.rossini@arm.com> | 2019-01-11 14:04:43 +0000 |
---|---|---|
committer | Giuseppe Rossini <giuseppe.rossini@arm.com> | 2019-01-14 12:42:42 +0000 |
commit | 1c9efebf4344e8db97e6d9282b2bf48b52090b58 (patch) | |
tree | 40c8737a23ff289cb176618d7c16fd4b2d811689 /arm_compute/runtime/CL/functions/CLGEMM.h | |
parent | 685c181d5ac36948f8df22f627ef9b3a74b5f87e (diff) | |
download | ComputeLibrary-1c9efebf4344e8db97e6d9282b2bf48b52090b58.tar.gz |
Issue COMPMID-1835: Remove CLGEMMInterleave4x4Kernel and replace with CLGEMMReshapeLHSMatrixKernel
Change-Id: Id6a1bd78f9b1698b64a004e4adebc41002b15745
Reviewed-on: https://review.mlplatform.org/496
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Diffstat (limited to 'arm_compute/runtime/CL/functions/CLGEMM.h')
-rw-r--r-- | arm_compute/runtime/CL/functions/CLGEMM.h | 25 |
1 files changed, 11 insertions, 14 deletions
diff --git a/arm_compute/runtime/CL/functions/CLGEMM.h b/arm_compute/runtime/CL/functions/CLGEMM.h index c4accde23d..624df33ef6 100644 --- a/arm_compute/runtime/CL/functions/CLGEMM.h +++ b/arm_compute/runtime/CL/functions/CLGEMM.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2016-2018 ARM Limited. + * Copyright (c) 2016-2019 ARM Limited. * * SPDX-License-Identifier: MIT * @@ -24,7 +24,6 @@ #ifndef __ARM_COMPUTE_CLGEMM_H__ #define __ARM_COMPUTE_CLGEMM_H__ -#include "arm_compute/core/CL/kernels/CLGEMMInterleave4x4Kernel.h" #include "arm_compute/core/CL/kernels/CLGEMMMatrixAdditionKernel.h" #include "arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyKernel.h" #include "arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyReshapedKernel.h" @@ -41,8 +40,7 @@ class ICLTensor; /** Basic function to execute GEMM on OpenCL. This function calls the following OpenCL kernels: * - * -# @ref CLGEMMInterleave4x4Kernel (only if the reshaped GEMM is selected by the heuristic model and the GPU target is NOT Mali-G76) - * -# @ref CLGEMMReshapeLHSMatrixKernel (only if the reshaped GEMM is selected by the heuristic model and the GPU target IS Mali-G76) + * -# @ref CLGEMMReshapeLHSMatrixKernel (only if the reshaped GEMM is selected by the heuristic model) * -# @ref CLGEMMReshapeRHSMatrixKernel (only if the reshaped GEMM is selected by the heuristic model) * -# @ref CLGEMMMatrixMultiplyKernel (if GPU target is NOT G76 or if the reshaped GEMM is NOT selected) * -# @ref CLGEMMMatrixMultiplyReshapedKernel (only if the reshaped GEMM is selected by the heuristic model and the GPU target IS Mali-G76) @@ -86,13 +84,13 @@ public: void configure(const ICLTensor *a, const ICLTensor *b, const ICLTensor *c, ICLTensor *output, float alpha, float beta, const GEMMInfo &gemm_info = GEMMInfo()); /** Static function to check if given info will lead to a valid configuration of @ref CLGEMM. * - * @param[in] a First input tensor info (Matrix or Vector A). Data types supported: F16/F32 - * @param[in] b Second input tensor info (Matrix B). Data type supported: same as @p a. - * @param[in] c Third input tensor info (Matrix C). It can be a nullptr if just the multiplication between @p a and @p b is needed. Data type supported: same as @p a. - * @param[out] output Output tensor info. Data type supported: same as @p a - * @param[in] alpha Weight of the matrix product - * @param[in] beta Weight of matrix C - * @param[in] gemm_info (Optional) Specifies if the matrix A and/or matrix B have been reshaped and + * @param[in] a First input tensor info (Matrix or Vector A). Data types supported: F16/F32 + * @param[in] b Second input tensor info (Matrix B). Data type supported: same as @p a. + * @param[in] c Third input tensor info (Matrix C). It can be a nullptr if just the multiplication between @p a and @p b is needed. Data type supported: same as @p a. + * @param[in] output Output tensor info. Data type supported: same as @p a + * @param[in] alpha Weight of the matrix product + * @param[in] beta Weight of matrix C + * @param[in] gemm_info (Optional) Specifies if the matrix A and/or matrix B have been reshaped and * if the reshape of matrix B should happen only for the first run * * @return a status @@ -105,7 +103,6 @@ public: private: CLMemoryGroup _memory_group; - CLGEMMInterleave4x4Kernel _interleave_kernel; // TODO - COMPMID-1835: Remove this kernel and use CLGEMMReshapeLHSMatrixKernel CLGEMMMatrixMultiplyKernel _mm_kernel; CLGEMMMatrixAdditionKernel _ma_kernel; CLGEMMReshapeLHSMatrixKernel _reshape_lhs_kernel; @@ -118,8 +115,8 @@ private: bool _run_addition; bool _reshape_b_only_on_first_run; bool _is_prepared; - bool _is_G76_path; // TODO: To be removed once completed COMPMID-1835 and COMPMID-1836 + bool _is_G76_path; }; -} +} // namespace arm_compute #endif /* __ARM_COMPUTE_CLGEMM_H__ */ |