From 7026b303d636e7639f8877ae8d5eff54f39c1121 Mon Sep 17 00:00:00 2001 From: Gian Marco Iodice Date: Wed, 26 Jun 2019 17:18:11 +0100 Subject: COMPMID-1979: Fuse Activation Function in CLGEMM - part 1 Implementing a new struct to contains the information for the OpenCL GEMM kernels Change-Id: I6c641c312f9c3b025a7c69dd0df3b730d2d2c2cb Signed-off-by: Gian Marco Iodice Reviewed-on: https://review.mlplatform.org/c/1434 Tested-by: Arm Jenkins Reviewed-by: Giuseppe Rossini --- arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyNativeKernel.h | 6 ++++-- arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyReshapedKernel.h | 6 ++++-- .../core/CL/kernels/CLGEMMMatrixMultiplyReshapedOnlyRHSKernel.h | 6 ++++-- 3 files changed, 12 insertions(+), 6 deletions(-) (limited to 'arm_compute/core/CL') diff --git a/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyNativeKernel.h b/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyNativeKernel.h index 79689a2894..96f412c6a5 100644 --- a/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyNativeKernel.h +++ b/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyNativeKernel.h @@ -26,6 +26,8 @@ #include "arm_compute/core/CL/ICLKernel.h" +#include "arm_compute/core/KernelDescriptors.h" + namespace arm_compute { class ICLTensor; @@ -62,7 +64,7 @@ public: */ void configure(const ICLTensor *input0, const ICLTensor *input1, const ICLTensor *input2, ICLTensor *output, float alpha, float beta, const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, - const GEMMReshapeInfo &gemm_info); + const GEMMKernelInfo &gemm_info); /** Static function to check if given info will lead to a valid configuration of @ref CLGEMMMatrixMultiplyNativeKernel * * @param[in] input0 Input tensor info for the LHS matrix. Data type supported: F32/F16. The number of dimensions for the LHS matrix must be less or equal than 4. @@ -83,7 +85,7 @@ public: */ static Status validate(const ITensorInfo *input0, const ITensorInfo *input1, const ITensorInfo *input2, const ITensorInfo *output, float alpha, float beta, const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, - const GEMMReshapeInfo &gemm_info); + const GEMMKernelInfo &gemm_info); // Inherited methods overridden: void run(const Window &window, cl::CommandQueue &queue) override; diff --git a/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyReshapedKernel.h b/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyReshapedKernel.h index 68ab94a31d..47916b3019 100644 --- a/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyReshapedKernel.h +++ b/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyReshapedKernel.h @@ -26,6 +26,8 @@ #include "arm_compute/core/CL/ICLKernel.h" +#include "arm_compute/core/KernelDescriptors.h" + namespace arm_compute { class ICLTensor; @@ -69,7 +71,7 @@ public: */ void configure(const ICLTensor *input0, const ICLTensor *input1, const ICLTensor *input2, ICLTensor *output, float alpha, float beta, const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, - const GEMMReshapeInfo &gemm_info); + const GEMMKernelInfo &gemm_info); /** Static function to check if given info will lead to a valid configuration of @ref CLGEMMMatrixMultiplyReshapedKernel * * @param[in] input0 Input tensor containing the LHS reshaped matrix. Data type supported: F32/F16. The number of dimensions for the LHS matrix must be less or equal than 4 @@ -94,7 +96,7 @@ public: */ static Status validate(const ITensorInfo *input0, const ITensorInfo *input1, const ITensorInfo *input2, const ITensorInfo *output, float alpha, float beta, const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, - const GEMMReshapeInfo &gemm_info); + const GEMMKernelInfo &gemm_info); // Inherited methods overridden: void run(const Window &window, cl::CommandQueue &queue) override; diff --git a/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyReshapedOnlyRHSKernel.h b/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyReshapedOnlyRHSKernel.h index e3b3880a37..3315331e87 100644 --- a/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyReshapedOnlyRHSKernel.h +++ b/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyReshapedOnlyRHSKernel.h @@ -26,6 +26,8 @@ #include "arm_compute/core/CL/ICLKernel.h" +#include "arm_compute/core/KernelDescriptors.h" + namespace arm_compute { class ICLTensor; @@ -65,7 +67,7 @@ public: */ void configure(const ICLTensor *input0, const ICLTensor *input1, const ICLTensor *input2, ICLTensor *output, float alpha, float beta, const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, - const GEMMReshapeInfo &gemm_info); + const GEMMKernelInfo &gemm_info); /** Static function to check if given info will lead to a valid configuration of @ref CLGEMMMatrixMultiplyReshapedOnlyRHSKernel * * @param[in] input0 Input tensor info for the LHS matrix. Data type supported: F32/F16. The number of dimensions for the LHS matrix must be less or equal than 4. @@ -86,7 +88,7 @@ public: */ static Status validate(const ITensorInfo *input0, const ITensorInfo *input1, const ITensorInfo *input2, const ITensorInfo *output, float alpha, float beta, const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, - const GEMMReshapeInfo &gemm_info); + const GEMMKernelInfo &gemm_info); // Inherited methods overridden: void run(const Window &window, cl::CommandQueue &queue) override; -- cgit v1.2.1