Implement OpenCL MatMul for Lhs NT Rhs T/NT FP32/16
- Implement ClNativeMatMulKernel class
- Implement opencl kernel for LHS non-transposed and RHS non-transposed
- Implement opencl kernel for LHS non-transposed and RHS transposed
- Add test fixture and dataset for matmul
- Implement transpose_tensor() for reference implementation to transpose high dimensional tensors
Resolves: COMPMID-5944, COMPMID-5951
Co-authored-by: Gunes Bayir <gunes.bayir@arm.com>
Co-authored-by: Ramy Elgammal <ramy.elgammal@arm.com>
Change-Id: I1d5b8978f41be27baddb3153ade880472141573f
Signed-off-by: Gunes Bayir <gunes.bayir@arm.com>
Signed-off-by: Ramy Elgammal <ramy.elgammal@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/9333
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Benchmark: Arm Jenkins <bsgcomp@arm.com>
diff --git a/arm_compute/core/KernelDescriptors.h b/arm_compute/core/KernelDescriptors.h
index 19ac254..016e03d 100644
--- a/arm_compute/core/KernelDescriptors.h
+++ b/arm_compute/core/KernelDescriptors.h
@@ -21,8 +21,8 @@
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
-#ifndef ARM_COMPUTE_CORE_KERNEL_DESCRIPTORS_H
-#define ARM_COMPUTE_CORE_KERNEL_DESCRIPTORS_H
+#ifndef ACL_ARM_COMPUTE_CORE_KERNELDESCRIPTORS
+#define ACL_ARM_COMPUTE_CORE_KERNELDESCRIPTORS
#include "arm_compute/core/PixelValue.h"
#include "arm_compute/core/Types.h"
@@ -223,5 +223,19 @@
bool align_corners; /**< Align corners of input and output */
DataLayout data_layout; /**< Data layout to use */
};
+
+struct MatMulKernelInfo
+{
+ MatMulKernelInfo(bool adj_lhs = false, bool adj_rhs = false, int m0 = 1, int n0 = 1, int k0 = 1, bool export_rhs_to_cl_image = false)
+ : adj_lhs{ adj_lhs }, adj_rhs{ adj_rhs }, m0{ m0 }, n0{ n0 }, k0{ k0 }, export_rhs_to_cl_image{ export_rhs_to_cl_image }
+ {
+ }
+ bool adj_lhs{ false }; /**< Get Adjoint LHS flag value */
+ bool adj_rhs{ false }; /**< Get Adjoint RHS flag value */
+ int m0{ 1 }; /**< Number of output rows processed by each work-item*/
+ int n0{ 1 }; /**< Number of output columns processed by each work-item*/
+ int k0{ 1 }; /**< Number of inner accumulations */
+ bool export_rhs_to_cl_image{ false }; /**< Flag to know whether the RHS tensor should be exported to cl_image*/
+};
} // namespace arm_compute
-#endif /* ARM_COMPUTE_CORE_KERNEL_DESCRIPTORS_H */
+#endif /* ACL_ARM_COMPUTE_CORE_KERNELDESCRIPTORS */