diff options
author | Giorgio Arena <giorgio.arena@arm.com> | 2018-07-18 16:13:12 +0100 |
---|---|---|
committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:54:54 +0000 |
commit | 0f170396e84836ad8c54d54421e95c61812968be (patch) | |
tree | b8993f251e3c023aca2856b2784e67eb9f11c8a4 /src/runtime/CL | |
parent | b6eb35371d222c6b7f61210d97ebd7dd9e197458 (diff) | |
download | ComputeLibrary-0f170396e84836ad8c54d54421e95c61812968be.tar.gz |
COMPMID-1342 Add grouping support to CLIm2ColKernel
Change-Id: I4afb19751520a90fee27fb49b775cd10e92a94f5
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/140476
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/runtime/CL')
-rw-r--r-- | src/runtime/CL/functions/CLGEMM.cpp | 6 | ||||
-rw-r--r-- | src/runtime/CL/functions/CLGEMMConvolutionLayer.cpp | 2 |
2 files changed, 3 insertions, 5 deletions
diff --git a/src/runtime/CL/functions/CLGEMM.cpp b/src/runtime/CL/functions/CLGEMM.cpp index a8d7058f2a..1ad8531920 100644 --- a/src/runtime/CL/functions/CLGEMM.cpp +++ b/src/runtime/CL/functions/CLGEMM.cpp @@ -181,7 +181,6 @@ Status CLGEMM::validate(const ITensorInfo *a, const ITensorInfo *b, const ITenso TensorInfo tmp_a_info{}; TensorInfo tmp_b_info{}; - TensorInfo tmp_output_info{}; // Get the GPU target const GPUTarget gpu_target = CLScheduler::get().target(); @@ -229,13 +228,12 @@ Status CLGEMM::validate(const ITensorInfo *a, const ITensorInfo *b, const ITenso } // Validate matrix multiply - auto_init_if_empty(tmp_output_info, matrix_a_info->clone()->set_tensor_shape(compute_mm_shape(*matrix_a_info, *matrix_b_info, run_interleave_transpose, reshape_info))); - ARM_COMPUTE_RETURN_ON_ERROR(CLGEMMMatrixMultiplyKernel::validate(matrix_a_info, matrix_b_info, &tmp_output_info, alpha, run_interleave_transpose, reshape_info, gpu_target)); + ARM_COMPUTE_RETURN_ON_ERROR(CLGEMMMatrixMultiplyKernel::validate(matrix_a_info, matrix_b_info, output, alpha, run_interleave_transpose, reshape_info, gpu_target)); if(beta != 0 && c != nullptr) { // Validate matrix addition kernel - ARM_COMPUTE_RETURN_ON_ERROR(CLGEMMMatrixAdditionKernel::validate(c, &tmp_output_info, beta)); + ARM_COMPUTE_RETURN_ON_ERROR(CLGEMMMatrixAdditionKernel::validate(c, output, beta)); } return Status{}; diff --git a/src/runtime/CL/functions/CLGEMMConvolutionLayer.cpp b/src/runtime/CL/functions/CLGEMMConvolutionLayer.cpp index ca6157ef13..26fd906dd1 100644 --- a/src/runtime/CL/functions/CLGEMMConvolutionLayer.cpp +++ b/src/runtime/CL/functions/CLGEMMConvolutionLayer.cpp @@ -391,7 +391,7 @@ Status CLGEMMConvolutionLayer::validate(const ITensorInfo *input, const ITensorI const Size2D kernel_dims(kernel_width, kernel_height); // Output tensor auto initialization if not yet initialized - TensorShape expected_output_shape = compute_im2col_conv_shape(input, kernel_dims, conv_info, append_bias, dilation, true); + TensorShape expected_output_shape = compute_im2col_conv_shape(input, kernel_dims, conv_info, append_bias, dilation, true /* num_groups == 1, num_groups */); auto_init_if_empty(im2col_reshaped_info, input->clone()->set_tensor_shape(expected_output_shape)); |