diff options
-rw-r--r-- | src/core/CL/gemm/CLGEMMHelpers.cpp | 15 | ||||
-rw-r--r-- | src/runtime/NEON/functions/NEWinogradConvolutionLayer.cpp | 8 |
2 files changed, 17 insertions, 6 deletions
diff --git a/src/core/CL/gemm/CLGEMMHelpers.cpp b/src/core/CL/gemm/CLGEMMHelpers.cpp index 5abe224a3f..61aa962198 100644 --- a/src/core/CL/gemm/CLGEMMHelpers.cpp +++ b/src/core/CL/gemm/CLGEMMHelpers.cpp @@ -73,10 +73,17 @@ void update_padding_for_cl_image(ITensorInfo *tensor) constexpr unsigned int num_floats_per_pixel = 4; const unsigned int stride_y_in_elements = tensor->strides_in_bytes()[1] / tensor->element_size(); - const unsigned int pixel_aligment = get_cl_image_pitch_alignment(CLKernelLibrary::get().get_device()); - const unsigned int row_pitch_alignment = pixel_aligment * num_floats_per_pixel; - const unsigned int round_up_width = ((stride_y_in_elements + row_pitch_alignment - 1) / row_pitch_alignment) * row_pitch_alignment; - const unsigned int padding = round_up_width - stride_y_in_elements; + const unsigned int pixel_alignment = get_cl_image_pitch_alignment(CLKernelLibrary::get().get_device()); + + ARM_COMPUTE_ERROR_ON_MSG(pixel_alignment == 0, "Cannot retrieve cl_image pitch alignment"); + if(pixel_alignment == 0) + { + return; + } + + const unsigned int row_pitch_alignment = pixel_alignment * num_floats_per_pixel; + const unsigned int round_up_width = ((stride_y_in_elements + row_pitch_alignment - 1) / row_pitch_alignment) * row_pitch_alignment; + const unsigned int padding = round_up_width - stride_y_in_elements; tensor->extend_padding(PaddingSize(0, padding, 0, 0)); } diff --git a/src/runtime/NEON/functions/NEWinogradConvolutionLayer.cpp b/src/runtime/NEON/functions/NEWinogradConvolutionLayer.cpp index 21a4b7973a..dc3bbbe562 100644 --- a/src/runtime/NEON/functions/NEWinogradConvolutionLayer.cpp +++ b/src/runtime/NEON/functions/NEWinogradConvolutionLayer.cpp @@ -336,8 +336,8 @@ void NEWinogradConvolutionLayer::configure(const ITensor *input, const ITensor * _output = output; _is_prepared = false; - int n_gemms = 0; - int N_BLOCK = 0; // Size of block used by GEMM. + int n_gemms = 1; + int N_BLOCK = 1; // Size of block used by GEMM. std::unique_ptr<INEWinogradLayerTransformInputKernel> transform_input_kernel; std::unique_ptr<INEWinogradLayerTransformWeightsKernel> transform_weights_kernel; @@ -452,6 +452,10 @@ void NEWinogradConvolutionLayer::configure(const ITensor *input, const ITensor * } } #endif // __ARM_FEATURE_FP16_VECTOR_ARITHMETIC + else + { + ARM_COMPUTE_ERROR("Not supported."); + } const PaddingType use_padding_type = (conv_info.pad_top() != 0u || conv_info.pad_left() != 0) ? PADDING_SAME : PADDING_VALID; const bool use_same_padding = use_padding_type == PADDING_SAME; |