| /external/tensorflow/tensorflow/core/util/ |
| D | cuda_device_functions.h | 486 __device__ detail::ToTypeIfConvertible<U, T> CudaAtomicAdd(T* ptr, U value) { in CudaAtomicAdd() function
490 __device__ inline Eigen::half CudaAtomicAdd(Eigen::half* ptr, in CudaAtomicAdd() function
498 __device__ inline double CudaAtomicAdd(double* ptr, double value) { in CudaAtomicAdd() function
505 __device__ inline double CudaAtomicAdd(double* ptr, double value) { in CudaAtomicAdd() function
520 __device__ inline std::complex<float> CudaAtomicAdd(std::complex<float>* ptr, in CudaAtomicAdd() function
523 return std::complex<float>(CudaAtomicAdd(ptr_scalar, value.real()), in CudaAtomicAdd()
524 CudaAtomicAdd(ptr_scalar + 1, value.imag())); in CudaAtomicAdd()
527 __device__ inline std::complex<double> CudaAtomicAdd( in CudaAtomicAdd() function
530 return std::complex<double>(CudaAtomicAdd(ptr_scalar, value.real()), in CudaAtomicAdd()
531 CudaAtomicAdd(ptr_scalar + 1, value.imag())); in CudaAtomicAdd()
[all …]
|
| D | cuda_kernel_helper_test.cu.cc | 100 CudaAtomicAdd(failure_count, 1); in CudaShuffleGetSrcLaneTest()
|
| /external/tensorflow/tensorflow/core/kernels/ |
| D | resize_bilinear_op_gpu.cu.cc | 120 CudaAtomicAdd(output_grad + in ResizeBilinearGradKernel()
126 CudaAtomicAdd(output_grad + in ResizeBilinearGradKernel()
134 CudaAtomicAdd(output_grad + in ResizeBilinearGradKernel()
140 CudaAtomicAdd(output_grad + in ResizeBilinearGradKernel()
231 CudaAtomicAdd(output_grad + in LegacyResizeBilinearGradKernel()
237 CudaAtomicAdd(output_grad + in LegacyResizeBilinearGradKernel()
245 CudaAtomicAdd(output_grad + in LegacyResizeBilinearGradKernel()
251 CudaAtomicAdd(output_grad + in LegacyResizeBilinearGradKernel()
|
| D | crop_and_resize_op_gpu.cu.cc | 188 CudaAtomicAdd(grads_image_ptr + in CropAndResizeBackpropImageKernel()
194 CudaAtomicAdd(grads_image_ptr + in CropAndResizeBackpropImageKernel()
202 CudaAtomicAdd(grads_image_ptr + in CropAndResizeBackpropImageKernel()
208 CudaAtomicAdd(grads_image_ptr + in CropAndResizeBackpropImageKernel()
217 CudaAtomicAdd(grads_image_ptr + in CropAndResizeBackpropImageKernel()
336 CudaAtomicAdd(grads_boxes_ptr + b * 4 + 0, dy1); in CropAndResizeBackpropBoxesKernel()
337 CudaAtomicAdd(grads_boxes_ptr + b * 4 + 1, dx1); in CropAndResizeBackpropBoxesKernel()
338 CudaAtomicAdd(grads_boxes_ptr + b * 4 + 2, dy2); in CropAndResizeBackpropBoxesKernel()
339 CudaAtomicAdd(grads_boxes_ptr + b * 4 + 3, dx2); in CropAndResizeBackpropBoxesKernel()
|
| D | bias_op_gpu.cu.cc | 106 CudaAtomicAdd(bias_backprop + bias_offset, ldg(output_backprop + index)); in BiasGradNHWC_Naive()
118 CudaAtomicAdd(bias_backprop + bias_offset, ldg(output_backprop + index)); in BiasGradNCHW_Naive()
138 CudaAtomicAdd(s_data + bias_offset, AccT(ldg(output_backprop + index))); in BiasGradNHWC_SharedAtomics()
143 CudaAtomicAdd(bias_backprop + index, T(s_data[index])); in BiasGradNHWC_SharedAtomics()
179 CudaAtomicAdd(s_data + bias_offset, sum); in BiasGradNCHW_SharedAtomics()
191 CudaAtomicAdd(bias_backprop + bias_index, T(data)); in BiasGradNCHW_SharedAtomics()
|
| D | scatter_nd_op_gpu.cu.cc | 47 CudaAtomicAdd(out, val); in operator ()()
66 CudaAtomicAdd(ptr, val.real()); in operator ()()
67 CudaAtomicAdd(ptr, val.imag()); in operator ()()
|
| D | sparse_tensor_dense_matmul_op_gpu.cu.cc | 50 CudaAtomicAdd(out_location, std::numeric_limits<T>::quiet_NaN()); in SparseTensorDenseMatMulKernel()
59 CudaAtomicAdd(out_location, a_value * b_value); in SparseTensorDenseMatMulKernel()
|
| D | segment_reduction_ops_gpu.cu.cc | 85 CudaAtomicAdd(output + output_index, sum); in SortedSegmentSumCustomKernel()
100 CudaAtomicAdd(output + output_index, sum); in SortedSegmentSumCustomKernel()
|
| D | resize_nearest_neighbor_op_gpu.cu.cc | 122 CudaAtomicAdd(bottom_diff_n + idx, ldg(top_diff + index)); in ResizeNearestNeighborBackwardNHWC()
151 CudaAtomicAdd(bottom_diff_n + idx, ldg(top_diff + index)); in LegacyResizeNearestNeighborBackwardNHWC()
|
| D | segment_reduction_ops.h | 89 CudaAtomicAdd(dest, value); in operator()
|
| D | dilation_ops_gpu.cu.cc | 120 CudaAtomicAdd( in DilationBackpropInputKernel()
169 CudaAtomicAdd( in DilationBackpropFilterKernel()
|
| D | scatter_functor_gpu.cu.h | 44 __device__ void operator()(T* dest, T src) const { CudaAtomicAdd(dest, src); }
|
| D | maxpooling_op_gpu.cu.cc | 214 CudaAtomicAdd(bottom_diff + n * height * width * channels + maxidx, in MaxPoolBackwardNoMaskNHWC()
247 CudaAtomicAdd(bottom_diff + offset + mask[index], top_diff[index]); in MaxPoolBackward()
|
| D | depthwise_conv_op_gpu.h | 1096 CudaAtomicAdd(addr, partial_sum);
1127 CudaAtomicAdd(addr, partial_sum);
1308 CudaAtomicAdd(filter_offset + filter, static_cast<T>(val));
1384 CudaAtomicAdd(addr, partial_sum);
1416 CudaAtomicAdd(addr, partial_sum);
1576 CudaAtomicAdd(filter_offset + filter, static_cast<T>(val));
|
| D | svd_op_gpu.cu.cc | 69 CudaAtomicAdd(V + batch, v); in ComputeValueOfVKernel()
|