| /external/clang/test/SemaCUDA/ |
| D | cuda-builtin-vars.cu | 28 out[i++] = gridDim.x; in kernel()
29 gridDim.x = 0; // expected-error {{no setter defined for property 'x'}} in kernel()
30 out[i++] = gridDim.y; in kernel()
31 gridDim.y = 0; // expected-error {{no setter defined for property 'y'}} in kernel()
32 out[i++] = gridDim.z; in kernel()
33 gridDim.z = 0; // expected-error {{no setter defined for property 'z'}} in kernel()
|
| /external/eigen/unsupported/Eigen/CXX11/src/Tensor/ |
| D | TensorReductionCuda.h | 116 const Index num_threads = blockDim.x * gridDim.x; in ReductionInitKernel()
130 if (gridDim.x == 1) { in FullReductionKernel()
158 eigen_assert(gridDim.x == 1 || *semaphore >= 2u); in FullReductionKernel()
178 if (gridDim.x > 1 && threadIdx.x == 0) { in FullReductionKernel()
180 atomicInc(semaphore, gridDim.x + 1); in FullReductionKernel()
193 eigen_assert(gridDim.x == 1); in ReductionInitFullReduxKernelHalfFloat()
206 const Index num_threads = blockDim.x * gridDim.x; in ReductionInitKernelHalfFloat()
226 if (gridDim.x == 1 && first_index == 0) { in FullReductionKernelHalfFloat()
256 if (gridDim.x == 1 && first_index == 0) { in FullReductionKernelHalfFloat()
380 eigen_assert(gridDim.y == 1);
[all …]
|
| D | TensorRandom.h | 25 gridDim.x * blockDim.x * (blockIdx.y * blockDim.y + threadIdx.y); in get_random_seed()
|
| D | TensorExecutor.h | 240 const Index step_size = blockDim.x * gridDim.x;
|
| D | TensorConvolution.h | 585 const int plane_stride = blockDim.y * gridDim.y;
639 const int plane_stride = blockDim.z * gridDim.z;
|
| /external/clang/test/CodeGenCUDA/ |
| D | cuda-builtin-vars.cu | 21 out[i++] = gridDim.x; // CHECK: call i32 @llvm.nvvm.read.ptx.sreg.nctaid.x() in kernel()
22 out[i++] = gridDim.y; // CHECK: call i32 @llvm.nvvm.read.ptx.sreg.nctaid.y() in kernel()
23 out[i++] = gridDim.z; // CHECK: call i32 @llvm.nvvm.read.ptx.sreg.nctaid.z() in kernel()
|
| /external/tensorflow/tensorflow/core/kernels/ |
| D | concat_lib_gpu_impl.cu.cc | 43 for (; gidx < total_cols; gidx += blockDim.x * gridDim.x) { in concat_fixed_kernel()
50 for (; gidy < total_rows; gidy += blockDim.y * gridDim.y) { in concat_fixed_kernel()
97 for (; gidx < total_cols; gidx += blockDim.x * gridDim.x) { in concat_variable_kernel()
109 for (; gidy < total_rows; gidy += blockDim.y * gridDim.y) in concat_variable_kernel()
|
| D | random_op_gpu.h | 136 const int32 total_thread_count = gridDim.x * blockDim.x;
175 const int32 total_thread_count = gridDim.x * blockDim.x;
|
| D | check_numerics_op_gpu.cu.cc | 41 const int32 total_thread_count = gridDim.x * blockDim.x; in CheckNumericsKernel()
|
| D | split_lib_gpu.cu.cc | 152 for (; gidx < total_cols; gidx += blockDim.x * gridDim.x) { in split_v_kernel()
164 for (; gidy < total_rows; gidy += blockDim.y * gridDim.y) in split_v_kernel()
|
| D | stateful_random_ops_gpu.cu.cc | 47 auto total_thread_count = gridDim.x * blockDim.x; in FillKernel()
|
| D | reduction_gpu_kernels.cu.h | 201 const int stride = blockDim.x * gridDim.x;
325 row += rows_per_warp * gridDim.y * blockDim.y;
326 for (; row < num_rows; row += rows_per_warp * gridDim.y * blockDim.y) {
354 out[col * gridDim.y + blockIdx.y] = s;
378 row += gridDim.y * blockDim.y;
381 for (; row < num_rows; row += gridDim.y * blockDim.y) {
411 out[col * gridDim.y + blockIdx.y] = s;
|
| D | relu_op_gpu.cu.cc | 42 const int32 total_device_threads = gridDim.x * blockDim.x; in ReluGradHalfKernel()
|
| D | parameterized_truncated_normal_op_gpu.cu.cc | 85 max_samples_per_item * (gridDim.x * blockDim.x); in TruncatedNormalKernel()
|
| D | bias_op_gpu.cu.cc | 136 index += blockDim.x * gridDim.x) { in BiasGradNHWC_SharedAtomics()
|
| D | depthwise_conv_op_gpu.h | 252 for (int b = blockIdx.x; b < in_blocks; b += gridDim.x) { in DepthwiseConv2dGPUKernelNHWCSmall()
539 for (int b = blockIdx.x; b < in_blocks; b += gridDim.x) { in DepthwiseConv2dGPUKernelNCHWSmall()
1247 for (int b = blockIdx.x; b < in_blocks; b += gridDim.x) {
1517 for (int b = blockIdx.x; b < in_blocks; b += gridDim.x) {
|
| D | conv_2d_gpu.h | 227 eigen_assert(gridDim.y == 1);
228 eigen_assert(gridDim.z == 1);
|
| /external/tensorflow/tensorflow/core/util/ |
| D | cuda_device_functions.h | 91 gridDim.x * blockDim.x, count); in CudaGridRangeX()
99 gridDim.y * blockDim.y, count); in CudaGridRangeY()
107 gridDim.z * blockDim.z, count); in CudaGridRangeZ()
387 assert(blockDim.x * gridDim.x / blockDim.x == gridDim.x); in SetZero()
398 assert(blockDim.x * gridDim.x / blockDim.x == gridDim.x); in SetToValue()
|
| /external/tensorflow/tensorflow/examples/adding_an_op/ |
| D | cuda_op_kernel.cu.cc | 23 i += blockDim.x * gridDim.x) { in AddOneKernel()
|
| /external/tensorflow/tensorflow/tools/ci_build/builds/user_ops/ |
| D | cuda_op_kernel.cu.cc | 23 i += blockDim.x * gridDim.x) { in AddOneKernel()
|
| /external/clang/lib/Headers/ |
| D | cuda_builtin_vars.h | 115 __CUDA_BUILTIN_VAR __cuda_builtin_gridDim_t gridDim; variable
|
| /external/tensorflow/tensorflow/contrib/mpi_collectives/ |
| D | ring.cu.cc | 91 i += blockDim.x * gridDim.x) { in elemwise_accum()
|
| /external/tensorflow/tensorflow/contrib/mpi_collectives/kernels/ |
| D | ring.cu.cc | 91 i += blockDim.x * gridDim.x) { in elemwise_accum()
|
| /external/skia/src/compute/hs/cuda/sm_35/ |
| D | hs_cuda_macros.h | 181 #define HS_GLOBAL_SIZE_X() (gridDim.x * blockDim.x)
|
| /external/skqp/src/compute/hs/cuda/sm_35/ |
| D | hs_cuda_macros.h | 181 #define HS_GLOBAL_SIZE_X() (gridDim.x * blockDim.x)
|