Home
last modified time | relevance | path

Searched refs:total_work_block (Results 1 – 3 of 3) sorted by relevance

/external/pytorch/aten/src/ATen/native/cuda/
DFunctionOfAMatrixUtilsKernel.cu17 constexpr int total_work_block = n_threads * n_elems_per_thread; in C10_LAUNCH_BOUNDS_2() local
18 int idx = total_work_block * blockIdx.x + threadIdx.x; in C10_LAUNCH_BOUNDS_2()
36 constexpr int total_work_block = n_threads * n_elems_per_thread; in _lauch_kernel() local
37 dim3 grid((total_n_elems + total_work_block - 1) / total_work_block); in _lauch_kernel()
DLinearAlgebra.cu78 constexpr int total_work_block = n_threads * n_elems_per_thread; in C10_LAUNCH_BOUNDS_2() local
79 int idx = total_work_block * blockIdx.x + threadIdx.x; in C10_LAUNCH_BOUNDS_2()
97 constexpr int total_work_block = n_threads * n_elems_per_thread; in _launch_kernel() local
98 dim3 grid((total_n_elems + total_work_block - 1) / total_work_block); in _launch_kernel()
DUnfoldBackwardKernel.cu25 constexpr int total_work_block = n_threads * n_elems_per_thread; in C10_LAUNCH_BOUNDS_2() local
26 int idx = total_work_block * blockIdx.x + threadIdx.x; in C10_LAUNCH_BOUNDS_2()
44 constexpr int total_work_block = n_threads * n_elems_per_thread; in _launch_unfold_backward_kernel() local
45 dim3 grid((total_n_elems + total_work_block - 1) / total_work_block); in _launch_unfold_backward_kernel()