Compiles kernel stored in code
and caches pointer to CUfunction
Type | Intent | Optional | Attributes | Name | ||
---|---|---|---|---|---|---|
type(MPI_Comm), | intent(in) | :: | comm |
MPI Communicator |
||
character(len=*), | intent(in) | :: | kernel_name |
Kernel name |
||
type(kernel_type_t), | intent(in) | :: | kernel_type |
Type of kernel to build |
||
type(dtfft_transpose_t), | intent(in) | :: | transpose_type |
Type of transposition to perform |
||
type(kernel_codegen), | intent(in) | :: | code |
Kernel code to compile |
||
type(device_props), | intent(in) | :: | props |
GPU architecture properties |
||
integer(kind=int64), | intent(in) | :: | base_storage |
Number of bytes needed to store single element |
||
integer(kind=int32), | intent(in) | :: | tile_size |
Tile size to use in shared memory |
||
integer(kind=int32), | intent(in) | :: | padding |
Padding to use in shared memory |
Compiled kernel to return