NebulaStream
0.6.213
NebulaStream is a data and application management framework for the internet of things
|
#include <jitify.hpp>
Public Member Functions | |
ExecutionPolicy (Location location_=DEVICE, jitify::detail::vector< std::string > headers_=0, jitify::detail::vector< std::string > options_=0, file_callback_type file_callback_=0, cudaStream_t stream_=0, int device_=0, int block_size_=256, size_t cache_size_=JitCache::DEFAULT_CACHE_SIZE) | |
Public Attributes | |
Location | location |
std::vector< std::string > | headers |
std::vector< std::string > | options |
file_callback_type | file_callback |
cudaStream_t | stream |
int | device |
int | block_size |
size_t | cache_size |
Specifies location and parameters for execution of an algorithm.
stream | The CUDA stream on which to execute. |
headers | A vector of headers to include in the code. |
options | Options to pass to the NVRTC compiler. |
file_callback | See jitify::Program. |
block_size | The size of the CUDA thread block with which to execute. |
cache_size | The number of kernels to store in the cache before overwriting the least-recently-used ones. |
|
inline |
int jitify::ExecutionPolicy::block_size |
CUDA block size with which to execute.
Referenced by jitify::parallel_for().
size_t jitify::ExecutionPolicy::cache_size |
The number of instantiations to store in the cache before overwriting the least-recently-used ones.
Referenced by jitify::parallel_for().
int jitify::ExecutionPolicy::device |
CUDA device on which to execute.
Referenced by jitify::parallel_for().
file_callback_type jitify::ExecutionPolicy::file_callback |
Optional callback for loading source files.
Referenced by jitify::parallel_for().
std::vector<std::string> jitify::ExecutionPolicy::headers |
List of headers to include when compiling the algorithm.
Referenced by jitify::parallel_for().
Location jitify::ExecutionPolicy::location |
Location (HOST or DEVICE) on which to execute.
Referenced by jitify::parallel_for().
std::vector<std::string> jitify::ExecutionPolicy::options |
List of compiler options.
Referenced by jitify::parallel_for().
cudaStream_t jitify::ExecutionPolicy::stream |
CUDA stream on which to execute.
Referenced by jitify::parallel_for().