Implementations requiring the definitions of multiple CUDA entity proxy classes, and which regard modules.
More...
Go to the source code of this file.
|
| cuda |
| Definitions and functionality wrapping CUDA APIs.
|
|
Implementations requiring the definitions of multiple CUDA entity proxy classes, and which regard modules.
Specifically:
- Functions in the
cuda::module
namespace.
- Methods of cuda::module_t and possibly some relates classes.
- The
context_t::create_module()
methods; see issue #320 on the issue tracker.
◆ load_from_file() [1/2]
module_t cuda::module::load_from_file |
( |
const device_t & |
device, |
|
|
const char * |
path |
|
) |
| |
|
inline |
- Parameters
-
device | The device in whose primary context to create a module containing the loaded data |
◆ load_from_file() [2/2]
module_t cuda::module::load_from_file |
( |
const char * |
path | ) |
|
|
inline |
- Note
- loads the module into the primary context of the device of the current context (although not necessarily in the current context!)