|
|
Sets through bytes the amount of dynamic shared memory that will be available to each thread block when the kernel given by hfunc is launched.
- Parameters:
-
| hfunc | - Kernel to specify dynamic shared-memory size for |
| bytes | - Dynamic shared-memory size per thread in bytes |
- Returns:
- CUDA_SUCCESS, CUDA_ERROR_DEINITIALIZED, CUDA_ERROR_NOT_INITIALIZED, CUDA_ERROR_INVALID_CONTEXT, CUDA_ERROR_INVALID_HANDLE, CUDA_ERROR_INVALID_VALUE
- Note:
- Note that this function may also return error codes from previous, asynchronous launches.
- See also:
- cuFuncSetBlockShape, cuFuncSetCacheConfig, cuFuncGetAttribute, cuParamSetSize, cuParamSeti, cuParamSetf, cuParamSetv, cuLaunch, cuLaunchGrid, cuLaunchGridAsync
|
