You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Beginning with CUDA 12.4, we are able to "enumerate" the functions in a module, or more literally - get an array of all cuFunction's for all compiled functions in a module. These can then be wrapped into cuda::kernel_t's queried for names, attributes etc.
Relevant API function is:
CUresult cuModuleEnumerateFunctions ( CUfunction* functions, unsigned int numFunctions, CUmodule mod );
We also need to think about whether we want a container of cuda::kernel_t's, or a container of cuFunction's, or both.
The text was updated successfully, but these errors were encountered:
Beginning with CUDA 12.4, we are able to "enumerate" the functions in a module, or more literally - get an array of all cuFunction's for all compiled functions in a module. These can then be wrapped into
cuda::kernel_t
's queried for names, attributes etc.Relevant API function is:
We also need to think about whether we want a container of
cuda::kernel_t
's, or a container ofcuFunction
's, or both.The text was updated successfully, but these errors were encountered: