I am searching for information about cuGetExportTable or cudaGetExportTable. In particular, I want to know what functions are are included in the table. A previous post (cudaGetExportTable a total hack) mentions that " I concluded that the export table is a list of function pointers into internal driver functions". If this post is correct, there are functions that are not included in the CUDA driver API. So there is a way to perform allocations, frees, memcpy to the GPU without using cuMemAlloc etc? Or the functions included in the export table are some kind of GPU management functions?
cuGetExportTable is undocumented by NVIDIA.
Using an undocumented function means you are subject to various uncertainties, as well as the possibility of behavioral change from one CUDA version to the next, with no formal notice.
The cuHook sample code may be of interest.