Jan Sikorski
Jan Sikorski
Hi, I figured this would be a good place to post this. For more memory reusing one could use a lower level interface bridging between accelerate and cuda. Something along...
On NOMMU architectures _map_remove() and _map_add() both take process' lock to protect it's list of map_entries. Currently it will cause deadlock e.g. when killing the process.
When allocating descriptors from the pool, we should only request as much as needed for the variable number of descriptors, if applicable. It looks slightly dicey to use the fixed...
Works in some cases.