Renato Golin
Renato Golin
> Truth be told, I never thought about padding and always assumed nice users passing blocks that perfectly divide the iteration domain. > 4. This brings me to my next...
> rather than its current behaviour. I could easily imagine a "stronger"/"enhanced" version that would call `snmalloc` instead of returning -1. I'm not an expert, but here are some considerations......
IIUC, the idea behind the `__builtin_instrumented_object_size` case would be to provide an actual simple implementation that returns a reasonable answer without having to instrument the whole code. Sure, sanitizers can...
After having some of our changes going through IREE (pack/unpack tensor ops), I am sceptical of this process, but I'm not against it. There may not be a better one,...
GH Bad: * GitHub confuses discussion comments with review comments in the code view. * You can only answer multiple comments in the code view. * Old discussion view keep...
While I'm unhappy with github PRs, @llvm-beanz point is more important than my personal preference. While I'm also unhappy with Phabricator, the reverse is also true. To me, the best...
I would suggest that all series have to be managed in the developers repos, and only one of their branches to be requested against the official repo's main branch. So...
> @joker-eph Yeah, we could end up with hundreds of branches in the main repo. I guess we could agree that branches that are merged should be deleted, and, if...