Am 21.02.26 um 4:12 AM schrieb Benjamin McGuire: > I’ve come up with a different patch where the wait happens in the child > process that runs the qm cleanup (With help from Cursor).
If I wanted to scrutinize LLM output, I could ask an LLM myself. You don't need to play middle man. >From a quick look, the generated patch completely ignores the part of the issue with cleanup being attempted twice. And no, this is not an invitation to ask the LLM again for such a patch. > This way, we avoid any Perl-side polling or sleeping while holding the > VM config lock, which was the way we did it in the v2 patch. Holding the config lock is desired in this case. It's better if no other operation should take the lock and interfere before cleanup is done. Please do not post such things to the mailing list. It just side-tracks the discussion, is a legal gray zone and a wasteful use of resources. Best Regards, Fiona
