Hello,
Michael Banck, le jeu. 05 févr. 2026 09:27:37 +0100, a ecrit:
> With the glibc patch from [2] applied, I have been running the above SQL
> for 5 million iterations without a hang just now. So those semi-frequent
> hangs on fruitcrow should be fixed now, yay!
Yay!
Thanks for taking the time to get the backtrace that explained it all :)
> > * I see special code paths for threads that were in (its notion of)
> > critical sections, which must be rare, but it looks like that just
> > leave it pending which seems reasonable
That's what was missing in cancellation checks. A very small window, but
still there so happened of course :)
Take the two most tricky things in Unix (signals and thread
cancellation) ; that was deemed to pose very tricky problems :)
Samuel
