I wonder if we should establish some sort of policy around this, because I've seen different maintainers take different positions. (And if checking for allocation failure isn't useful, I don't really see why these patches should be accepted regardless; better not to clutter the code I'd think.)
There can of course be good reasons to check for allocation failures - it depends on the situation. Further, if new code is written that includes checks then I don't care and that could to some extent be deemed personal choice. What we probably should avoid is MRs that add unnecessary checks to existing code - we all have better things to spend our time on. I did accept these MRs because I didn't really want to get into a debate about this... -- https://gitlab.winehq.org/wine/wine/-/merge_requests/3846#note_45550