Junio C Hamano <gitster@xxxxxxxxx> writes: > Sorry, but you still have lost me---I do not see if/why we even care > about atexit codepath. As far as the end users are concered, they > are running "git" and observing the exit code from "git". There, > reporting that "git" was killed by SIGPIPE, instead of exiting > normally, is not something they want to hear about after quitting > their pager, and that is why the signal reception codepath matters. (something I noticed that I left unsaid...) On the other hand, "git" spawns not just pager but other subprocesses (e.g. "hooks"), and it is entirely up to us what to do with the exit code from them. When we care about making an external effect (e.g. post-$action hooks that are run for their side effects), we can ignore their exit status just fine. And I do not see why the "we waited before leaving, and noticed the pager exited with non-zero status" that we could notice in the atexit codepath has to be so special. We _could_ (modulo the "exit there is not portable" you noted) make our exit status reflect that, but I do not think it is all that important a "failure" (as opposed to, say, we tried to show a commit message but failed to recode it into utf-8, or we tried to spawn the pager but failed to start a process) to clobber _our_ exit status with pager's exit status. So...
