On 04/03, Eric W. Biederman wrote: > > Oleg Nesterov <[email protected]> writes: > > > I meant that may_hang == 0 implies zap_other_threads(do_count => -1) which > > should > > return the number of threads which didn't pass exit_notify(). The returned > > value > > can be wrong unless you change exit_notify() to set exit_state under > > siglock.
but I forgot to add that, of course, this problem is very minor because we can only miss a thread which is already at the end of exit_notify() so nothing bad can happen. But imo should be fixed anyway, simply because this looks wrong/racy. Your recent 4/5 has the same problem. > Interesting an existing bug. Hmm... what do you mean? The current code looks fine. Oleg.
