On 07/28, Roland McGrath wrote: > > > + * While in TASK_STOPPED, we can be considered safely > > + * stopped by utrace_do_stop(). Clear ->stopped if we > > + * were woken by signal. > > s/signal/SIGKILL/, no? It's not really "while in TASK_STOPPED" any more, > it's that utrace_do_stop could have changed TASK_STOPPED to TASK_TRACED. > Then the only case is that we wake up now because of SIGKILL.
Yes sure, but only after we change utrace_do_stop() to set TASK_TRACED. OK, I'll do s/signal/SIGKILL/ in the comment. Actually I was going to do this in 4/4 but forgot. > If utrace_wakeup woke us, it would have cleared ->stopped itself. Yes, that is why the comment says "woken by signal". Oleg.
