fix pthread_detach inadvertently acting as cancellation point in race case

disabling cancellation around the pthread_join call seems to be the
safest and logically simplest fix. i believe it would also be possible
to just perform the unmap directly here after __tl_sync, removing the
dependency on pthread_join, but such an approach duplicately encodes a
lot more implementation assumptions.
This commit is contained in:
Rich Felker 2023-02-11 09:54:12 -05:00
parent 115149c023
commit c3cd04fa5f

View file

@ -5,8 +5,12 @@ static int __pthread_detach(pthread_t t)
{
/* If the cas fails, detach state is either already-detached
* or exiting/exited, and pthread_join will trap or cleanup. */
if (a_cas(&t->detach_state, DT_JOINABLE, DT_DETACHED) != DT_JOINABLE)
return __pthread_join(t, 0);
if (a_cas(&t->detach_state, DT_JOINABLE, DT_DETACHED) != DT_JOINABLE) {
int cs;
__pthread_setcancelstate(PTHREAD_CANCEL_DISABLE, &cs);
__pthread_join(t, 0);
__pthread_setcancelstate(cs, 0);
}
return 0;
}