Hello! On Tue, Aug 18, 2009 at 01:28:13PM -0700, Roland McGrath wrote: > > do we need to add this check? Or is the glibc code already supposed to > > handle that correctly? Roland? > > pthread_kill should short-circuit for signal 0 after it's validated the thread. > That usage mode is not actually kosher any more as a thread-validity check > like using kill that way has always been a kosher PID-liveness and > permission-to-signal check, since pthread_kill is not required to > gracefully diagnose a bogus pthread_t. But the degenerate use on > pthread_self () is always going to work, so I guess that configure check is OK. Understood. Samuel now pushed the same short-circuiting code snippet that I had in my tree (and forgot :-( to push at that time...). I also enhanced libpthread's test-16.c with respect to that -- does that look alright? <http://git.savannah.gnu.org/cgit/hurd/libpthread.git/log/?h=master-test> Regards, Thomas
Attachment:
signature.asc
Description: Digital signature