On Mon, Jan 18, 2016 at 12:37:01PM +0200, Boris Astardzhiev wrote: > Hello, > > Sorry for the delay of my reply. As far as I understand pthread_testcancel() > is not necessary in the recvmmsg syscall since cancellation is not quite > common > among apps. But if there is cancellation attempts as long as I use > __sys_recvmsg() instead > of the interposing approach on a cancel attempt recvmmsg() will return > EINTR which will get > me out? Yes.
The corner case is the cancellation attempt (SIGCANCEL == SIGTHR) coming while the thread is executing code around the syscall. > > Secondly, I guess it's better to use __sys_sendmmsg() similarly instead of > the > insterposing table regarding sendmmsg(). Sure, sendmmsg and recvmmsg are same. > > Lastly, regarding the manpage - should I extend send/recv(2) for the new > calls or > create new manpage files? IMO it is more logical to extend the existing page than write a new one. _______________________________________________ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
