Luke Small <[email protected]> schreef op 28 december 2016 09:03:43 CET: >You have to make the socket call nonblocking > >On Wed, Dec 28, 2016, 01:11 Otto Moerbeek <[email protected]> wrote: > >> On Wed, Dec 28, 2016 at 03:10:13AM +0000, Luke Small wrote: >> >> > Thanks for the help. I'm not trying to kill the process. I stated >that >> the >> > process didn't die to show that it didn't reach the "got here2" >because >> it >> > was blocking. >> > >> > On Tue, Dec 27, 2016 at 6:40 PM Jeremie Courreges-Anglas ><[email protected] >> > >> > wrote: >> > >> > > Luke Small <[email protected]> writes: >> > > >> > > > It says "got here1", doesn't say "got here2", and the process >doesn't >> > > die. >> > > > Since it isn't reading anything from sockfd which is >"listen"ing, I >> > > expect >> > > > it to fail. Is it too much to ask it to fail? >> > > >> > > SOCK_NONBLOCK doesn't do what you think it does. The manpage >says >> under >> > > which conditions an accept() call doesn't block. >> > > >> > > -- >> > > jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE >1524 >> E7EE >> > > >> >> Read the man page again, the SOCK_NONBLOCK flags does not apply to >the >> accepting socket, but the new fd that is created on a succceful >accept. >> >> -Otto >>
I do not have to do anything. But you should work on your bug reporting skills. -Otto
