Sebastian Hagedorn wrote: > -- Ken Murchison <murch@xxxxxxxxxxxxxx> is rumored to have mumbled on > 19. November 2007 13:17:07 -0500 regarding Re: One more attempt: stuck > processes: > >> The only other potential downside >>> the patch has is that stracing or gdb'ing it causes the timeout to >>> trigger prematurely. AFAIK that's a common issue for signals. I'm not >>> sure if there's something that could be done about that, i.e. catching >>> ERESTART or something like that? >> >> Is this only a problem with pop3s, or does it also effect the STLS >> command? > > Yes, it does: > > telnet cyrus 110 > ... > stls > +OK Begin TLS negotiation now > -ERR [SYS/PERM] Starttls failed > > That's when I straced the process ... If you want to write and test a patch to resolve this, I'll certainly take a look at it. Another way to resolve the blocking SSL_accept() problem would be to make the socket non-blocking and then check the SSL_accept() return code to see if we need to call SSL_accept() again. I took the easy way out by setting SO_RCVTIMEO. -- Kenneth Murchison Systems Programmer Project Cyrus Developer/Maintainer Carnegie Mellon University ---- Cyrus Home Page: http://cyrusimap.web.cmu.edu/ Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html
