Date: Fri, 11 Apr 2003 13:14:14 EDT From: BelletJr@aol.com To: barney@pit.databus.com Cc: current@freebsd.org Subject: Re: connect(2) behavior with unreacheable hosts Message-ID: <113.21af9c71.2bc851e6@aol.com>
next in thread | raw e-mail | index | archive | help
>Dans un e-mail dat=E9 du 10/04/03 22:56:31 Paris, Madrid (Heure d'=E9t=E9),= =20 barney@pit.databus.com a =E9crit : > > >On Thu, Apr 10, 2003 at 03:45:54PM -0400, BelletJr@aol.com wrote: >> Why does not connect(2) return any error when trying to connect to a host= =20 >> unreachable because of an infinite loop in the routes? No time-out occurs= =20 and=20 >> the value 0 is returned by connect(2). >> My test was done with TCP/IPv4. > >By any chance, are you using UDP (SOCK_DGRAM) rather than TCP (SOCK_STREAM)= ? >That sets up data structures but sends no packets, so even if an icmp >time-exceeded will be returned when you do send packets, the problem >can't be detected yet. Thanks for your time. No, it happens well and truly with SOCK_STREAM. Jerome
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?113.21af9c71.2bc851e6>