Date: Tue, 20 Apr 2004 14:12:51 -0500 (CDT) From: Mike Silbersack <silby@silby.com> To: Nate Lawson <nate@root.org> Cc: cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/netinet tcp_subr.c tcp_var.h Message-ID: <20040420141059.Q25391@odysseus.silby.com> In-Reply-To: <20040420054638.E27872@root.org> References: <200404200633.i3K6XdXn067858@repoman.freebsd.org> <20040420032850.H20848@odysseus.silby.com> <20040420054638.E27872@root.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 20 Apr 2004, Nate Lawson wrote: > > I think that we may have to break away from standard RFC handling and > > change the TIME_WAIT code in tcp_input so that it will accept any SYN > > packet coming in without regard to the sequence number, forcing the > > TIME_WAIT socket to be recycled. > > It's been a while since I looked at all the RFCs, but can the window scale > option be taken into account for this? I'm thinking that if you receive a > packet while in TIME_WAIT with the proper window scale + sequence, accept > it, otherwise discard. As for initial sequences, make them less dependent > on port/address combos. Not sure if this will solve your problem. > > -Nate I don't see how the window scale option would change the situation at all. Making ISNs less dependent on the port/address pair isn't a solution either, their dependence on port/address is RFC1948's best property! In the face of high speed networks, we may just have to drop the check completely. Mike "Silby" Silbersack
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040420141059.Q25391>