Date: Thu, 12 Sep 2002 10:25:59 +0300 From: Maxim Sobolev <sobomax@FreeBSD.org> To: Munechika Sumikawa <sumikawa@ebina.hitachi.co.jp> Cc: brooks@one-eyed-alien.net, cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/net if_gif.c if_gif.h Message-ID: <3D804187.512BE49C@FreeBSD.org> References: <200209051535.g85FZdq2038989@freefall.freebsd.org> <20020905094452.A3044@Odin.AC.HMC.Edu> <20020912.154452.47469593.sumikawa@ebina.hitachi.co.jp>
next in thread | previous in thread | raw e-mail | index | archive | help
Munechika Sumikawa wrote: > > > On Thu, Sep 05, 2002 at 08:35:39AM -0700, Maxim Sobolev wrote: > > > sobomax 2002/09/05 08:35:39 PDT > > > > > > Modified files: > > > sys/net if_gif.c if_gif.h > > > Log: > > > Make recursion prevention variable per-instance and remove XXX comment > > > about thread-unsafety. > > > > Good solution. > > > > On minor issue. This changes the meaning of max_gif_nesting. Before it > > was the number of gif over gif tunnels you could nest. Now it's > > the number of times you can traverse the same tunnel while routing > > a packet. I'm not sure it makes any sense for it to be tunable any more. > > 1. it's still thread unsafe when two different threads use same gif > tunnel. Ok, good point - I'll add /* XXX */ back. > 2. As Brooks said, when several tunnel are nested for example gif0 -> > gif1 -> gif2 -> gif0, it's not counted rightly. Why not? The packet will be dropped down after the second attempt to enter gif0. To me it makes more sense than dropping it down in gif1. -Maxim To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3D804187.512BE49C>