Date: Thu, 19 Jun 1997 10:17:01 +0400 From: "Alexander V. Tischenko" <flash@intech.hway.ru> To: "Alexander V. Tischenko" <flash@hway.ru>, "Louis A. Mamakos" <louie@TransSys.COM> Cc: "David S. Miller" <davem@jenolan.rutgers.edu>, <freebsd-hackers@FreeBSD.ORG>, <Eric.Schenk@dna.lth.se> Subject: Re: RFCs and Urgent pointers Message-ID: <199706190620.KAA28211@thorin.hway.ru>
next in thread | raw e-mail | index | archive | help
> From: Louis A. Mamakos <louie@TransSys.COM> > To: Alexander V. Tischenko <flash@hway.ru> > Cc: David S. Miller <davem@jenolan.rutgers.edu>; freebsd-hackers@FreeBSD.ORG; Eric.Schenk@dna.lth.se > Subject: Re: RFCs and Urgent pointers > Date: 19 èþíÿ 1997 ã. 7:31 > > > On Wed, 18 Jun 1997, David S. Miller wrote: > > > > > Date: Wed, 18 Jun 1997 16:50:13 +0400 (MSD) > > > From: "Alexander V. Tischenko" <flash@hway.ru> > > > > > > Anybody thought of adding the RFC style Urgent pointers to the TCP, > > > say, as TCP level socket option ? > > > > > > We've made this a sysctl() tunable under Linux, I don't think we > > > considered the benefits of making it a socket option, that may in fact > > > be a better approach. Comments? > > > > > I suppose it is better to make it an option, 'cause this way you can set > > it on per-socket basis from your applications. > > TCP urgent data is how the socket out-of-band-data abstraction is realized. I > don't understand what else you might "add" to TCP to do "Urgent Pointers". > There ought to already be a option for "inline" out-of-band data, which I > think is the default in most modern BSD-based TCPs. Just read RFC-1122, especially the part concerning TCP and Urgent Pointers (4.2.2.4 Urgent Pointer), you will understand what i meant. > > louie > > ---------------------------------------------------------------------------- -------------------------------------------- Alexander V. Tischenko Integrated Network Technologies AT55-RIPE
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199706190620.KAA28211>