Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 4 Jul 2005 10:29:53 +0900
From:      Pyun YongHyeon <pyunyh@gmail.com>
To:        Mike Silbersack <silby@silby.com>
Cc:        cvs-src@FreeBSD.org, src-committers@FreeBSD.org, cvs-all@FreeBSD.org, Andrew Thompson <thompsa@FreeBSD.org>
Subject:   Re: cvs commit: src/sys/net bridge.c
Message-ID:  <20050704012953.GA896@rndsoft.co.kr>
In-Reply-To: <20050703184013.U21939@odysseus.silby.com>
References:  <200507031824.j63IO3Bs009536@repoman.freebsd.org> <20050703184013.U21939@odysseus.silby.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, Jul 03, 2005 at 06:41:36PM -0500, Mike Silbersack wrote:
 > 
 > On Sun, 3 Jul 2005, Andrew Thompson wrote:
 > 
 > >thompsa     2005-07-03 18:24:03 UTC
 > >
 > > FreeBSD src repository
 > >
 > > Modified files:
 > >   sys/net              bridge.c
 > > Log:
 > > Check the alignment of the IP header before passing the packet up to the
 > > packet filter. This would cause a panic on architectures that require 
 > > strict
 > > alignment such as sparc64, ia64 and ppc.
 > 
 > What's the route that a packet takes before it gets to the bridge?  If it 
 > goes through ether_input, it would make sense to put the alignment there 
 > when necessary.
 > 
Agreed.

 > By the way, which network drivers are not aligning to the ip header right 
 > now?
 > 

FYI: em(4) produces non-aligned ip header when it is configured to
     use JUMBO frames. This is the reason why em(4) does not work
     on strict alignment architectures. In addition em(4) should
     remove its use of inl/outl in order to work on sparc64/ppc etc.
-- 
Regards,
Pyun YongHyeon



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050704012953.GA896>