Date: Tue, 2 Sep 2008 11:32:43 -0700 From: Andrew Thompson <thompsa@FreeBSD.org> To: Maksim Yevmenkin <maksim.yevmenkin@gmail.com> Cc: current@freebsd.org Subject: Re: m_uiotombuf alignment Message-ID: <20080902183243.GC12367@citylink.fud.org.nz> In-Reply-To: <bb4a86c70809021054rf131f7dh7aa9c5bba90ac3b0@mail.gmail.com> References: <20080902174540.GB12367@citylink.fud.org.nz> <bb4a86c70809021054rf131f7dh7aa9c5bba90ac3b0@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Sep 02, 2008 at 10:54:08AM -0700, Maksim Yevmenkin wrote: > Andrew, > > > I have a patch here to removing the alignment of the align parameter. I > > can not see why it was added as it up to the caller to specify this, it > > breaks tap(4) on strict alignment machines as m_uiotombuf is called with > > ETHER_ALIGN. Also 'align' isnt a great description of this field, its > > more a padding or data offset. > > hmm... strange... from cvs > > === > > Revision 1.53 > Wed May 4 18:55:02 2005 UTC (3 years, 4 months ago) by emax > Branches: MAIN > > Change m_uiotombuf so it will accept offset at which data should be copied > to the mbuf. Offset cannot exceed MHLEN bytes. This is currently used to > fix Ethernet header alignment problem on alpha and sparc64. Also change all > users of m_uiotombuf to pass proper offset. > > Reviewed by: jmg, sam > Tested by: Sten Spans "sten AT blinkenlights DOT nl" > MFC after: 1 week > > === > > could you please explain how and on which platforms it breaks tap(4)? That revision had the correct behaviour, it was broken in r1.169 Rewrite m_uiotombuf() to use m_getm2() for mbuf allocation and do the uiomove() in a tight loop over the mbuf chain. Add a flags parameter to accept mbuf flags to be passed to m_getm2(). Adjust all callers for the extra parameter. Sponsored by: TCP/IP Optimization Fundraise 2005 Andrew
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20080902183243.GC12367>