Date: Fri, 16 Sep 2005 16:37:10 +0400 From: Gleb Smirnoff <glebius@FreeBSD.org> To: Yar Tikhiy <yar@comp.chem.msu.su> Cc: cvs-src@FreeBSD.org, src-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/net if_vlan.c Message-ID: <20050916123710.GX53891@cell.sick.ru> In-Reply-To: <20050916123421.GB23109@comp.chem.msu.su> References: <200509161144.j8GBihiM082730@repoman.freebsd.org> <20050916115312.GT53891@cell.sick.ru> <20050916123421.GB23109@comp.chem.msu.su>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, Sep 16, 2005 at 04:34:21PM +0400, Yar Tikhiy wrote: Y> On Fri, Sep 16, 2005 at 03:53:12PM +0400, Gleb Smirnoff wrote: Y> > On Fri, Sep 16, 2005 at 11:44:43AM +0000, Yar Tikhiy wrote: Y> > Y> yar 2005-09-16 11:44:43 UTC Y> > Y> Y> > Y> FreeBSD src repository Y> > Y> Y> > Y> Modified files: Y> > Y> sys/net if_vlan.c Y> > Y> Log: Y> > Y> Test the new M_VLANTAG packet flag before calling Y> > Y> m_tag_locate(). This adds little overhead of a simple Y> > Y> bitwise operation in case hardware VLAN acceleration Y> > Y> is on, yet saves the more expensive function call if Y> > Y> the acceleration is off. Y> > Y> Y> > Y> Reviewed by: ru, glebius Y> > Y> X-MFC-after: 6.0 Y> > ^^^^^^^^^^^^^^^^^^^ Y> > Why? Y> Y> 'Cause it's more about code consistency than stability, or performance, Y> or features. So I'd rather not bother re@ with this when they are Y> being flooded with really important bugfixes. It shouldn't be the Y> extra m_tag_locate() call that is likely to be a bottleneck in 6.0 ;) Not a bottleneck, but a significant processing addon for every inbound packet. If you don't mind I will request MFC from re@ after some testing. -- Totus tuus, Glebius. GLEBIUS-RIPN GLEB-RIPE
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050916123710.GX53891>