From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ethan Sommer Subject: Re: [Fwd: [ANNOUNCE] Layer-7 Filter for Linux QoS] Date: Tue, 20 May 2003 00:07:33 -0500 Sender: netdev-bounce@oss.sgi.com Message-ID: <3EC9B815.4000504@ethanet.com> References: <1053313298.3909.5.camel@rth.ninka.net> <20030519202756.I39498@shell.cyberus.ca> Mime-Version: 1.0 Content-Type: text/plain; format=flowed; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: "David S. Miller" , linux-net@vger.kernel.org, netdev@oss.sgi.com Return-path: In-reply-to: <20030519202756.I39498@shell.cyberus.ca> To: Jamal Hadi Errors-to: netdev-bounce@oss.sgi.com List-Id: netdev.vger.kernel.org Jamal Hadi wrote: >Hi, > ... >BTW, i am not sure how efficient the stuff that Henry spencer >wrote is in comparison to newer research on variants of boyer-moore >for regex searches. comment? > > I picked Spencer's mainly because it was a pain to port the glibc version into kernel space. They are both basically the posix regexp interface, so if someone wants to get another version working in kernel space we can test and see which one performs better with real load pretty easily. >The code also does seem inefficient but thats beside the point at the >moment (ex you seem to malloc for every incoming packet so you can do a >regex). > > Yep. We haven't spent a lot of time on optimizations. Obviously that example can be fixed pretty quickly... except I'm not sure we can avoid it and stay thread safe? (linux can route multiple packets at the same time on an smp box right? so we can't just use a staically defined buffer...) Ethan Sommer http://l7-filter.sf.net