From owner-cvs-all@FreeBSD.ORG Thu Dec 9 15:51:01 2004 Return-Path: Delivered-To: cvs-all@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0DC3E16A4CE; Thu, 9 Dec 2004 15:51:01 +0000 (GMT) Received: from fledge.watson.org (fledge.watson.org [204.156.12.50]) by mx1.FreeBSD.org (Postfix) with ESMTP id B120E43D60; Thu, 9 Dec 2004 15:51:00 +0000 (GMT) (envelope-from robert@fledge.watson.org) Received: from fledge.watson.org (localhost [127.0.0.1]) by fledge.watson.org (8.13.1/8.13.1) with ESMTP id iB9FmVWZ055968; Thu, 9 Dec 2004 10:48:31 -0500 (EST) (envelope-from robert@fledge.watson.org) Received: from localhost (robert@localhost)iB9FmVOT055965; Thu, 9 Dec 2004 15:48:31 GMT (envelope-from robert@fledge.watson.org) Date: Thu, 9 Dec 2004 15:48:31 +0000 (GMT) From: Robert Watson X-Sender: robert@fledge.watson.org To: Andre Oppermann In-Reply-To: <41B8709C.329DD195@freebsd.org> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII cc: src-committers@freebsd.org cc: SUZUKI Shinsuke cc: cvs-src@freebsd.org cc: cvs-all@freebsd.org cc: glebius@freebsd.org cc: Ruslan Ermilov Subject: Re: cvs commit: src/sys/netinet in_gif.c X-BeenThere: cvs-all@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: CVS commit messages for the entire tree List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 09 Dec 2004 15:51:01 -0000 On Thu, 9 Dec 2004, Andre Oppermann wrote: > Ruslan Ermilov wrote: > > > > On Tue, Dec 07, 2004 at 11:18:15AM +0100, Andre Oppermann wrote: > > > I hope to remove the ability to have pointers into the routing table > > > before 6.0 release. > > > > > This would make my WIP OBE, which is nice. But how are you going to > > compensate for the performance loss of *not* having cached routes? > > Up to a couple of hundred kpps it doesn't matter and the only place > where we have cached routes is in the various IP[46] tunnels. The > caching makes locking of the routing table more complex and expensive. > Once that can be simplified there is a significant amount of lookup cost > going away. It is also my belief that the locking around the cached routes is currently incorrect, and with that locking added, there will also be a further hit. Robert N M Watson FreeBSD Core Team, TrustedBSD Projects robert@fledge.watson.org Principal Research Scientist, McAfee Research