From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752032Ab0JUPTz (ORCPT ); Thu, 21 Oct 2010 11:19:55 -0400 Received: from moutng.kundenserver.de ([212.227.126.171]:60355 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751783Ab0JUPTy (ORCPT ); Thu, 21 Oct 2010 11:19:54 -0400 From: Arnd Bergmann To: Jan Harkes Subject: Re: [RFC][PATCH 0/3] Coda: remove BKL Date: Thu, 21 Oct 2010 17:20:28 +0200 User-Agent: KMail/1.12.2 (Linux/2.6.35-16-generic; KDE/4.3.2; x86_64; ; ) Cc: linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org References: <1287606184-26889-1-git-send-email-jaharkes@cs.cmu.edu> <201010211108.41320.arnd@arndb.de> <4CC058EC.1060200@cs.cmu.edu> In-Reply-To: <4CC058EC.1060200@cs.cmu.edu> MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-15" Content-Transfer-Encoding: 7bit Message-Id: <201010211720.28775.arnd@arndb.de> X-Provags-ID: V02:K0:cPL7nvNMGY/hq8yXG8VoGVFTXuDoTODfTOI/oedzBFM CO/Ez93Uc1N4oGxjLGE98SQf1XJeC+puFx5HkpMMmVReF+kElb 5wVzE/Z2c1qD50h8kDL4BU/oPYD3usAiM8s3ZThEl9REIqwxI/ L9gTyjN+aQwKI16zyYZ2c1xrOLsZn1wb4tbdhl9vezsP6+rduO bsZjiJKNM4qKPewLkXl5w== Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thursday 21 October 2010, Jan Harkes wrote: > On 10/21/2010 05:08 AM, Arnd Bergmann wrote: > > On Wednesday 20 October 2010, Jan Harkes wrote: > >> The following three patches remove the big kernel lock from the Coda > >> file system kernel module. We initially add a spinlock to protect the > >> Coda-specific inode data. After this we can push the BKL all the way > >> down to where the requests are queued for the userspace Coda cache > >> manager process and finally we replace the remaining BKL uses with a > >> mutex that protects the upcall queues. > > > > Very nice! > > > > Are you pushing this to Linus or should I take care of it along with > > the other file systems? > > > > This patch is larger than most of the others, but I can add it to > > my bkl/vfs tree if you like. > > In that case it might be better when pushed separately, just to avoid > any chances on a merge conflict for the bkl/vfs tree. I was thinking of > letting the initial merges hit, rebase and retest over the weekend if > there were any Coda related changes, and sending the final version to > Linus by monday. > > Besides, that gives me a couple of more days to make sure I weed out any > remaining issues. Ok, good. I'll leave everything as is then, to avoid having to rebase myself. This means you will get the conflict against the pushdown patch into coda_fill_super, which is trivial to clean up by reverting part of my patch and applying yours. Arnd