From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jens Axboe Subject: Re: [PATCH 05/10] writeback: support > 1 flusher thread per bdi Date: Mon, 31 Aug 2009 19:32:24 +0200 Message-ID: <20090831173223.GY12579@kernel.dk> References: <1251720891-19793-1-git-send-email-jens.axboe@oracle.com> <1251720891-19793-6-git-send-email-jens.axboe@oracle.com> <20090831130125.GC20465@infradead.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, chris.mason@oracle.com, david@fromorbit.com, akpm@linux-foundation.org, jack@suse.cz To: Christoph Hellwig Return-path: Received: from brick.kernel.dk ([93.163.65.50]:60244 "EHLO kernel.dk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751765AbZHaRcW (ORCPT ); Mon, 31 Aug 2009 13:32:22 -0400 Content-Disposition: inline In-Reply-To: <20090831130125.GC20465@infradead.org> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Mon, Aug 31 2009, Christoph Hellwig wrote: > On Mon, Aug 31, 2009 at 02:14:46PM +0200, Jens Axboe wrote: > > Build on the bdi_writeback support by allowing registration of > > more than 1 flusher thread. File systems can call bdi_add_flusher_task(bdi) > > to add more flusher threads to the device. If they do so, they must also > > provide a super_operations function to return the suitable bdi_writeback > > struct from any given inode. > > Please drop this for now. It adds a lot of unused and untestede code > including a dead on arrival inode operation. OK, will drop the multi thread support, I'll retain the worker items though. > Once we are going to add actual multiple flusher threads we can revise > it. But I don't think for example that the operation in it's current > form makese too much sense - the most logical way to add multiple > flushers per filesystems would be having multiple bdis of some sort. Sure, it was more of a proof-of-concept, since I think it'll be important at higher buffered output rates with a single thread becoming a bottle neck. But I have no issues with dropping it for now, it'll be easy enough to re-introduce. I'll post an updated patch series later tonight/tomorrow with the changes. -- Jens Axboe