From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: [dm-devel] [PATCH v4 01/11] block: make generic_make_request handle arbitrarily sized bios Date: Wed, 27 May 2015 10:20:18 +0200 Message-ID: <20150527082018.GB25575@lst.de> References: <1432318723-18829-1-git-send-email-mlin@kernel.org> <1432318723-18829-2-git-send-email-mlin@kernel.org> <20150526143626.GA4315@redhat.com> <20150526153414.GA31182@agk-dp.fab.redhat.com> <20150527090640.44c346e2@notabene.brown> <20150527004022.GD31182@agk-dp.fab.redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <20150527004022.GD31182@agk-dp.fab.redhat.com> Sender: linux-kernel-owner@vger.kernel.org To: device-mapper development , Alasdair G Kergon , Mike Snitzer , Ming Lei , Al@redhat.com, Christoph Hellwig , Lars Ellenberg , Oleg@redhat.com, Philip Kelleher , Joshua Morris , Christoph Hellwig , Kent@redhat.com, Overstreet , Nitin Gupta , Drokin , Viro , Dave Chinner , Ming Lin , Jens Axboe , Andreas Dilger , Geoff Levand , Jiri Kosina , lkml , Jim Paris , Minchan Kim , Dongsu Park List-Id: dm-devel.ids On Wed, May 27, 2015 at 01:40:22AM +0100, Alasdair G Kergon wrote: > It does sometimes make a significant difference to device-mapper stacks. > DM only uses it for performance reasons - it can already split bios when > it needs to. I tried to remove merge_bvec_fn from DM several years ago but > couldn't because of the adverse performance impact of lots of splitting activity. Does it still? Since the move to immutable biovecs the bio splits are pretty cheap now, but I'd really like to see this verified by benchmarks. From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Date: Wed, 27 May 2015 10:20:18 +0200 From: Christoph Hellwig To: device-mapper development , Alasdair G Kergon , Mike Snitzer , Ming Lei , Al@redhat.com, Christoph Hellwig , Lars Ellenberg , Oleg@redhat.com, Philip Kelleher , Joshua Morris , Christoph Hellwig , Kent@redhat.com, Overstreet , Nitin Gupta , Drokin , Viro , Dave Chinner , Ming Lin , Jens Axboe , Andreas Dilger , Geoff Levand , Jiri Kosina , lkml , Jim Paris , Minchan Kim , Dongsu Park , drbd-user@lists.linbit.com Message-ID: <20150527082018.GB25575@lst.de> References: <1432318723-18829-1-git-send-email-mlin@kernel.org> <1432318723-18829-2-git-send-email-mlin@kernel.org> <20150526143626.GA4315@redhat.com> <20150526153414.GA31182@agk-dp.fab.redhat.com> <20150527090640.44c346e2@notabene.brown> <20150527004022.GD31182@agk-dp.fab.redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20150527004022.GD31182@agk-dp.fab.redhat.com> Subject: Re: [Drbd-dev] [dm-devel] [PATCH v4 01/11] block: make generic_make_request handle arbitrarily sized bios List-Id: "*Coordination* of development, patches, contributions -- *Questions* \(even to developers\) go to drbd-user, please." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Wed, May 27, 2015 at 01:40:22AM +0100, Alasdair G Kergon wrote: > It does sometimes make a significant difference to device-mapper stacks. > DM only uses it for performance reasons - it can already split bios when > it needs to. I tried to remove merge_bvec_fn from DM several years ago but > couldn't because of the adverse performance impact of lots of splitting activity. Does it still? Since the move to immutable biovecs the bio splits are pretty cheap now, but I'd really like to see this verified by benchmarks. From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752703AbbE0IUX (ORCPT ); Wed, 27 May 2015 04:20:23 -0400 Received: from verein.lst.de ([213.95.11.211]:50826 "EHLO newverein.lst.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752232AbbE0IUU (ORCPT ); Wed, 27 May 2015 04:20:20 -0400 Date: Wed, 27 May 2015 10:20:18 +0200 From: Christoph Hellwig To: device-mapper development , Alasdair G Kergon , Mike Snitzer , Ming Lei , Al@redhat.com, Christoph Hellwig , Lars Ellenberg , Oleg@redhat.com, Philip Kelleher , Joshua Morris , Christoph Hellwig , Kent@redhat.com, Overstreet , Nitin Gupta , Drokin , Viro , Dave Chinner , Ming Lin , Jens Axboe , Andreas Dilger , Geoff Levand , Jiri Kosina , lkml , Jim Paris , Minchan Kim , Dongsu Park , drbd-user@lists.linbit.com Subject: Re: [dm-devel] [PATCH v4 01/11] block: make generic_make_request handle arbitrarily sized bios Message-ID: <20150527082018.GB25575@lst.de> References: <1432318723-18829-1-git-send-email-mlin@kernel.org> <1432318723-18829-2-git-send-email-mlin@kernel.org> <20150526143626.GA4315@redhat.com> <20150526153414.GA31182@agk-dp.fab.redhat.com> <20150527090640.44c346e2@notabene.brown> <20150527004022.GD31182@agk-dp.fab.redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20150527004022.GD31182@agk-dp.fab.redhat.com> User-Agent: Mutt/1.5.17 (2007-11-01) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, May 27, 2015 at 01:40:22AM +0100, Alasdair G Kergon wrote: > It does sometimes make a significant difference to device-mapper stacks. > DM only uses it for performance reasons - it can already split bios when > it needs to. I tried to remove merge_bvec_fn from DM several years ago but > couldn't because of the adverse performance impact of lots of splitting activity. Does it still? Since the move to immutable biovecs the bio splits are pretty cheap now, but I'd really like to see this verified by benchmarks.