From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jens Axboe Subject: Re: [PATCH 3/4] drop vmerge accounting Date: Fri, 22 Aug 2008 11:17:40 +0200 Message-ID: <20080822091740.GB20055@kernel.dk> References: <20080808055405.GF20055@kernel.dk> <20080808151147J.fujita.tomonori@lab.ntt.co.jp> <20080808061537.GI20055@kernel.dk> <20080808152845N.fujita.tomonori@lab.ntt.co.jp> <20080808063527.GJ20055@kernel.dk> <20080815094800.GI20055@kernel.dk> <20080822091044.GA20055@kernel.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from [93.163.65.50] ([93.163.65.50]:27756 "EHLO kernel.dk" rhost-flags-FAIL-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1751500AbYHVJRn (ORCPT ); Fri, 22 Aug 2008 05:17:43 -0400 Content-Disposition: inline In-Reply-To: <20080822091044.GA20055@kernel.dk> Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: Mikulas Patocka Cc: FUJITA Tomonori , adel.gadllah@gmail.com, matthew@wil.cx, linux-scsi@vger.kernel.org, James.Bottomley@hansenpartnership.com, dan.j.williams@intel.com, pjones@redhat.com, viro@zeniv.linux.org.uk, dougg@torque.net, drzeus-mmc@drzeus.cx On Fri, Aug 22 2008, Jens Axboe wrote: > On Fri, Aug 15 2008, Mikulas Patocka wrote: > > > I have applied this and the previous, with a fixup to raid5 to overload > > > both active stripe and process count in bi_phys_segments. > > > > > > -- > > > Jens Axboe > > > > Hi > > > > So here are two new patches. Patch 3 drops blk_queue_max_hw_segments > > function and max_hw_segments entry. I checked all the drivers, most of > > them set blk_queue_max_hw_segments equal to blk_queue_max_phys_segments, > > for the few ones that don't (for example scsi_lib.c), I set minimum of the > > two values with blk_queue_max_phys_segments. > > > > The next patch 4 drops the hw_segments logic from MMC card subsystem. > > Thanks, I've queued both up for review and inclusion for 2.6.28. Spoke too soon, they don't really apply to the for-2.6.28 branch of the block git tree that has your two previous patches. Can you please recheck and resend against that? -- Jens Axboe