From mboxrd@z Thu Jan 1 00:00:00 1970 From: NeilBrown Subject: Re: [PATCH 0/5] misc cleanups for RAID5 Date: Thu, 23 Jun 2011 10:55:54 +1000 Message-ID: <20110623105554.19c5fe09@notabene.brown> References: <1308718230-2536-1-git-send-email-namhyung@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <1308718230-2536-1-git-send-email-namhyung@gmail.com> Sender: linux-raid-owner@vger.kernel.org To: Namhyung Kim Cc: linux-raid@vger.kernel.org List-Id: linux-raid.ids On Wed, 22 Jun 2011 13:50:25 +0900 Namhyung Kim wrote: > Hello, > > These are assorted cleanup patches for RAID5 code. > Please take a look. Any comments are welcomed. > > Thanks. > > > Namhyung Kim (5): > md/raid5: use kmem_cache_zalloc() > md/raid5: factor out dev_need_read() > md/raid5: factor out dev_need_for_write() > md/raid5: use r5_for_each_bio() > md/raid5: get rid of duplicated call to bio_data_dir() > > drivers/md/raid5.c | 146 ++++++++++++++++++++++++++-------------------------- > 1 files changed, 73 insertions(+), 73 deletions(-) > Hi, thanks for these. I have applied the first and the last. The two "factor out" patches conflict with some much more substantial refactoring I have been doing in raid5.c. I have just pushed all of that into my for-next branch: git://neil.brown.name/md for-next so you can use that as a basis for any further review. The r5_for_each_bio() patch I'm not 100% sure I'm happy with, and in any case it would have conflicted with my other changes too. I'm not fond of macros that hide details that could be important. A "for_each" macro that purely and simply walks through a list is fine. A "for_each" macro that does anything more complicated I start to have doubts about... However if you really do like it and want to rebase it on the for-next branch I'll have another look and think harder about it. Maybe I'll end up liking it after all, but no promises. Thanks, NeilBrown