From mboxrd@z Thu Jan 1 00:00:00 1970 From: Neil Brown Subject: Re: [PATCH 1/3] md: ping userspace on 'write-pending' events Date: Fri, 2 May 2008 12:10:20 +1000 Message-ID: <18458.30732.453556.480932@notabene.brown> References: <20080430011826.20468.93265.stgit@dwillia2-linux.ch.intel.com> <20080430011933.20468.45086.stgit@dwillia2-linux.ch.intel.com> <18457.19136.168378.816525@notabene.brown> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: message from Dan Williams on Thursday May 1 Sender: linux-raid-owner@vger.kernel.org To: Dan Williams Cc: linux-raid@vger.kernel.org List-Id: linux-raid.ids On Thursday May 1, dan.j.williams@intel.com wrote: > > [ thinking out loud here to make sure we are on the same page ] > > I had to convince myself that it is still holding off writes while > MD_CHANGE_DEVS is set. And yes, MD_CHANGE_PENDING is not cleared > until after MD_CHANGE_DEVS is handled. So, the mddev->flags==0 test > was indeed more restrictive than it needed to be. Adding the > unqualified wait_event() to md_allow_write() also makes sense since it > is needed in the 'external' case and should be a nop in the > '!external' case since those bits are cleared in md_update_sb(). > > Simple test confirms this is working as expected. Good, thanks. NeilBrown