From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Martin K. Petersen" Subject: Re: BLKZEROOUT + pread should return zeroes, right? Date: Tue, 14 Oct 2014 21:32:25 -0400 Message-ID: References: <20141014030132.GA12013@birch.djwong.org> <20141014042711.GJ5267@dastard> <20141014060242.GA22878@birch.djwong.org> <20141014063210.GK9738@thunk.org> <20141015012534.GB12013@birch.djwong.org> Mime-Version: 1.0 Content-Type: text/plain Cc: "Theodore Ts'o" , Dave Chinner , Jens Axboe , "Martin K. Petersen" , linux-fsdevel@vger.kernel.org, linux-ext4 To: "Darrick J. Wong" Return-path: In-Reply-To: <20141015012534.GB12013@birch.djwong.org> (Darrick J. Wong's message of "Tue, 14 Oct 2014 18:25:34 -0700") Sender: linux-fsdevel-owner@vger.kernel.org List-Id: linux-ext4.vger.kernel.org >>>>> "Darrick" == Darrick J Wong writes: Darrick> Well then let's change BLKZEROOUT to require O_DIRECT instead Darrick> of hiding the coherency problem, That would break existing users, though. Darrick> A further optimization to mke2fs would be to detect that we've Darrick> run discard-with-zeroes and therefore can skip issuing Darrick> subsequent zeroouts on the same ranges, but I'm wary that Darrick> discard-zeroes-data does what it purports to do. It's dubious. I'm working on making sure we only set discard_zeroes_data when the device guarantees it for 3.19. -- Martin K. Petersen Oracle Linux Engineering