From: "Martin K. Petersen" Subject: Re: [PATCH] e2fsck: Discard free data and inode blocks. Date: Fri, 22 Oct 2010 17:01:03 -0400 Message-ID: References: <1287670556-23460-1-git-send-email-lczerner@redhat.com> <6388FD2D-50A8-42B9-A955-3824451ACBF4@dilger.ca> <4CC175E6.5000700@gmail.com> <4CC19BC2.9010503@gmail.com> <4CC1A3AA.6040004@gmail.com> <386E61B0-BF4D-4F96-9541-A614F63DE808@dilger.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Ric Wheeler , Lukas Czerner , linux-ext4@vger.kernel.org, tytso@mit.edu, sandeen@redhat.com To: Andreas Dilger Return-path: Received: from rcsinet10.oracle.com ([148.87.113.121]:31083 "EHLO rcsinet10.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755190Ab0JVVCL (ORCPT ); Fri, 22 Oct 2010 17:02:11 -0400 In-Reply-To: <386E61B0-BF4D-4F96-9541-A614F63DE808@dilger.ca> (Andreas Dilger's message of "Fri, 22 Oct 2010 11:50:40 -0600") Sender: linux-ext4-owner@vger.kernel.org List-ID: >>>>> "Andreas" == Andreas Dilger writes: Andreas> When this patch was first under discussion, I proposed that we Andreas> validate that the device is actually zeroed by doing a write a Andreas> non-zero block to the disk and then calling discard+zero for Andreas> that region, and reading back the block and verifying it. The problem is that the devices that get this wrong don't deterministically return zeroes. So they may return zeroes for the discarded test block but garbage for the stuff you actually depend on being cleared. -- Martin K. Petersen Oracle Linux Engineering