From: Andreas Dilger Subject: Re: [PATCH 15/23] jbd2: Change disk layout for metadata checksumming Date: Sat, 28 Apr 2012 16:58:12 -0600 Message-ID: References: <20120306204750.1663.96751.stgit@elm3b70.beaverton.ibm.com> <20120306204941.1663.56283.stgit@elm3b70.beaverton.ibm.com> <20120428141933.GB29481@thunk.org> Mime-Version: 1.0 (1.0) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8BIT Cc: "Darrick J. Wong" , Andreas Dilger , Sunil Mushran , Martin K Petersen , Greg Freemyer , Amir Goldstein , linux-kernel , Andi Kleen , Mingming Cao , Joel Becker , linux-fsdevel , "linux-ext4@vger.kernel.org" , Coly Li To: "Ted Ts'o" Return-path: Received: from mail130c7-2520.megamailservers.com ([69.49.98.25]:43331 "EHLO mail130c7.megamailservers.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1752068Ab2D1W6n convert rfc822-to-8bit (ORCPT ); Sat, 28 Apr 2012 18:58:43 -0400 In-Reply-To: <20120428141933.GB29481@thunk.org> Sender: linux-ext4-owner@vger.kernel.org List-ID: On 2012-04-28, at 8:19, Ted Ts'o wrote: > On Tue, Mar 06, 2012 at 12:49:41PM -0800, Darrick J. Wong wrote: >> @@ -177,11 +189,17 @@ typedef struct journal_block_tag_s >> __be32 t_blocknr; /* The on-disk block number */ >> __be32 t_flags; /* See below */ >> __be32 t_blocknr_high; /* most-significant high 32bits. */ >> + __be32 t_checksum; /* crc32c(uuid+seq+block) */ >> } journal_block_tag_t; >> >> #define JBD2_TAG_SIZE32 (offsetof(journal_block_tag_t, t_blocknr_high)) >> #define JBD2_TAG_SIZE64 (sizeof(journal_block_tag_t)) > > There's a problem with this patch here --- we are changing the size of > journal_block_tag_t, which is an on-disk data structure. So for > 64-bit journals, this represents a format change. This means that if > you have a 64-bit file system that needs to have its journal > recovered, if the journal was written with an older kernel, and then > we try to recover it with a new kernel, things won't be good. > Similarly, for e2fsck's recovery code, it's not going to be able to > recover 64-bit file systems using current coding, since this patch > series changes the size of JBD2_TAG_SIZE64. > > What we need to do is something like this: > > #define JBD2_TAG_SIZE64 (offsetof(journal_block_tag_t, t_checksum)) > #define JBD2_TAG_SIZE_CSUM (sizeof(journal_block_tag_t)) > > And then change the code appropriately in e2fsprogs and in the kernel > to use the correct tag size depending on the journal options. I thought we originally discussed using the high 16 bits of the t_flags field to store the checksum? This would avoid the need to change the disk format. Since there is still a whole transaction checksum, it isn't so critical that the per-block checksum be strong. One idea is to do the crc32c for each block, then store the high 16 bits into t_flags, and checksum the full 32-bit per-block checksums to make the commit block checksum, to avoid having to do the block checksums twice. Cheers, Andreas