linux-ext4 - Re: xfstests-bld generic/231 failure on 4.18-rc4 and later

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20180726180146.6dhe326vm3g6zuax@localhost.localdomain>
Date:   Thu, 26 Jul 2018 14:01:46 -0400
From:   Eric Whitney <enwlinux@...il.com>
To:     Eric Whitney <enwlinux@...il.com>
Cc:     linux-ext4@...r.kernel.org, tytso@....edu
Subject: Re: xfstests-bld generic/231 failure on 4.18-rc4 and later

* Eric Whitney <enwlinux@...il.com>:
> As mentioned in today's ext4 concall, I've been seeing generic/231 failures
> when running kvm-xfstests on recent 4.18 kernels, beginning with rc4 and
> continuing through rc5 and rc6.  To get the failure, it's necessary to
> run generic/230 and 231 in sequence.  Running generic/231 alone and
> repeatedly does not reproduce the failure for me.
> 
> The test failure occurs because e2fsck reports problems at the end of 231:
> 
> 
> e2fsck 1.44-WIP (06-Sep-2017)
> Pass 1: Checking inodes, blocks, and sizes
> Inode 19 passes checks, but checksum does not match inode.  Fix? no
> 
> Inode 20 passes checks, but checksum does not match inode.  Fix? no
> 
> Inode 21 passes checks, but checksum does not match inode.  Fix? no
> 
> Inode 22 passes checks, but checksum does not match inode.  Fix? no
> 
> Inode 23 passes checks, but checksum does not match inode.  Fix? no
> 
> 
> The failure bisects to a patch in 4.18-rc4:
> (8844618d8aa7) ext4: only look at the bg_flags field if it is valid
> 
> I can no longer reproduce the failure after reverting this patch in a clean
> rc4 build.
> 

I forgot to note that this failure only occurs for me in the nojournal test
case - I've not seen it in any other test case over my rc4, rc5, or rc6
regression runs.

> Thanks,
> Eric