linux-kernel - Re: 2.6.29 regression: ATA bus errors on resume

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <4A441A69.1010803@gmail.com>
Date:	Fri, 26 Jun 2009 09:46:33 +0900
From:	Tejun Heo <htejun@...il.com>
To:	Niel Lambrechts <niel.lambrechts@...il.com>
CC:	Alan Cox <alan@...rguk.ukuu.org.uk>,
	"linux.kernel" <linux-kernel@...r.kernel.org>,
	Theodore Tso <tytso@....edu>
Subject: Re: 2.6.29 regression: ATA bus errors on resume

Hello, Niel.

Niel Lambrechts wrote:
> I've recently switched my root partition from OpenSUSE 11.1 to Fedora 11
> and since then I've not again seen the issue. I'm still using vanilla
> 2.6.30 generated with the same .config and EXT4 as before, so I have no
> idea why I cannot reproduce the issue. I still use hibernate + sleep
> frequently, and I just checked - I have 5 days uptime with a mount count
> of 20 and the file-system is still clean.
> 
> The one big difference is that my original partition was a EXT2 -> EXT3
> -> EXT4 upgrade job over a long period of time, and some of the EXT4
> parameters now used by Fedora 11 on the reformatted root partition are
> different from what I had then. Here is a summary of the differences in
> case it matters at all:
> 
> Current settings:
> Default mount options:    user_xattr acl
> Inodes per group:         8192
> Inode blocks per group:   512
> Flex block group size:    16
> Required extra isize:     28
> Desired extra isize:      28
> Default directory hash:   half_md4
> 
> Previous settings:
> Default mount options:    (none)
> Inodes per group:         8176
> Inode blocks per group:   511
> Default directory hash:   tea
> 
> If I do notice any such errors again I'll apply the debug patch and let
> you know, but it does seem as if the upgrade made this issue disappear...

Eh... if my analysis is correct the bug will be dependent on some
non-deterministic things including timing and layout of blocks of the
file which the filesystem touchs during suspend/resume.  I'm fairly
sure the problem is still there.  Oh well, it seems I'll have to
construct a synthetic case to test it.  Anyways, please let me know if
you encounter the problem again.  I'll keep you in loop with further
patches.

Thanks.

-- 
tejun
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/