linux-kernel - Storage regression in v3.6-rc3 was: Storage related regression in linux-next 20120824

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]

Date:	Wed, 29 Aug 2012 23:58:09 +0200
From:	Dieter Ries <clip3@....de>
To:	Jeff Garzik <jgarzik@...ox.com>
CC:	Arvydas Sidorenko <asido4@...il.com>, linux-kernel@...r.kernel.org,
	IDE/ATA development list <linux-ide@...r.kernel.org>,
	Dieter Ries <mail@...terries.net>
Subject: Storage regression in v3.6-rc3 was: Storage related regression in
 linux-next 20120824

Hi again,

On 08/27/2012 06:39 PM, Dieter Ries wrote:
> On 08/27/2012 04:40 PM, Jeff Garzik wrote:
>> On Mon, Aug 27, 2012 at 4:59 AM, Arvydas Sidorenko <asido4@...il.com> wrote:
>>> linux-next 20120824 introduced regression on Mac mini 2011 - /root partition
>>> gets mounted as read-only and it stays so even when trying to reboot into
>>> previously working kernel. In order to make it work on older kernel again
>>> `fsck` needs to be executed:
>>> ----------------------------------------------------------------------
>>> Root partition uses ext4 filesystem on Intel 330 SSD connected to SATA3.
>>> dmesg shows that journaling was disabled:
>>> ----------------------------------------------------------------------
>>> I haven't bisected to which exactly tree introduces the regression,
>>> but I can tell
>>> that linux-next 20120813 and v3.6-rc3 work fine.
> 
> I think I experienced this with v3.6-rc3 today as well. I thought my ssd
> (ocz vertex3) was broken first, but then I saw this mail.
> 
> Never happened before on 3.5.something.
> 
> I am trying to reproduce it right now, it happened 2 times before I read
> the mail, I didnt keep the dmesg. Now it's taking it's time to occurr
> again...

It happened again today, and I managed to get a dmesg this time. The
root partition was first, which results in the dmesg you can find
attached. Later the same happened to the home partition, but then I
wasn't able to save a dmesg anymore.

Just to stress this again, this mangles my root partition, resulting in
another forced reboot for fixing it again after a reboot. So I guess
it's possibly dangerous.

It never happened on other kernels, and with 3.6-rc3 it happened on 3
partitions now, as I copied the content of the root partition to
somewhere else on the ssd after it happened first.

>> Can you pastebin 'dmesg' and 'lspci'?  Did this occur only once, or is
>> it reproducible?

Both are attached. I cannot really bisect it, because I cannot trigger
it. It just happens at inconvenien moments...

One funny thing is: The lspci output was different between one boot
where nothing happened, and today after it happened agian. Only 2
revisions changed, but I attached both anyway, the one suffixed "Bad" is
after the problem occured.

>>      Jeff

Cheers,

Dieter

>> --
>> 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/
>>
> 
> 

View attachment "kernel36rc3faildmesg" of type "text/plain" (26825 bytes)

View attachment "kernel36rc3faillspciBad" of type "text/plain" (1869 bytes)

View attachment "kernel36rc3faillspci" of type "text/plain" (1851 bytes)