Mailing List Archive

Suddenly cannot open root partition
Hello:

I have the following issue: after a power outage, my B2000 rebooted. It
loaded the kernel, mounted the root partition read-only, but when it should
open it read-write I got the following:

checking root partition
fask.ext3: No such file or directory while trying to open /dev/sda4
the superblock cannot be read or does not describe.....blah blah
*Filesystem couldn't be fixed.

But the partition is good as I can check it, mount it, read and write on it
from an install CD.

dmesg shows the following:

break 0,0: pid=970 command='udevd'
Your system ate a SPARC! Gah! ....and the cow :-)
udevd (pid 970). Breakpoint (code 0)

The problem seems to be with udev. Any idea or workaround?

Thanks,
Mauro
--
gentoo-hppa@gentoo.org mailing list
Re: Suddenly cannot open root partition [ In reply to ]
Guy:

Kernel 2.6.15.1-pa4
udev version 103

I am using this kernel since more than a year ago so I do not think there
is a problem with its config. I am not 100% sure but I believe udev had
been updated a couple of days before and the problem occured after that.

There is nothing suspicious on the logs, only what I wrote in my previous
email.

I also noticed that it only can see the root partition (in read-only
mode of course), but /boot (/dev/sda4) and /mnt/disk2 (/dev/sdb1) both
shows as if they were empty (ls shows nothing inside them)

Thanks,
Mauro


On Fri, 26 Jan 2007, Guy Martin wrote:

>
> Hi Mauro,
>
> What kernel and what version of udev are you running ?
>
> Could you provide full kernel config and kernel boot logs ?
>
>
> Guy
>
>
> On Thu, 25 Jan 2007 21:46:25 -0300
> Mauro Maroni <mmaroni@fi.uba.ar> wrote:
>
>> Hello:
>>
>> I have the following issue: after a power outage, my B2000 rebooted. It
>> loaded the kernel, mounted the root partition read-only, but when it should
>> open it read-write I got the following:
>>
>> checking root partition
>> fask.ext3: No such file or directory while trying to open /dev/sda4
>> the superblock cannot be read or does not describe.....blah blah
>> *Filesystem couldn't be fixed.
>>
>> But the partition is good as I can check it, mount it, read and write on it
>> from an install CD.
>>
>> dmesg shows the following:
>>
>> break 0,0: pid=970 command='udevd'
>> Your system ate a SPARC! Gah! ....and the cow :-)
>> udevd (pid 970). Breakpoint (code 0)
>>
>> The problem seems to be with udev. Any idea or workaround?
>>
>> Thanks,
>> Mauro
>> --
>> gentoo-hppa@gentoo.org mailing list
>>
>
>
> --
> Guy Martin
> Gentoo Linux - HPPA port Lead
>
--
gentoo-hppa@gentoo.org mailing list
Re: Suddenly cannot open root partition [ In reply to ]
On Fri, Jan 26, 2007 at 07:30:56PM -0300, Mauro Maroni wrote:
> I am not 100% sure but I believe udev had
> been updated a couple of days before and the problem occured after that.
>
If you would like to confirm if udev did get upgraded, You can
boot from CD, enter chroot (refer to the gentoo handbook,
just don't do the fdisk or formats ;-) and emerge genlop. This is a
great tool that reports on the emerge activity. From there you can type
"genlop -l" and is will show you all the packages emerged and when.

If you feel (or confirm) that the problem is related to udev, you may
want to back down to the previous version you had installed. Hopefully
it's still in the tree. If it's not, this is a good time to add the the
--buildpkg option to EMERGE_DEFAULT_OPTS in /etc/make.conf, that way you
will have binary packages for each package you emerge going forward, and
can easily revert in case of emergency.

> I also noticed that it only can see the root partition (in read-only
> mode of course), but /boot (/dev/sda4) and /mnt/disk2 (/dev/sdb1) both
> shows as if they were empty (ls shows nothing inside them)

Can you see the other partitions when booting off the CD? Before you
enter the chroot environment above, make sure you mount all your
partitions as you did during the install. Hopefully you can see the data
and the other partitions. If not it sounds like you have may have some
drive corruption.

Good luck!

Brett
--
gentoo-hppa@gentoo.org mailing list
Re: Suddenly cannot open root partition [ In reply to ]
On Friday 26 January 2007 20:08, Brett Johnson wrote:
> On Fri, Jan 26, 2007 at 07:30:56PM -0300, Mauro Maroni wrote:
> > I am not 100% sure but I believe udev had
> > been updated a couple of days before and the problem occured after that.
>
> If you would like to confirm if udev did get upgraded,

Confirmed. Udev was upgraded last december from 0.8x to 1.0.3.
This was the first time the machine have rebooted since December.
And I also found similar "udevd breakpoint" messages on /var/log/messages some
minutes after that upgrade.

>
> If you feel (or confirm) that the problem is related to udev, you may
> want to back down to the previous version you had installed. Hopefully
> it's still in the tree.

Well, the problem now is that the CD I had does not detect the network card,
even modprobing the tulip module manually. So no card, no emerge...
It is an old CD (2004.2) so now I am downloading a newer image (2006.1).
I hope it works with it.

>
> > I also noticed that it only can see the root partition (in read-only
> > mode of course), but /boot (/dev/sda4) and /mnt/disk2 (/dev/sdb1) both
> > shows as if they were empty (ls shows nothing inside them)
>
> Can you see the other partitions when booting off the CD? Before you
> enter the chroot environment above, make sure you mount all your
> partitions as you did during the install. Hopefully you can see the data
> and the other partitions. If not it sounds like you have may have some
> drive corruption.

The partitions are OK.

>
> Good luck!

Thanks. I just found in Google that a Debian user had a similar problem and
had to upgrade the kernel to 2.6.17. I will try that once I get the
image....and the network card working :-)

> Brett

Regards,
Mauro
--
gentoo-hppa@gentoo.org mailing list
Re: Suddenly cannot open root partition (solved) [ In reply to ]
On Monday 29 January 2007 21:29, Mauro Maroni wrote:
> On Friday 26 January 2007 20:08, Brett Johnson wrote:
> > On Fri, Jan 26, 2007 at 07:30:56PM -0300, Mauro Maroni wrote:
> > > I am not 100% sure but I believe udev had
> > > been updated a couple of days before and the problem occured after
> > > that.
> >
> > If you would like to confirm if udev did get upgraded,
>
> Confirmed. Udev was upgraded last december from 0.8x to 1.0.3.
> This was the first time the machine have rebooted since December.
> And I also found similar "udevd breakpoint" messages on /var/log/messages
> some minutes after that upgrade.
>
> > If you feel (or confirm) that the problem is related to udev, you may
> > want to back down to the previous version you had installed. Hopefully
> > it's still in the tree.
>

> Thanks. I just found in Google that a Debian user had a similar problem and
> had to upgrade the kernel to 2.6.17. I will try that once I get the
> image....and the network card working :-)

I finally upgraded the kernel to 2.6.19-1 and the issue with udev dissapeared.

Regards,
Mauro
--
gentoo-hppa@gentoo.org mailing list