For what it's worth, I seem to have found a way to
work around this bug.

Here are the details.

I have an MSI (Micro Star) MS-6712 (KT4V)
motherboard.

It was running version 2.6.32-35 of
package linux-image-2.6.32-5-686 in squeeze.

I wanted to upgrade its RAM from about half a
gigabyte on two sticks, to three gigabytes on three
sticks.

I bought three used and unmatched one gigabyte
memory sticks, pulled out the old small ones, and
put in the new.

After rebooting, this Debian bug (#582487)
happened.

Then I tried upgrading linux to the latest version
in stable (squeeze), 2.6.32-39.

That failed too.

Then I upgraded linux to the version in
debian-backports, 2.6.39+35.

Still no luck.

Then I pulled out the memory stick nearest to the
CPU.

That worked.

I rebooted twice more, and saw no more stack
traces in dmesg or /var/log/messages.

I suspect bad or mismatched RAM caused the error,
but I'm not sure.

~K



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111212060253.GA27665@loaner.com">http://lists.debian.org/20111212060253.GA27665@loaner.com

tags 582487 - moreinfo
merge 582088 582487
quit

Hi Kingsley,

Kingsley G. Morse Jr. wrote:

> For what it's worth, I seem to have found a way to
> work around this bug.

Having the same error message does not imply having the same bug.

The bug you're responding to looks like another episode in the
bug#582088 saga. Merging them.

Kind regards,
Jonathan



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111222110853.GE7180@elie.Belkin">http://lists.debian.org/20111222110853.GE7180@elie.Belkin

Hi Jonathan,

Happy holidays.

I checked out the bug#582088 saga that you
mentioned.

I happened to notice it has a different stack
trace than bug 582487.

If I asked you nicely, would you please elaborate
on why you think they're the same bug?

Thanks,
Kingsley

On 12/22/11 05:08, Jonathan Nieder wrote:
> tags 582487 - moreinfo
> merge 582088 582487
> quit
>
> Hi Kingsley,
>
> Kingsley G. Morse Jr. wrote:
>
> > For what it's worth, I seem to have found a way to
> > work around this bug.
>
> Having the same error message does not imply having the same bug.
>
> The bug you're responding to looks like another episode in the
> bug#582088 saga. Merging them.
>
> Kind regards,
> Jonathan




--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111222204436.GA19077@loaner.com">http://lists.debian.org/20111222204436.GA19077@loaner.com

Thank you for sharing your detailed thoughts.

I may owe you an apology for being too terse in my
bug report.

It seems to me that it may be important to
explain:

1.) I encountered a stack trace like the one
in bug#582487 after adding RAM, and then
pretty much every time I booted the computer.

My stack trace was consistent.

It looked like the one in bug#582487.

2.) I encountered a bug on a different machine
than the one(s) in bug#582487 and bug#582088.

I have an MSI (Micro Star) MS-6712 (KT4V)
motherboard, but the bug report you merged
(#582088) to says "Eee PC 1000HG, fuse".

At the moment, it seems to me that bug#582487 and
#582088 are different, and I humbly suggest
un-merging them.

Thanks,
Kingsley

On 12/22/11 14:49, Jonathan Nieder wrote:
> Kingsley G. Morse Jr. wrote:
>
> > I happened to notice it has a different stack
> > trace than bug 582487.
> >
> > If I asked you nicely, would you please elaborate
> > on why you think they're the same bug?
>
> Sure. It's because they both are examples of inconsistent state
> pointing to memory corruption, each only happened once, and both
> happened on the same machine.




--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111222212523.GB19077@loaner.com">http://lists.debian.org/20111222212523.GB19077@loaner.com

Kingsley G. Morse Jr. wrote:

> At the moment, it seems to me that bug#582487 and
> #582088 are different, and I humbly suggest
> un-merging them.

You didn't report bug#582487 in the first place!

Yes, you tripped the same assertion that Thiemo did. In your case, it
seems it was due to a hardware problem. If you really want to track
that in the BTS, please file a new bug.

Thanks for your interest, and hope that helps.

Regards,
Jonathan



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111222213241.GF8650@elie.Belkin">http://lists.debian.org/20111222213241.GF8650@elie.Belkin