Bug#630031: linux-image-2.6.39-2-686-pae: Kernel boots, but dmesg contains kernel bug trace

[Jonathan Nieder <jrnieder@gmail.com>]

Hi again,

Frank McCormick wrote:

>    I don't have either of those kernels now. And because I work at home I
> won't have the time to mess around with multiple boots for a while.
>
> Why are you interested? Are you having the same problem or what ?

No, I can't reproduce it myself.  A few reasons:

 - hearing that someone is "stuck" using a kernel with known security
   problems does not leave me very happy.  There could be more such
   people, and not addressing relevant problems (in package
   descriptions or functionality) when they happen would damage the
   reputation of Debian and health of the internet;

 - I occasionally run into kernel bugs, and I would like not to be
   contributing to a trend in which more bugs are reported than the
   kernel team is able to promptly deal with:

	http://qa.debian.org/data/bts/graphs/l/linux-2.6.png

   So it seems worthwhile to understand existing bugs and fix them
   or pass them upstream as appropriate.

 - I like Debian users.

 - The thought of a kernel/hardware combination being buggy in this
   way triggers an aesthetic dislike in me.  I would be happier
   hacking on a platform that behaves reasonably and predictably.

So yes, the kernel producing symptoms of memory corruption when and
only when PAE is enabled seems interesting.  Hope that helps, and
sorry we didn't get to this earlier.

Jonathan