On Mon, Apr 17, 2006 at 07:15:32AM -0400, Theodore Ts'o wrote:
> On Mon, Apr 17, 2006 at 12:30:23PM +0200, Keld J?rn Simonsen wrote:
> > open("/proc/apm", O_RDONLY) = 4
> ...
> > read(4, <unfinished ...>
>
> This was caused by e2fsck trying to read from /proc/apm to see whether
> or not your system was running on batteries or not. /proc/apm exists
> (or the open would have returned an error), but reading from it
> apparently causes a kernel oops. This is definitely a kernel bug, and
> I suspect can be reproduced by the shell command "cat /proc/apm".
>
> Recompiling the kernel with CONFIG_APM disabled is probably the most
> expedient answer, since for most systems ACPI is more functional (and
> in some cases, required). Indeed, the APM code has been sufferring
> progressive bitrot, which probably explains the kernel oops. You
> could try sending a complaint to LKML if you really need APM
> functionality for your laptop, and for some reason ACPI is not
> sufficent for your needs.
My problem here has vanished, I don't know why.
But why was e2fsck checking APM?
None of the other fs fsck's do, AFAIK.
Best regards
Keld
_______________________________________________
Ext3-users@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/ext3-users
