--hK8Uo4Yp55NZU70L
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Mon, Mar 07, 2005 at 08:29:49PM +0100, Vincent wrote:
> Bill Studenmund a ?crit :
>=20
> >All I can think is to add printf()s after that message and keep trying.=
=20
> >You should be able to find out which call is crashing (since all the=20
> >subsequent printf()s won't print), then figure out what is being passed=
=20
> >in.
>=20
> Good theoretical idea, bad practical one. I lose at least 3 minutes each=
=20
> time in FSCK'ing the disks... But if there is no other way...

Mount everything read-only, or as much as you can. Think of the system as=
=20
being in a full-featured single-user mode. :-)

Other folks have noted that if it's really userland that's causing the
problem, you just need to find the offending program. If it's the kernel,
you shouldn't have to worry about the file systems.

> Well, I guess it is somehow related to Acpi. The bug is, I couldn't=20
> pinpoint what has been changed in Acpi between 99.12 and 99.13.

It's irritating, but a binary date-based search should work too. :-|

Take care,

Bill

--hK8Uo4Yp55NZU70L
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (NetBSD)

iD8DBQFCLPCdWz+3JHUci9cRAot3AJsENcXNMvipL4NkqNEaeHqL5cy7BACdGxGL
HfcqkqS5+3aNCrX5VCyaf2s=
=UqIq
-----END PGP SIGNATURE-----

--hK8Uo4Yp55NZU70L--