Re: Oops on UHCI module unload (2.4.2-pre1)

Keith Owens (kaos@ocs.com.au)
Tue, 06 Feb 2001 20:04:24 +1100


On Tue, 6 Feb 2001 05:40:40 -0300,
John R Lenton <john@grulic.org.ar> wrote:
>When trying to figure out how to get USB to work (it was the MPS
>setting, more in other post) I got a repeatable Oops (is it an
>oops? it doesn't say "Oops!" like I thought they do).

The kernel makes you guess what the error messages are, to add some
spice to life :-)

>Attached are two ksymoops outputs, for the two times I did this.

Error (expand_objects): cannot stat(/lib/modules/2.4.2-pre1/kernel/drivers/usb/uhci.o) for uhci
Error (expand_objects): cannot stat(/lib/modules/2.4.2-pre1/kernel/drivers/usb/usbcore.o) for usbcore
Error (expand_objects): cannot stat(/lib/modules/2.4.2-pre1/kernel/drivers/input/mousedev.o) for mousedev
Error (expand_objects): cannot stat(/lib/modules/2.4.2-pre1/kernel/drivers/input/input.o) for input
Feb 4 04:32:20 burocracia kernel: Unable to handle kernel paging request at virtual address d10ff9e8

ksymoops could not find the modules you loaded. Did you rename, delete
or overwrite /lib/modules/2.4.2-pre1 between the oops and the time you
ran ksymoops?

Error (expand_objects): cannot stat(/lib/modules/2.4.2-pre1/kernel/drivers/usb/uhci.o) for uhci
Error (expand_objects): cannot stat(/lib/modules/2.4.2-pre1/kernel/drivers/usb/usbcore.o) for usbcore
Warning (expand_objects): object /lib/modules/2.4.2-pre1/kernel/drivers/net/ppp_deflate.o for module ppp_deflate has changed since load
Warning (expand_objects): object /lib/modules/2.4.2-pre1/kernel/drivers/net/bsd_comp.o for module bsd_comp has changed since load
Warning (expand_objects): object /lib/modules/2.4.2-pre1/kernel/drivers/net/ppp_async.o for module ppp_async has changed since load
Warning (expand_objects): object /lib/modules/2.4.2-pre1/kernel/drivers/net/ppp_generic.o for module ppp_generic has changed since load
Warning (expand_objects): object /lib/modules/2.4.2-pre1/kernel/drivers/net/slhc.o for module slhc has changed since load

ksymoops could could not find some modules, those it could find have
changed since the oops. I would treat all your traces with suspicion,
the base data is very unreliable. Try to reproduce with current
modules, instead of modules that have changed since the oops.

BTW, do not specify -O unless you have a good reason to ignore objects,
which you don't. You should supply a system map with -m instead of
suppressing it with -M.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/