problems with dev rpm
Karl DeBisschop
kdebisschop at alert.infoplease.com
Mon Dec 22 20:49:16 UTC 2003
On Mon, 2003-12-22 at 13:39, Stan Bubrouski wrote:
> On Mon, 2003-12-22 at 12:33, Paul Nasrat wrote:
> > On Mon, Dec 22, 2003 at 12:32:30PM -0500, Karl DeBisschop wrote:
> > > On Mon, 2003-12-22 at 12:17, Paul Nasrat wrote:
> > > > On Mon, Dec 22, 2003 at 11:49:39AM -0500, Karl DeBisschop wrote:
> > > > > The dev RPM in the development tree (dev-3.3.9-2.i386.rpm) seems to very
> > > > > reliably crash my test machine. Is anyone else seeing this sort of
> > > > > behavior?
> > > >
> > > > can you run rpm -Uvvvh from a console.
> > >
> > > 'console' meaning over a serial line? Not yet.
> >
> > No just as not in X
> >
> > > Currently, the trace pushes everything else off the screen, and since
> > > there's no way to get the machine to respond, I have not yet been able
> > > to get more data.
> > >
> > > Is the serial console the only way I can get this data?
> >
> > Umm I did it from a text console tty0 (alt-F1), it didn't hang the box that time.
> >
> > The oops I got is attached.
> >
> > If you've bugzilla'd it please post the bug #, else please do as I'll be afk for a bit.
> >
> > Paul
>
> From the looks of the OOPs someone isn't giving up a lock before trying
> to take on another.
In the bugzilla I note that /dev/compaq is the last device created -
maybe the problem is in one of the 2 devices there.
Or maybe /dev/compaq is just the last thing that made it out the screen
buffer. The following items are:
/dev/compaq
/dev/compaq/cpqphpc
/dev/compaq/cpqrid
/dev/console
/dev/core
/dev/cpu
/dev/cpu/0
/dev/cpu/0/cpuid
/dev/cpu/0/microcode
/dev/cpu/0/msr
But from looking at timestamps in the dev tree, it really does look like
the problem is in /dev/compaq
--
Karl DeBisschop <kdebisschop at alert.infoplease.com>
More information about the fedora-devel-list
mailing list