Re: [Kgdb-bugreport] KGDB/RT integration woes

[Wessel, J. <jas...@wi...>]

=20

> -----Original Message-----
> From: Sergei Shtylyov [mailto:ssh...@ru...]=20
> Sent: Tuesday, June 26, 2007 12:47 PM
> To: Wessel, Jason
> Cc: kgd...@li...
> Subject: Re: [Kgdb-bugreport] KGDB/RT integration woes
>=20
> Hello.
>=20
> Wessel, Jason wrote:
> > Interesting...=20
>=20
> > I would have thought the generic fix I put in recently with=20
> respect to=20
> > restoring the preempt count after a bad memory read would=20
> have solved=20
> > this unless there is a another case that is not obvious with the rt=20
> > patch.
>=20
>     No. I'm getting the last mentioned trace with just KGDB=20
> patchset atop of 2.6.21, no -rt patch.
>=20

I tried with x86_64 + CONFIG_PREEMPT=3Dy + SMP + + kgdboe + kgdbwait.  I
set a breakpoint at do_mount() and did a back trace each time for the 5
file systems I mount and each continue worked fine.

I can only figure that the kernel is configured differently and that we
probably have different hardware.

It is hard to further instrument to know what the problem is, but you
can probably extrapolate if the problem is related to the preempt count
issue by adding a printk to print the preempt count on entry and exit
the kgdb_handle_exception routine, or in the place where the BUG() is
called.

Jason.