I apologize for all the repeats. It appears our mail server is
reflecting some E-mail, and we're looking into it.
Did you get the patch, Michael?
--Matt
"Matt D. Robinson" wrote:
>
> Michael Walfish wrote:
> >
> > Matt D. Robinson wrote:
> > > Let me know if you have any problems.
> >
> > Hi Matt,
> >
> > We're big fans of lkcd (so far it's been really easy to use and understand).
> > Any help you can provide is greatly appreciated.
> >
> > I applied the 2.4.2 patch this morning. Here are some observations:
> > (dmesg and relevant parts of our .config at the end of this mail)
> >
> > 1) no problems when I force a kernel crash inside a user process
> > 2) when I force a crash inside an interrupt (for a device driver), I get the
> > following, in order:
> > -->standard oops message
> > -->interesting message, below
> > -->another oops for the second CPU
> > -->system reset, presumably driven by the lkcd patch
>
> Hmmm, I haven't seen this. I can see from schedule() as to what
> might be happening, though. Can you send me the code that you're
> running to generate the interrupt crash?
>
> Basically what this means is, removing smp_send_stop() has messed
> things up.
>
> I might have something that can fix this, but first I need to know,
> is this an SMP or non-SMP system?
>
> ... back to the drawing board. I'll fix this quickly.
>
> --Matt
>
> > Thanks again,
> > Mike Walfish
|