Re: What is the "do_BUG" and why does my rtl.o want it?
"Norm" <[email protected]>
| Newsgroups | gmane.linux.real-time.rtlinux.general |
|---|---|
| Message-ID | <010e01c4d132$86c6ec10$2f01a8c0@never> |
After fooling around for a day, I finally gave up and remade the kernel and
rtlinux modules on the new HD using gcc 2.95.3 compiler with the original
setup -- kernel 2.4.18 & RTL3.2-pre1. It works just fine -- at least as
well as the previous one did. Since I had tried remaking just the rtlinux
modules with no success, I have to conclude that there was something wrong
in the kernel image I had that only showed up when the rtlinux code was
activated.
One of the interim steps in my experiments was to rescue the kernel image
from the failed HD and try to load rtlinux with that. It also failed with
the do_BUG problem so I have to believe that on the original, working HD,
the kernel image didn't have a do_BUG entry point and the rtlinux modules
used there didn't reference them.
BUT ... I still don't know what the do_BUG is or why the new rtlinux
modules I have require that kernel symbol. Can anyone enlighten me?
Norm
----- Original Message -----
From: "Der Herr Hofrat" <[email protected]>
To: "Norm" <[email protected]>
Cc: "rtlinux" <[email protected]>
Sent: Sunday, November 21, 2004 4:06 AM
Subject: Re: [Rtl] What is the "do_BUG" and why does my rtl.o want it?
> > THE DIRTY DETAILS:
> > The kernel is a "raw" 2.4.18 source tree downloaded from kernel.org
and
> > both kernels were patched with RTL3.2-pre1 and were in fact, made from
> > exactly the same source tree. What's really puzzling is that I had
another
> > HD with the same setup -- 2.4.28 patched with RTL3.2-pre1 -- and it
worked
> > just fine [until a system crash destroyed the root partition, but that's
a
> > different problem]. The current HD is a secondary disk which isn't an
exact
> > backup but contains much the same stuff.
> >
> > SO ...
> > 1. What is do_BUG? Is it something that's only included in RTLinux
> > with a P-III? Could I remake RTLinux to avoid it?
> the do_BUG is the function that produces the oops messages (reg dump,
stack
> etc.) - I'm supprised that this is not defined in a 2.4.18 kernel - I
would
> have guessed that it must be much older to not have this built in.
>
> > 2. Other than completely downloading a kernel and rtlinux patch and
> > completely remaking the system, what's the most reasonable strategy for
> > solving the system hang problem? [If I have no choice, then I have no
> > choice and I'll do it but I'm trying to save a full day's work in
> > downloading and compiling if I can.]
>
> pre1 had some problems (especially on SMP) so maby updating is a good idea
> any way - pleas compile it for a generic 686 CPU and let me know if it
hangs
> I have a DMP Celeron running here with pre1/2/3/rc1 and it does not hang
> (compiled as generic 686).
>
> hofrat
_______________________________________________
Rtl mailing list
[email protected]
http://www2.fsmlabs.com/mailman/listinfo/rtl