Re: Moab and socktest

• To: Vaughan Knight <vsk01@uow.edu.au>
• Subject: Re: Moab and socktest
• From: Patrick Tullmann <tullmann@cs.utah.edu>
• Date: Wed, 2 May 2001 10:33:13 -0600
• Cc: OSkit Users <oskit-users@fast.cs.utah.edu>, Janos <janos-discuss@fast.cs.utah.edu>
• In-Reply-To: <3AEEF1DD.AA24BCA3@uow.edu.au>; from vsk01@uow.edu.au on Wed, May 02, 2001 at 03:26:53AM +1000
• Mail-Followup-To: Patrick Tullmann <tullmann@cs.utah.edu>, Vaughan Knight <vsk01@uow.edu.au>, OSkit Users <oskit-users@fast.cs.utah.edu>, Janos <janos-discuss@flux.cs.utah.edu>
• References: <3AEEF1DD.AA24BCA3@uow.edu.au>
• Sender: owner-oskit-users@fast.cs.utah.edu
• User-Agent: Mutt/1.2.5i

Vaughan Knight wrote:
> Then it crashes, giving a HEX backtrace and then giving this
> output: "NodeOS: Got signal 11"

That's an error.  It means the OSKit or Moab segfaulted in an
unexpected place.  You can convert the hex backtraces to symbolic
backtraces by running the sym-trace script installed by the OSKit over
the kernel (not the bmod).  You'll want to look the script over and
perhaps hack it up a bit for your environment.

Do you have more than four ethernet cards in the machine this fails
on?  There's a bug in the OSKit startup library that would cause
problems, if you do.

> Secondly, I have been reading through some of the code and I have been
> thinking that maybe the SocketTest example needs to be recoded with
> OSkit specific initialization code.  Any help here is greatly
> appreciated too.

It shouldn't be necessary to add OSKit-specific initialization, that
should all be taken care of during Moab's setup phase.

-Pat

----- ----- ---- ---  ---  --   -    -      -         -               -
Pat Tullmann                                       tullmann@cs.utah.edu
	      The early bird gets the worm, so sleep in.

• Moab and socktest
• From: Vaughan Knight <vsk01@uow.edu.au>

• Prev: Kernel/Shell newbie question
• Next: Useing Asembly with the GCC compliler
• Index(es):
  • Main
  • Thread