*BSD News Article 74597


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!munnari.OZ.AU!news.ecn.uoknor.edu!news.ysu.edu!odin.oar.net!malgudi.oar.net!imci4!newsfeed.internetmci.com!pier2.bayarea.net!baygate.bayarea.net!thorpej
From: thorpej@baygate.bayarea.net (Jason R. Thorpe)
Newsgroups: comp.unix.bsd.netbsd.misc
Subject: Re: How to make DDB dump core
Date: 24 Jul 1996 19:51:13 GMT
Organization: George's NetBSD answer man
Lines: 34
Message-ID: <4t5urh$46m@pier2.bayarea.net>
References: <4sebp2$fhj@redwood.cs.sc.edu>
NNTP-Posting-Host: baygate.bayarea.net

In article <4sebp2$fhj@redwood.cs.sc.edu>,
Kevin Brown <kbrown@tupelo.sc.edu> wrote:

>The changes I've added to the TCP code in NetBSD cause it to crash (of
>course).  This I don't mind.  However, I've compiled in DDB which
>catches the panic (which is useful).  I cannot get DDB to dump core
>though so that I can later use kgdb to analyze the crashdump.
>
>I've looked at the FreeBSD site which suggests typing either:
>
>db> call diediedie()
>or
>db> call panic()

...or, you can do:

 db> continue

...and it should continue on to call boot() with the appropriate arguments
that should generate the crashdump.

Alternatively, you can call boot() manually with the appropriate arguments.

>On a related note, I think I made the swap partition too small (just
>over 32 MB) - which seems to be keeping the crash dump from working
>correctly even when DDB is not configured.  Is it possible to use a
>crash file like L*n*x does?

Really, a crash file is a Bad Idea.  What happens when, for example,
the panic is the result of a filesystem problem?

-- 

	Jason R. Thorpe <thorpej@bayarea.net>