In comp.os.linux.misc Noah Roberts <nroberts@stmartin.edu>:

cairpre409@yahoo.com wrote:
>I need to crash and generate multiple system core files on my system.
>I am not sure how to force a crash and or core dump
>Any help is appreciated.
[]

Try writing to a null pointer?

cat /
#include <linux/config.h>
#include <linux/module.h>
#include <linux/init.h>

int __init crash_init(void)
{
printk ("crashing kernel\n");
*((char *) 0) = 0;
return 0;
}

module_init(crash_init);

Comes with the redhat netdump package, to test functionality of a
netdump installation (server/client).

In comp.os.linux.misc cairpre409@yahoo.com:
Thanks for the code. I am running redhat 9

EL since some time, better use a recent distro.

I forgot to mention that I am not a C programmer.
I assume I am suppposed to do something with gcc
Specific instructions would be quite appreciated.

Instructions come with the netdump package docs (README.client)
or/and can be found in a few seconds searching the archives.

BTW
Would be great if you could keep some context while replying,
sadly the default brain dump way groups.google uses, doesn't.
Making it a constant annoyance for us. Please check:

"Summarize what you are following up. When you follow up an
existing article, Google Groups includes the full article in
quotes, with the cursor at the top of the article. Tempting
though it is to just start typing your message, please STP and
do two things first. Look at the quoted text and delete parts
that are irrelevant. Then, go to the BTTM of the article and
start typing there. Doing this makes it much easier for your
readers to get through your post. They'll have a reminder of the
relevant text before your comment, but won't have to re-read the
entire article. And if your reply appears on a site before the
original article does, they'll get the gist of what you're
talking about."

Since most people aren't using groups.google to write on usenet.

In addition try our newreader FAQ, just scroll a bit down.

Good luck

cairpre409@yahoo.com wrote:
I need to crash and generate multiple system core files on my system.
I am not sure how to force a crash and or core dump
Any help is appreciated.
abort() ?

gcin wrote:
cairpre409@yahoo.com wrote:
>
>>I need to crash and generate multiple system core files on my system.
>>I am not sure how to force a crash and or core dump
>>Any help is appreciated.
>
>
abort() ?

Does that work in the kernel? It is not clear if the P. wants to crash an
application, where abort() is the normal way to do it, or the kernel, where
I do not suppose abort() is available.

In comp.os.linux.misc cairpre409 <cairpre409@speakeasy.net>:
Sat, 03 Sep 2005 07:51:55 +0000, Unruh wrote:

>Then why the hell are you trying to do core dumps?

I am a basic level SA. There are certain simple skills I
can perform on Solaris that I would like to have on Linux.
They are:

You could care less about the problem with proper maintenance
you won't see Linux crash like you won't see Solaris crash. This
isn't doze.

1. Recognizing a number of standard panic or opps messages.

Read kernel source.

2. Running a basic crash analysis using a debugger.

if you have something like the rh netdump client/server
setup you'll get a crash dump at all, otherwise there won't be
any unlike a $$ Unix which has hardware made for that.

3 Crash a system on request for testing purposes

I don't expect to ever be a kernel engineer or developer, but
I should be able communicate intelligently with them.
And also be able to handle and interpret common crashes.

It's very rare to crash Linux at all, usually it's some broken hw
or perhaps some buggy distro kernel.

ray wrote:

Seems from the discussion that you're not really talking about crashing
Linux, but rather having an app you've built crash - big difference. In
five years with over a dozen Linux distros on a couple of dozen machines I
have NEVER seen Linux crash short of a hardware problem.

I keep log books of my machines.

I have had three machines: in 1996, a 166MHz Pentium machine which started
out with 32Meg RAM and Windows 95 that crashed several times a day until I
got a lot of their updates. I still got about 3 blue screens of death a
week. But applications crashed many times a day. After a lot of work, I
determined I was not having hardware problems. I stuck Red Hat Linux 5.0 on
it and it crashed much less. In fact, most of the crashing was in X Window
System, not the kernel that continued to run. I upgraded the Linux on there
to Red Hat Linux 6.0 and later to RHL 7.3. Along the line, that machine got
its memory increased to 256 MBytes. I do ot have the logbook for that one.

In March 2000 I got a dual 550MHz Pentiun III machine with 512MBytes of RAM
that came with RHL 6.0, upgraded to RHL 6.2, and RHL 7.3, running
Linux-only. It is now running RHL 9 and Windows XP Home. X Window system
locked up May 11 (and I did not have the LAN working, so I could not kill
it), but I could tell the kernel was still running (occasional disk I).
May 15, same thing.
June 26, same thing.
24: kernel crashed when modem trying to dial out.
November 22: kernel crashed when modem trying to dial out.
January 3, 2001: kernel crashed when modem trying to dial out.
January 7; Installed VA Linux Systems version of Red Hat Linux 6.2.
January 7, 2001: kernel crashed when modem trying to dial out.
May 22, 2001: kernel crashed when modem trying to dial out.
May 30, 2001: kernel crashed when modem trying to dial out.
June 6, 2001: kernel crashed when modem trying to dial out.
July 31, 2001: kernel crashed when modem trying to dial out.
August 7, 2001: kernel crashed when modem trying to dial out.
September 19, 2001: kernel crashed when trying to hang up modem.
January 16, 2002: kernel crashed twice same day when trying to hang up modem.
Jan 21, 2002: X Window System crashed, but kernel K.
April 1, 2002: X Window System crashed, but kernel K.
Mid June, 2002: Installed new distribution (RHL 7.3, I think)
September 12, 2002: crashed.
March 3, 2003: crashed.
June 17, 2003: crashed.
Never crashed after that.

In about March 2004, I put this machine together. Dual 3.06GHz Xeon
processors, 4 GBytes RAM, etc. I have run Red Hat Enterprise Linux 3 ES on
it ever since, 24/7 except when the power company has let me down. I reboot
it only when the power is off for over an hour and my UPS poops out, or when
I install a new kernel. Most recently, I rebooted it to run memtest86
because some program (I forget which) was flakey and people suggested a
marginal memory. Memory was fine and I never found out what the trouble was.
As far as I can tell, this machine has never crashed.

Michael Heiming writes:
if you have something like the rh netdump client/server setup you'll
get a crash dump at all, otherwise there won't be any unlike a $$ Unix
which has hardware made for that.
Both System III and BSD produced crash dumps without special hardware.

In comp.os.linux.misc John Hasler <john@dhh.gt.org>:
Michael Heiming writes:
>if you have something like the rh netdump client/server setup you'll
>get a crash dump at all, otherwise there won't be any unlike a $$ Unix
>which has hardware made for that.

Both System III and BSD produced crash dumps without special hardware.

Great, my point was Linux doesn't produce them unless you are
using something like rh netdump, in opposite to $$ Unix which
usually write the dump to the swap partition which is checked for
dumps on reboot and recovers them.

There's lkcd for Linux which is supposed to do just that, however
I haven't seen this working reliable.

While the rh provided netdump works flawlessly. AFAIR the reasons
for better using the network to dump to another system then swap
has something to do with x86 hardware according to the rh docs
about netdump (iirc). Nice if BSD allows that, but alas despite
using never saw a BSD system crashing. ;-)

Michael Heiming writes:
Nice if BSD allows that, but alas despite using never saw a BSD system
crashing.
My BSDI installation crashed frequently (it was beta software, though).