Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Crash test

4 views
Skip to first unread message

Gordon Miller

unread,
Mar 24, 1992, 12:25:34 PM3/24/92
to
I took the challenge and ran crash.com on my Zenith 486/25E.
I ran it in a DOS window (6.304 beta) and it DID NOT hang the system.
After only three seconds OS/2 informed me that the program had
executed an illegal instruction and gave me the option of killing
the process. I selected kill and the DOS session went away whilst
everything else continued to run. I love it.


Gordon Miller
g...@cray.com


P.S. This is my first attempt at posting to this newsgroup so forgive me
if I end up somewhere I do not belong.

Timothy F. Sipples

unread,
Mar 24, 1992, 1:26:12 PM3/24/92
to
In article <1992Mar24.1...@hemlock.cray.com> g...@cray.com (Gordon Miller) writes:
>I took the challenge and ran crash.com on my Zenith 486/25E.
>I ran it in a DOS window (6.304 beta) and it DID NOT hang the system.
>After only three seconds OS/2 informed me that the program had
>executed an illegal instruction and gave me the option of killing
>the process. I selected kill and the DOS session went away whilst
>everything else continued to run. I love it.

Does 486/25E mean an EISA system? If so, that could be the reason.
Apparently ISA systems do not have watchdog timers while EISA (and
MCA) systems do.

Who said OS/2 2.0 didn't support EISA? :-)
--
Timothy F. Sipples Keeper of the OS/2 Frequently Asked Questions
si...@ellis.uchicago.edu List, available via anonymous ftp from
Dept. of Economics 128.123.35.151, directory pub/os2/faq, or via
Univ. of Chicago 60637 netmail from LIST...@BLEKUL11.BITNET.

Pat Duffy

unread,
Mar 24, 1992, 1:56:47 PM3/24/92
to
In article <1992Mar24.1...@hemlock.cray.com> g...@cray.com (Gordon Miller) writes:
>I took the challenge and ran crash.com on my Zenith 486/25E.
>I ran it in a DOS window (6.304 beta) and it DID NOT hang the system.
>After only three seconds OS/2 informed me that the program had
>executed an illegal instruction and gave me the option of killing
>the process. I selected kill and the DOS session went away whilst
>everything else continued to run. I love it.
>

What's crash.com? Is it that little 6-line BASIC program that scribbles all
over memory?

I love 6.304E too (after the bug fixes :-). Gawd, I can't waaaaaait for GA.
--
Patrick Duffy, E-Mail: du...@theory.chem.ubc.ca

Hi! I am a .signature virus. Copy me into your .signature to join in!

Gordon Miller

unread,
Mar 24, 1992, 4:38:52 PM3/24/92
to
>>I took the challenge and ran crash.com on my Zenith 486/25E.
>>I ran it in a DOS window (6.304 beta) and it DID NOT hang the system.
>>After only three seconds OS/2 informed me that the program had
>>executed an illegal instruction and gave me the option of killing
>>the process. I selected kill and the DOS session went away whilst
>>everything else continued to run. I love it.

>Does 486/25E mean an EISA system? If so, that could be the reason.


>Apparently ISA systems do not have watchdog timers while EISA (and
>MCA) systems do.
>
>Who said OS/2 2.0 didn't support EISA? :-)


Yes, the E means EISA.

Come on IBM, tell the world what a great product OS/2 2.0 is.


---

Gordon Miller
g...@cray.com


Mikael Wahlgren

unread,
Mar 25, 1992, 12:44:31 PM3/25/92
to
In article <1992Mar24.1...@unixg.ubc.ca> du...@theory.chem.ubc.ca (Pat Duffy) writes:

>What's crash.com? Is it that little 6-line BASIC program that scribbles all
>over memory?

No, it is a three byte long program that craches Windows (and OS/2 on some
hardware).

Write:

COPY CON CRASH.COM

and type the following

ALT-250, ALT-235, ALT-254, CTRL-Z

(this means press and hold the ALT-key and type 2-5-0 on the numeric keypad).
Save all work before running the program...

Mikael Wahlgren d9mi...@dtek.chalmers.se

Rick Stoen

unread,
Mar 26, 1992, 1:34:40 PM3/26/92
to
In article <10...@chalmers.se>, d9mi...@dtek.chalmers.se (Mikael Wahlgren) writes:

>
> COPY CON CRASH.COM
>
> and type the following
>
> ALT-250, ALT-235, ALT-254, CTRL-Z
>
> (this means press and hold the ALT-key and type 2-5-0 on the numeric keypad).
> Save all work before running the program...
>
> Mikael Wahlgren d9mi...@dtek.chalmers.se

Well, I tried it on my Compaq 486/25 (EISA) running OS/2 2.0 6.304 and
it wasn't very exciting. It just sat there, so I started up another
one, and another one, etc. With 4 of them running my other DOS windows
became VERRRY slow, however nothing else showed any signs of slowness.
After closing the windows running crash my other DOS boxes returned to
normal. I tried it with a DOS 5.0 virtual machine and then it did
come back and say that the session encountered a problem and can not
continue.

Gee, I'm 2 for 2 (OS/2 sees and uses all 32MB of memory and crash
doesn't crash it. I guess the old adage "you get what you pay for" is
true. :):)

========================================================================
| Rick Stoen | |
| Systems Engineer | Phone: (713) 374-4514 |
| Compaq Computer Corporation | FAX: (713) 374-7305 |
| P.O. Box 692000 - M050701 | EMail: st...@compaq.com |
| Houston, TX 77269-2000 | |
========================================================================
Disclaimer: The views expressed here are my own, and do not reflect
those of my employer.

Timothy F. Sipples

unread,
Mar 26, 1992, 3:24:05 PM3/26/92
to
In article <1992Mar26....@twisto.eng.hou.compaq.com> st...@katana.eng.hou.compaq.com (Rick Stoen) writes:
>> COPY CON CRASH.COM
>> and type the following

(in a DOS session, incidently...)

>> ALT-250, ALT-235, ALT-254, CTRL-Z
>> (this means press and hold the ALT-key and type 2-5-0 on the numeric keypad).
>> Save all work before running the program...

>Well, I tried it on my Compaq 486/25 (EISA) running OS/2 2.0 6.304 and
>it wasn't very exciting. It just sat there, so I started up another
>one, and another one, etc. With 4 of them running my other DOS windows
>became VERRRY slow, however nothing else showed any signs of slowness.
>After closing the windows running crash my other DOS boxes returned to
>normal. I tried it with a DOS 5.0 virtual machine and then it did
>come back and say that the session encountered a problem and can not
>continue.
>Gee, I'm 2 for 2 (OS/2 sees and uses all 32MB of memory and crash
>doesn't crash it. I guess the old adage "you get what you pay for" is
>true. :):)

Maybe (for the sake of ISA machines which don't have watchdog timers)
we could have yet another DOS Setting:

Prompt_CLI: {on|off}

If off, CLI would be executed in any DOS session (with the rare
possibility that if, for some reason, the DOS program doesn't manage
to issue a subsequent STI, the system locks; incidently, I can't think
of any DOS program which could do this and not lock single-tasking DOS
itself, for all intents and purposes, but you never know).

If on, a pop up dialog would appear:

"DOS program attempted to execute a CLI instruction (to turn
off system interrupts). The following options are available:

End Program

Ignore CLI and Continue
In rare cases the DOS program will exhibit erratic
behavior.

Execute CLI and Continue (Default)
DOS program, in rare cases, may compromise system
integrity. Save work in other sessions before
selecting this option."

This dialog would appear only once. If either Ignore or Execute was
selected and another CLI encountered in the same program, the system
would take whichever action was selected previously.

Maybe for OS/2 2.1.

Rick Stoen

unread,
Mar 26, 1992, 8:00:06 PM3/26/92
to

< My ramblings of how it worked on my Compaq 486/25 (EISA) deleted


>
> Maybe (for the sake of ISA machines which don't have watchdog timers)
> we could have yet another DOS Setting:
>
> Prompt_CLI: {on|off}
>

< deleted Timothy's explanation on how to implement this.


>
> Maybe for OS/2 2.1.
> --

Just for kicks I tried this on a Compaq 386/25 (ISA) machine running
6.307C (instead of 6.304). It did basically the same thing as my
486/25 did. So did they fix something in 6.307 that makes OS/2 immune
to crash.com? Has anyone else tried it on a ISA machine running
6.307C? Maybe we don't have to wait for 2.1 after all.

Colin Goldstein

unread,
Mar 27, 1992, 1:51:24 PM3/27/92
to
>Just for kicks I tried this on a Compaq 386/25 (ISA) machine running
>6.307C (instead of 6.304). It did basically the same thing as my
>486/25 did. So did they fix something in 6.307 that makes OS/2 immune
>to crash.com? Has anyone else tried it on a ISA machine running
>6.307C? Maybe we don't have to wait for 2.1 after all.

Sorry to be the bearer of bad news, but on an AST 386/33 running
the latest 307 Beta, crash.com is still deadly. Maybe it has to
do with the fact you're running on a Compaq.

The rest of us will have to wait until 2.1.

Colin
--
/-------------------------------------------------------------------\
| The views expressed here are my own. | How's life treating you |
| They do not necessarily represent | Norm??? |
| the views expressed by my employer. | |
| ---------------| Like a dog treats a |
| co...@novell.com | Novell Inc., | fire hydrant. |
| uunet!novell!colin | San Jose, CA | - Cheers |
\-------------------------------------------------------------------/

jste...@desire.wright.edu

unread,
Mar 27, 1992, 7:45:55 PM3/27/92
to

It'd be simple enough to filter for three byte dos progs that exacly
match crash...or you could just always check right after a cli to see if they
go into a infinate loop.

John Stewart

0 new messages