CRASH.COM revisited
Harald Boegeholz
Since I posted my original article containing crash.com in
comp.os.os2.programmer, it has raised considerable discussion in all
three OS/2 groups.
Some people don't consider the CLI problem a serious one (i.e., people
with machines that have a watchdog timer). For those, I have an update
:-) of crash.com. Enter crash2.com as in the following sample debug
session:
I:\ >debug
-a
0FC1:0100 cli
0FC1:0101 jmp 100
0FC1:0103
-rcx
CX 0000
:3
-ncrash2.com
-w
Writing 00003 bytes
-q
I:\ >
Then, make sure you don't have any important processes running and run
crash2.com in a DOS session under OS/2 2.0. Please report the result!
In case you missed my original article in c.o.o.p, please mail me and
I'll send you a copy.
Harald
--
Harald Boegeholz |Home: h...@texnix.stgt.sub.org (read daily)
|University: h...@phoebus.mathematik.uni-stuttgart.de
|please don't send large (>100k) mail to my home address.
Barr3y Jaspan
duplicated on all OS/2 newsgroups.
I ran the new crash.com (which looks to me to be the same as the old one,
although I don't know what the rcx 3 command does) on a PS/2 running the LA
release. The machine got VERY SLOW (about a 1 second response time to mouse
events) but did not hang. Perhaps my config.sys is not set up intelligently
to handle runaway DOS programs. When I killed the DOS window running
crash.com, everything went back to normal.
--
Barr3y Jaspan, bja...@mit.edu
Rick Stoen
> I have an update
> :-) of crash.com. Enter crash2.com as in the following sample debug
> session:
>
<Debug script deleted>
> Then, make sure you don't have any important processes running and run
> crash2.com in a DOS session under OS/2 2.0. Please report the result!
>
> In case you missed my original article in c.o.o.p, please mail me and
> I'll send you a copy.
>
Well, I tried this "new and improved" crash on a Compaq 486/25 (EISA)
running 6.304 and a Compaq 386/25 (ISA) running 6.307C and they didn't
crash. In fact I had so much confidence I didn't even bother to close
down my other stuff before running it. This new version did slow down
the whole machine as opposed to just the DOS boxes with the original
crash.com. It also made the mouse basically unusable, but <Alt>CY is
your friend.
========================================================================
| 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.
David W. Levine
heck out of the system, made the mouse a little erratic. I used ctrl-esc
to pull up the task list, ticked down to the errant task, seleteced it
with shift-f10 (to get the context menu) and killed it. Blip, the dos
session went away, and off I went. No big deal.
David W. Levine -- IBM Thomas J. Watson Research Center
d...@watson.ibm.com -- (914) 784-7427
My Opinions, IBM's hardware. --
robert cook
Their is a difference. The Old Crash.com, jumped to the JMP instruction,
while the new one jumps to the CLI instruction. His thought that the
watchdog timer would reset itself everytime CLI was executed. From
the results people are posting, this is apparently not true. Both
programs crash my el-cheapo 386sx machine though.
BTW "rcx 3" justs tells debug that the program is three bytes long.
--
Robert Kelley Cook rc...@darwin.cc.nd.edu
Univ. of Notre Dame
Physics Dept. If you think these are ND opinions, speak to the pope!
zl...@vax5.cit.cornell.edu
> Some people don't consider the CLI problem a serious one (i.e., people
> with machines that have a watchdog timer). For those, I have an update
> :-) of crash.com. Enter crash2.com as in the following sample debug
> session:
>
> Then, make sure you don't have any important processes running and run
> crash2.com in a DOS session under OS/2 2.0. Please report the result!
This is all fine, but I'm not sure if I understand the motivation. It this a
response to my query as to whether there was a program that could crash OS/2
but not Windows? [I'm the person who posted that famous 6-line BASIC program]
So: does this .com crash OS/2 but leave Win up and running, to reveal a
relative system stability problem? Or, does it take Windows down also?
- John Hwang.
Standard Disclaimer.
David R. Giller
Actually, it was an attempt to come up with a DOS program that would be
guaranteed to crash OS/2. That's the first step. THEN we can try to find one
of these that DOESN'T crash Windows. Good luck, though.
-Dave
--
David Giller, Box 134 | Q: How many Oregonians does it take to screw in a light
Occidental College | bulb? A: Three. One to replace the bulb, and two to
1600 Campus Road | fend off all the Californians trying to share the
Los Angeles, CA 90041 | experience. ---------------------------rafe...@oxy.edu
Stephan Zeisset
In article <1992Mar31.1...@cheshire.oxy.edu>, rafe...@cheshire.oxy.edu (David R. Giller) writes:
|> In article <1992Mar30.1...@vax5.cit.cornell.edu> zl...@vax5.cit.cornell.edu writes:
|> >boeg...@hermes.informatik.uni-stuttgart.de (Harald Boegeholz) writes:
|> >> Some people don't consider the CLI problem a serious one (i.e., people
|> >> with machines that have a watchdog timer). For those, I have an update
|> >> :-) of crash.com. Enter crash2.com as in the following sample debug
|> >> session:
|> >> [interesting program deleted]
|> >>
|> >> Then, make sure you don't have any important processes running and run
|> >> crash2.com in a DOS session under OS/2 2.0. Please report the result!
Hey, can anyone explain to me, why crash2.com should lock up
even machines with an watchdog timer ???
No matter how often it is invoked, CLI should only prevent
maskable Interrupts and not the Watchdog-Timeout - or am I
misguided ?
Alan Ballard
> His thought that the
>watchdog timer would reset itself everytime CLI was executed. From
>the results people are posting, this is apparently not true.
My understanding from the little documentation I have is that the "watchdog
timer" is a separate timer channel (of the 8254A PIT) that is enabled
by the IRQ0 line, so if IRQ0 stays up for too long, it will trigger,
presumably a NMI. This appears to be a feature of PS/2 hardware + the
Advanced BIOS software. I don't think it existed on pre-microchannel
machines, but I don't see any reason why it couldn't be included on
ISA and EISA bus machines. Anybody know if IBM's ISA PS/2's include it?
--
Alan Ballard | Internet: bal...@ucs.ubc.ca
University Computing Services | Bitnet: USERAB1@UBCMTSG
University of British Columbia | Phone: 604-822-3074
Vancouver B.C. Canada V6T 1Z2 | Fax: 604-822-5116
Mikael Wahlgren
>response to my query as to whether there was a program that could crash OS/2
>but not Windows? [I'm the person who posted that famous 6-line BASIC program]
>So: does this .com crash OS/2 but leave Win up and running, to reveal a
>relative system stability problem? Or, does it take Windows down also?
The three byte crash.com progam definitely take down Windows 3.0 and Windows
3.1 (I was going to try it on Windows NT, but they wouldn't let me...:-).
It takes down OS/2 2.0 also, on some machines, and on some other machines
with a better hardware support, it doesn't.
Mikael Wahlgren d9mi...@dtek.chalmers.se