RE: Kernel Panic

From: Ted Mittelstaedt (tedm@toybox.placo.com)
Date: 06/24/01 

From: "Ted Mittelstaedt" <tedm@toybox.placo.com>
To: <js43064n@pace.edu>, <freebsd-questions@FreeBSD.ORG>, <freebsd-stable@FreeBSD.ORG>, <freebsd-security@FreeBSD.ORG>
Date: Sun, 24 Jun 2001 02:01:34 -0700

That would be impossible unless you had "." in your path. If
you did (which is a very BAD thing) then yes your script probably
loaded itself (assuming you named it "pine). This is why the
system defaults to NOT having "." in the path.

However, if the script DID load itself, a recursive script
under an ordinary user ID isn't allowed to crash the
system. I'd be very interested in seeing someone who
thinks differently post a script that could do this.
The ordinary users have limits restrictions that styme
forkbombs and such like that.

Of course, if a setuid-to-root binary has a programming bug
that allows an ordinary user to call it recursively, then
all bets are off. But pine is not normally setuid to root.

But, since you are apparently so determined to NOT believe that
your system has a hardware problem, in the interests of putting this
issue to final rest, here is what I did on my own system (FreeBSD
4.2)

1) Logged in as an ordinary user

2) Issued the commands

PATH=.:$PATH
export PATH

3) Created the script:

#!/bin/sh
./pine -i
rm -rf $HOME/dead.letter

(note that not only did I put current directory in my path - I forced the
script to reload itself just to be as convincing as possible)

4) Then issued the command:

pine

5) After about 16 seconds, I got this error message:

./pine: Cannot fork: Resource temporarily unavailable

and the script exited and came back to the $ prompt.

And, just for the sake of argument, I ALSO did the exact same thing
as the root user, and got the exact same response.

Please, you need to give the FreeBSD developers credit for SOME sense at
least.
Your not the only one that has ever built a recursive script by accident,
you know. FreeBSD is not so fragile that this kind of thing is going to
kill it.

Like I said, you need to focus your attention on your disk subsystem.
The script you built, called recursively, creates an intense burst of
disk activity until the system runs out of resources. It's that burst
of activity that's taking your system down, not the fact that the recursive
script is consuming all free resources. Undoubtedly, if you were to
run other equally disk-intensive programs on your machine they would cause
it to crash as well.

If you don't want to believe you have a hardware problem then fine, but
please don't throw unwarranted stones at the OS in place of that.

Ted Mittelstaedt tedm@toybox.placo.com
Author of: The FreeBSD Corporate Networker's Guide
Book website: http://www.freebsd-corp-net-guide.com

>-----Original Message-----
>From: owner-freebsd-questions@FreeBSD.ORG
>[mailto:owner-freebsd-questions@FreeBSD.ORG]On Behalf Of Jonathan Slivko
>Sent: Friday, June 22, 2001 8:56 AM
>To: js43064n@pace.edu; freebsd-questions@FreeBSD.ORG;
>freebsd-stable@FreeBSD.ORG; freebsd-security@FreeBSD.ORG; Ted
>Mittelstaedt
>Subject: RE: Kernel Panic
>
>
>I think I see what caused the kernel to crash. What happened was
>this, I believe, is that since I didn't specify the regular pine
>binary, the script just loaded itself, thus throwing it into a
>loop. It's really a sad situation. -- Jonathan
>
>______________________________________________
>Jonathan M. Slivko <js43064n@pace.edu>
>Technical Support, Black Lotus Communications
>http://www.blacklotus.net -- check us out!
>----------------------------------------------
>
>
>
>---------- Original Message ----------------------------------
>From: "Ted Mittelstaedt" <tedm@toybox.placo.com>
>Date: Thu, 21 Jun 2001 23:34:02 -0700
>
>>That absolutely will not crash a FreeBSD system that's not
>>got other problems.
>>
>>However, what I think is going on here is that the system that
>>you ran this on has buggy disk hardware. It's probably some
>>IDE disk, right?
>>
>>I've got a system here that I've tried 5 different IDE paddle
>>cards in, and on every one I've tried installing FreeBSD and
>>doing different operations and within about 20 minutes I
>>had crashed it and screweged the filesystem.
>>
>>I finally got so annoyed I dug up an old AHA1520 SCSI card
>>and slapped a 1GB SCSI disk on it (the system isn't intended
>>to be doing anything fancy) and it's been solid as a rock
>>ever since. The best conclusion I have is that the ISA bus
>>in the system has some clock speed error that doesen't affect
>>the SCSI disk system.
>>
>>Ted Mittelstaedt tedm@toybox.placo.com
>>Author of: The FreeBSD Corporate Networker's Guide
>>Book website: http://www.freebsd-corp-net-guide.com
>>
>>
>>>-----Original Message-----
>>>From: owner-freebsd-questions@FreeBSD.ORG
>>>[mailto:owner-freebsd-questions@FreeBSD.ORG]On Behalf Of
>Jonathan Slivko
>>>Sent: Thursday, June 21, 2001 4:59 PM
>>>To: freebsd-questions@FreeBSD.ORG; freebsd-stable@FreeBSD.ORG;
>>>freebsd-security@FreeBSD.ORG
>>>Subject: Kernel Panic
>>>
>>>
>>>Hello,
>>>
>>>I just wrote a little shell script that, on the machine I tested
>>>it on, crashed the box and forced a reboot. The contents of the
>>>script was:
>>>
>>>#!/bin/sh
>>>pine -i
>>>rm -rf $HOME/dead.letter
>>>
>>>Thats the whole script. I don't see how something like that
>could
>>>cause a kernel to crash. Would anyone mind trying to replicate
>>>this on a test box. If it's a security issue, i'll forward it to
>>>security when I get more information.
>>>
>>>-- Jonathan
>>>
>>>______________________________________________
>>>Jonathan M. Slivko <js43064n@pace.edu>
>>>Technical Support, Black Lotus Communications
>>>http://www.blacklotus.net -- check us out!
>>>----------------------------------------------
>>>
>>>_________________________________________________________________
>__
>>>___
>>>Sent via the Pace University Mail system at stmail.pace.edu
>>>
>>>
>>>
>>>
>>>
>>>To Unsubscribe: send mail to majordomo@FreeBSD.org
>>>with "unsubscribe freebsd-questions" in the body of the message
>>>
>>
>>To Unsubscribe: send mail to majordomo@FreeBSD.org
>>with "unsubscribe freebsd-stable" in the body of the message
>>
>
>
>___________________________________________________________________
>___
>Sent via the Pace University Mail system at stmail.pace.edu
>
>
>
>
>
>To Unsubscribe: send mail to majordomo@FreeBSD.org
>with "unsubscribe freebsd-questions" in the body of the message
>

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-security" in the body of the message

Relevant Pages

  • [Summary] puzzle with cold mirror script
    ... puzzle with cold mirror script ... DUMP: Date of last level 0 dump: the epoch ... all the disk information and a listing of the script follows. ... /export/home/chris# format ...
    (SunManagers)
  • Re: to lvm or not to lvm?
    ... A typo in a script can do more harm in 20 ... with disk space. ... As long as it's not possible for a growing partition to take space away from another partition, I can't see how this would do anything but give you more time to react, you run your messed up script that starts pushing the contents of /dev/zero into a text file and eating harddrive space, if your sitting next to the machine you see the HDD lights come on and hear the seeking, if your remote the first clue you get is an email or SMS saying /blah has been automatically grown by 5GB because it was more than 85% full, you can kill your script, delete the text file, reduce the partition again and nothing crashed, if all it had done was email you a message pointing out /blah was running out of space, and /blah was also required by some other vital process that ran out of space and crashed before you could kill the script, you'd wish you had auto growth. ... The LV extending commands were never ...
    (Debian-User)
  • Re: FGC to Grapheus
    ... probable/unsure sign's signifiance from sign's (probable/unsure, ... Phaistos Disk is NOT related in one way or the other to the LinearA/ ... Civilization, Greek yes, but NOT "Mycenaean"... ... "Cultures" (in particular Script) were somewhat different... ...
    (sci.lang)
  • ufsdump/ufsrestore error
    ... I am using ufsdump and ufsrestore in a script to copy one disk to ... Warning: `.' ... missing from directory ./sadm/pkg/SUNWcsr/save ...
    (SunManagers)
  • puzzle with cold mirror script
    ... This is the mirror script from BigAdmin modified just ... DUMP: Date of last level 0 dump: the epoch ... all the disk information and a listing of the script follows. ... /export/home/chris# format ...
    (SunManagers)
Quantcast