[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: A newbie recursion problem...:-)))) (fwd)



Yesterday i make a test like this in a virtual console
not as root(X is not running).

I started the program and after a few seconds it began
to give the message below. 

./test: fork: try again

then i changed th virtual console and try pressing
Ctrl-Alt-Del but the system didn't respond.
The after 10 minutes i pressed the reset button.

i think too that there must be a way to restrict the
processes to call them recursively in a limit.
---------- Forwarded message ----------
Date: Thu, 05 Dec 1996 20:06:57 +0100
From: a2863057@smail1.rrz.Uni-Koeln.DE
To: "ken (k.y.) chan" <kenchan@nortel.ca>
Cc: debian-user@lists.debian.org
Subject: Re: A newbie recursion problem...:-))))
Resent-Date: Thu, 5 Dec 1996 22:06:11 +0300
Resent-From: debian-user@lists.debian.org

ken (k.y.) chan wrote:
> 
> In message "A newbie recursion problem...:-))))",
> 'a2863057@smail1.rrz.Uni-Koeln.de' writes:
> 
> >Hi to all,
> >
> >well I do not know whether this is a bug or not...
> >Suppose a user would create a file with say ten lines, each line
> >containig just the word test. Then s/he will save it as test and s/he
> >will make it executable. Now s/he will open say 4 xterminals and in each
> >he will just type test & (followed by enter). On my Debian this will
> >cause the system to nearly "stop" and the load of the system will never
> >end...
> 
> In fact, if you have 10 lines of "test &" in each test file instead,
> you would be creating new process at a rate of 4*10 process per
> *cycle* :).
> 
> Also, you don't need to have ten lines of  "test"  in your executable
> called test. All you need is one line of "test", then you will have
> an infinite execution recursion if you are running it under the
> same directory..etc, at a rate of 4 new processes per *execution*  :)
> 
> >I remember that when this user types test & in each line of the file the
> >system will slow down too and I think it never stops swapping...
> >The system did not crash but was no more usable untill killing of X.
> >Well I'm a newbie, but someone knows, whether there is a cure to this???
> >I think that the kernel should prevent this recursion...any
> >propositions??? Solutions are welcome....
> >
> If you are not running as a root, then you should be restricted to
> a limited of number of processes that you can have at any time (I
> forgot where it is defined). Linux should always reserve one process
> *space* in the system so that if you have too many processes running,
> you can still
> spawn a KILL process to kill them all. If not, I consider that
> a bug in the kernel.
I was running as a user not root, and I think that I was not restricted
to
a limited number of processes, well in a way...
I only got the error Message:
./test: fork: Try again

This message "never" stopped and the system slowed down every second...
The only way for me to stop all this was to close the xterminal...the
time 
I could move the pointer...or later only by killing X with <Ctrl-Alt-F1>
I suppose that this IS a bug, because I know, that e.g. on a sparc or
dec or HP a
recursion like this is just killed by the OS. My system has 40MB of RAM
and finally I
had just 400KB of free memory, only after running this little
recursion...:-)))
BTW I'm running the 2.0.25 kernel. Oh...where can I define the max
number of processes
for a user???

Greetings

Andreas


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-user-REQUEST@lists.debian.org . Trouble? e-mail to Bruce@Pixar.com



--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-user-REQUEST@lists.debian.org . Trouble? e-mail to Bruce@Pixar.com


Reply to: