Signal 24 (SIGXCPU) - CPU time limit exceeded

3624 views
Skip to first unread message

Geoff McHugh

unread,
Apr 21, 1998, 3:00:00 AM4/21/98
to

I keep getting intermittent errors with Linux on my PC. Often, gcc will
stop compiling with an internal error reporting signal 24. I compiled
v2.0.33 of the kernel with sound & network support for my hardware, and
now it crashes out of the X server with a 'signal 24'

My PC is a 133MHz Intel Pentium, 128MB RAM, 1.2GB & 3.1GB WD HDD, 2
ATAPI IDE CDROM drives (24x & 4x). I checked the signal.h file to find
out what signal 24 was, and it's SIGXCPU - CPU time limit exceeded
(which is also an error message I have been seeing occasionally.

I've searched the web, and the only detailed info about signals is
referring to signal 11 and seems to indicate faulty RAM from what I have
read.

The 2.0.33 kernel *always* bombs out if the X server (XFree86 3.3.2)
while the original kernel (Caldera OpenLinux Base 1.1 - 2.0.29 if I
remember correctly) runs X perfectly well.

After all that (phew) is there anyone who has any experience of what may
be causing signal 24 to be causing me such grief? If anyone can suggest
any possible solutions (hopefully apart from 'buy a new PC'!!) or can
point me in the direction of some information which may help I'd be very
grateful.

Thanks
Geoff McHugh
gmc...@enterprise.net

r...@greenend.org.uk

unread,
Apr 21, 1998, 3:00:00 AM4/21/98
to

Geoff McHugh <gmc...@enterprise.net> writes:

> I checked the signal.h file to find out what signal 24 was, and it's
> SIGXCPU - CPU time limit exceeded (which is also an error message I
> have been seeing occasionally.

Do "ulimit -a" to list all limits; you'll probably see that cpu time
is limited to some specific amount. This is bad l-)

"ulimit -t unlimited" should fix it for the login that you run it in
and any future children of that process (i.e. all commands you type in
that login before next logging out), but if you look through startup
scripts you'll probably find some location where the limit is set to
its current value - and that's where the problem is.

Geoff McHugh

unread,
Apr 22, 1998, 3:00:00 AM4/22/98
to r...@greenend.org.uk

Thanks for the help.

I tried "ulimit -a" as you suggested and the cpu time limit was set to
-1 seconds. I set it to unlimited and the X server started up fine.
Now I will have to spend some time and try to track down where it is
setting the time limit to -1 in the first place.

Then I'm going to download the System Administrators Guide and other LDP
documents and read them very carefully... :-)

Reply all
Reply to author
Forward
0 new messages