Re: (longish) tcsh should be removed from hamm
I got the bug to appear on a Sparc and an Alpha. I also produce it
a different way. Somehow my message this moring seems to have gone
unnoticed. The text below is available through the bug tracking system.
However, I reproduce my report here. Looking at the strace, it seems also
that tcsh is trying to write "exit" to stdout and can't.
This bug is causing many runaway tcsh procesess to pile up from
several users on several machines. In particular, people dialing and
using slirp cause the problem, although it can happen with people using
X11 from the console.
This bug occurs under sunos and digital unix as well. Under
these two operating systems, it causes a core dump and kills parent
processes as well. First, here is how to witness the bug:
1. login ( or use xon, etc. if accessing machine remotely)
2. start a new shell from the login shell, (i've tried bash or
tcsh )
3. start a new tcsh from the shell started in (2)
4. start a shell (bash or tcsh ) from the shell started in (3).
5. Do 'kill -6' on the pid of the shell created in (2).
At this point the tcsh started in (3) finds that the shell created
in (4) has exited and shortly thereafter either a core dump (digital and
solaris) or an infinite loop eating proccesor time (Debian) occurs. In
addition, on solaris and digital the shell from step (2) exits as well.
Under Digital:
version tcsh 6.05.00 (Cornell) 94/19/06 (alpha-dec-osf1) options
8b,nls,dl,al,rh
message printed on occurence of bug is:
IOT/Abort trap (core dumped)
Under Sunos:
version tcsh 6.05.00 (Cornell) 94/19/06 (sparc-sun-sunos4) options
8b,dl,al,rh
message printed on occurence of bug is:
Abort (core dumped)
G John Lapeyre <lapeyre@physics.arizona.edu>
Tucson,AZ http://www.physics.arizona.edu/~lapeyre
--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: