Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!cs.utexas.edu!uunet!ctbilbo!ray
From: ray@ctbilbo.UUCP (Ray Ward)
Newsgroups: comp.unix.programmer
Subject: Re: Problems with execl("bin/csh", ...) in my OWN shell from login
Message-ID: <169@ctbilbo.UUCP>
Date: 22 Mar 91 19:53:00 GMT
References: <12602@ucrmath.ucr.edu> <12860@ucrmath.ucr.edu>
Lines: 33

In article <12860@ucrmath.ucr.edu>, yakker@ucrmath.ucr.edu (matt robinson) writes:
> In article <12602@ucrmath.ucr.edu> yakker@ucrmath.ucr.edu writes:
> >The problem I'm having is that I need to start up a new /bin/csh from my
> >/bin/eggsh, [...] 
> 
> Well, this was my own article, and I didn't have anyone reply to it (except
> for two individuals, whose answers didn't work.)  [...]
> 
> The suggestions given were to try execl("/bin/csh", "-csh", (char *)0),
> or execl("/bin/csh", "-", (char *)0), both of which do not work.

However, both of these DO work on the machine I am on at the moment, a
XENIX SysV 386 :-( .  So it would appear that the problem lies elsewhere.

I only ran the execl calls in a small test  program; I did not run your
entire source excerpt as your utmp structure used fields this system's
utmp.h did not have.

You did not mention the machine that you are running on or the extent of
the code in your eggsh ( but, then, I didn't see your original posting ).  
On some systems I have worked on I have experienced similar it-just-hangs 
symptoms from someone's wild pointer stepping on something.

Try the calls in a small test program that only does an execl().  If
the calls work in the same context without the other code present, I would
bet the problem is in the other code.


-- 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Ray Ward                                          Email:  uunet!ctbilbo!ray  
Voice:  (214) 991-8338x226, (800) 331-7032        Fax  :  (214) 991-8968     
=-=-=-=-  There _are_ simple answers, just no _easy_ ones. -- R.R. -=-=-=-=