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

Re: son of Delta


I've seen the same thing happen with running "./rsts 0".  I think it's the
Berkeley job control stuff getting in the way.  The process is able to write
to the terminal, but blocks on input waiting to become the controlling
process for the terminal (which never happens).  If you logoff and back-on,
or run it from another tty then it works fine.

To tell you the truth, I'm kind of surprised job control doesn't always get
in the way.  Before I modified the simulator I ran some experiments on
Solaris and Linux to see if I could have a nohuped background process open
the tty of a logged in process and do reads and writes, and to my surprise
it actually worked.  It wasn't until much later that I found there was this
case where the terminal that started the simulator was not able to then run
the little rsts front-end.  I suspect it has something to do with the
terminal that initiated the simulator being in the same process group as the
simulator and this causes the job control stuff to kick into effect.
There's probably a way to fix this by issuing an ioctl() on the terminal to
reset its process group, but I haven't tried that yet.  I tried doing a
setpgrp() from the simulator after the fork(), but that didn't fix it.  For
now, the workaround is to run ./rsts 0 from a different tty than the
simulator.  I have this in the README file in /home/delta.

-----Original Message-----
From: Eric Thayer <eht@cs.cmu.edu>
To: 'Gary L. Luckenbaugh' <garyll@ibm.net>
Date: Wednesday, July 22, 1998 9:41 AM
Subject: son of Delta

>I got it working.  For some reason, the tty that is closed by the forked
>process is not able to successfully issue a ./rsts 0.  It behaves as if the
>input side is closed, but the output is still open (i.e. I see the
>prompt, but by <LF> is ignored).  Issuing a ./rsts 0 from some other tty
>worked (this is on a DEC Alpha running Digital Unix; still on Digital (er,
>uh Compaq) hardware after all these years).
>-----Original Message-----
>From: Gary L. Luckenbaugh [mailto:garyll@ibm.net]
>Sent: Tuesday, July 21, 1998 8:01 PM
>To: Eric Thayer
>Subject: Re: what's the right parameter to use...
>Eric,  I modified the simulator to support 10 additional KL-11s beyond the
>one used for KB0:  The registers for KB1:-KB10: are in a block that is
>separate from the ones for KB0:.
>When I generated the RSTS monitor I only built it for 8 terminals, and I
>haven't had a chance to re-build it yet.  The net of this is that the
>simulator knows there is a KB9: and KB10: but RSTS does not, and won't
>respond to those keyboards as it is presently built.
>Regards, Gary
>-----Original Message-----
>From: Eric Thayer <Eric_Thayer@alf19.speech.cs.cmu.edu>
>To: garyll@ibm.net <garyll@ibm.net>
>Date: Tuesday, July 21, 1998 2:59 PM
>Subject: what's the right parameter to use...
>>for # KL11's for your modified simulator 10 or 11?
>>I'm having trouble getting keyboard input/output to work
>>for it.