Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site mit-eddie.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!daver@csvax.caltech.edu
From: daver@csvax.caltech.edu
Newsgroups: net.emacs
Subject: SUN2 17.36 Hardware bug?
Message-ID: <996@mit-eddie.UUCP>
Date: Fri, 17-Jan-86 19:48:13 EST
Article-I.D.: mit-eddi.996
Posted: Fri Jan 17 19:48:13 1986
Date-Received: Sun, 19-Jan-86 03:45:04 EST
Sender: daemon@mit-eddi.UUCP
Organization: MIT, Cambridge, MA
Lines: 44

From: daver@csvax.caltech.edu (David Robinson)
Here is a bug in emacs version 17.36 on SUN2's:

When loading up xemacs with temacs make dies with the following error:

Termination code 132

A system error is sent to the console:

sendsig: bad user stack pid=6181, sig=11
usp is f7ffe0, action is 4a29c, upc is 287ba

The catch to this bug is that it is MACHINE dependant and NOT OS dependant!!

We have a network of SUN's sharing a common /usr and userfiles.  It is
made up of SUN 2/120 150 170 and 160's.  Each has 2-4 megabytes of memory
and up to 16Mbyte of swap space.

temacs dies on the following machines:
120 2M physical  8M swap sky board (Multibus)
150 4M physical 16M swap sky board (Multibus)
170 4M physical 16M swap no sky board (Multibus)

temacs successfully loads xemacs on:
160 4M physical  8M swap no sky (VME bus)

They all run the same SUN 2.0 UNIX with a shared /usr,/bin,/lib directories.
The same executeable code is run on each machine (because of NFS) and
it fails or completes depending on the HARDWARE!  The completed GNU emacs
runs on ALL machines.  All machines successfully ran 17.30 and previous
versions back to 15.XX.
The config.h is standard except for m-sun2.h instead of m-vax.h.
Sources files were directly FTP'd from mit-prep.arpa across the ARPANET (twice!)

On a seperate 170 (different SUN distribution tape) it also dies.
NO documentation is found on the "sendsig" message.

I am completely stumped!  Any help on what the errors mean or fixes would
be greatly apreciated.

				-David Robinson
				daver@cit-vax.arpa
				ihnp4!cithep!cit-vax!daver