Subject: Re: random signals kill my processes with NetBSD-1.2
To: None <port-sparc@NetBSD.ORG>
From: Mike Frisch <mfrisch@saturn.tlug.org>
List: port-sparc
Date: 02/03/1997 14:35:31
On Mon, 3 Feb 1997, Paul Boven wrote:

> machine completely hangs during the linking of octave (wich uses about 
> 70M of VM to link). I wonder if this is a VM-problem?

When I was netbooting my Sun 4/110 from my Linux NFS server, I had thought
that the swap over NFS code was broken when NetBSD/sparc exhibited this
SIG11 behaviour.  When I added a disk to the Sun 4/110, I was surprised to
see that the same bevahiour was still there.  I personally don't think
it's a swap problem because in daily operations of my, somewhat memory
impaired, Sun 4/110, I've seen the swap file grow to 20MB and shrink
again.  This functionality appears to be working correctly. 

I found that the SIG11s were happening mostly when compiling bash and
other applications of that magnitude.  Compiling such apps as ncftp, pine,
etc all went off without a hitch.

> > After a core dump I can restart the make and
> > it will continue no problem (i.e. the core dump
> > is random).
> Yup, same here.

I'll third this, but sometimes it took several retries to finally get it
to finish.

Mike.

======================================================================
Mike Frisch	                        Email: mfrisch@saturn.tlug.org
Northstar Technologies       WWW: http://a19b89.rogerswave.ca/~mfrisch
Newmarket, Ontario, Canada