| Date Of Receipt: 12-DEC-1994 11:38:57.25
From: ALPHA::jmf "Joshua M. Friedman OSF/UNIX SDE"
To: grava odehelp
CC: buildhelp@DEC:.zko.alpha
Subj: Re: ld core dumps?
Bill, what image produced the core? (run 'file core') Also, how
much memory/swap do you have? If you can reproduce the Memory fault,
and run 'swapon -s' just after, that might be useful.
No tools were in transition, nor was the goldminos.bl4 tree.
-josh
---------
From grava Fri Dec 9 14:24:59 1994
Delivery-Date: Fri, 09 Dec 94 14:25:03 -0500
Return-Path: grava
Received: from wasted.zk3.dec.com by flambe.zk3.dec.com; (5.65/1.1.8.2/30Mar94-0502PM)
id AA20768; Fri, 9 Dec 1994 14:24:59 -0500
Received: from localhost by wasted.zk3.dec.com; (5.65/1.1.8.2/01Sep94-0829AM)
id AA20169; Fri, 9 Dec 1994 14:24:35 -0500
Message-Id: <[email protected]>
X-Mailer: exmh version 1.5 11/22/94
To: odehelp
Subject: ld core dumps?
Cc: grava
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Date: Fri, 09 Dec 94 14:24:35 -0500
From: grava
X-Mts: smtp
Hi--
I get core dumps when trying to create vmunix:
rm -f vmunix vmunix.sys
loading vmunix.sys
sh: 769 Memory fault - core dumped
*** Exit 1 (ignored)
chmod 755 a.out
chmod: a.out: No such file or directory
*** Exit 4
Stop.
*** Exit 1
Stop.
My sandbox is backed against goldminos.bl4:
<sablelab1:/big1/sandbox/minor/src/kernel> currentsb -all
minor William_Grava_minor /big1/sandbox . /big1/sandbox/minor/rc_files/lo=
cal =
/tmp_mnt/share/nsa/build/goldminos.dsk3/goldminos.bl4 =
Am I doing something wrong, or are the ODE tools in transition? I ask th=
at =
because I encountered a similar inexplicable situation in the cluster poo=
l =
this morning, where a define that was defined in an include file that was=
=
included aborted a compile saying the define was undefined. ... ;-) (The =
file =
that was giving us problems was ./kernel/src/mcconfig/mcconfig.c, which =
#include(s) <cfg.h>, which #define(s) CFG_FLAG_NOFLAGS 0, but the compil=
er =
kept telling us that CFG_FLAG_NOFLAGS was undefined. This cluster issue w=
as =
an annoyance that got in the way of a build, and we just forced the defin=
e to =
be zero in the source file and continued on our way, assuming the problem=
=
would go away just as mysteriously as it appeared. =
But core dumps from a load aren't something I can work around very easil=
y. =
Does this ring any bells?
thanks,
-bill (1-6177)
|