[893] in arla-drinkers

home help back first fref pref prev next nref lref last post

Two problems with arla 0.25

daemon@ATHENA.MIT.EDU (Rainer Schoepf)
Tue Jun 22 21:37:43 1999

From owner-arla-drinkers@stacken.kth.se Wed Jun 23 01:37:42 1999
Return-Path: <owner-arla-drinkers@stacken.kth.se>
Delivered-To: arla-drinkers-mtg@bloom-picayune.mit.edu
Received: (qmail 28046 invoked from network); 23 Jun 1999 01:37:42 -0000
Received: from unknown (HELO sundance.stacken.kth.se) (130.237.234.41)
  by bloom-picayune.mit.edu with SMTP; 23 Jun 1999 01:37:42 -0000
Received: (from majordom@localhost)
	by sundance.stacken.kth.se (8.8.8/8.8.8) id DAA13989
	for arla-drinkers-list; Wed, 23 Jun 1999 03:30:27 +0200 (MET DST)
Received: from assaris.sics.se (assaris.sics.se [193.10.66.108])
	by sundance.stacken.kth.se (8.8.8/8.8.8) with ESMTP id DAA13984
	for <arla-drinkers@stacken.kth.se>; Wed, 23 Jun 1999 03:30:21 +0200 (MET DST)
Received: (from assar@localhost) by assaris.sics.se (8.9.3/8.7.3) id DAA66175; Wed, 23 Jun 1999 03:30:40 +0200 (CEST)
From: Rainer Schoepf <schoepf@uni-mainz.de>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="mpGOjiqzet"
Content-Transfer-Encoding: 8bit
Message-ID: <14174.24707.683955.762702@perdita.zdv.Uni-Mainz.de>
Date: Wed, 9 Jun 1999 14:39:31 +0200 (MEST)
To: arla-drinkers@stacken.kth.se
CC: schoepf@uni-mainz.de
Subject: Two problems with arla 0.25
X-Mailer: VM 6.62 under Emacs 19.34.1
Organization: Johannes Gutenberg-Universitaet Mainz
Lines: 102
Sender: owner-arla-drinkers@stacken.kth.se
Precedence: bulk


--mpGOjiqzet
Content-Type: text/plain; charset=iso-8859-1
Content-Description: message body text
Content-Transfer-Encoding: 8bit

[moderator: I have cut down a bit on the attached files...
 thanks for submitting all the traces, but I'm not sure everyone on
 the list are that interested in those.
 I guess the best thing would be to make the files available somewhere
 and send pointers to them, in that case, the people interested (like
 me) can go and fetch them.

 /assar
]


Hello,

I'm experimenting with Arla on Solaris7/X86. It works quite well, but
has some problems.

1. The build process fails because the name of the xfs module in
xfs/solaris is the same as the name of the include directory. That's
easy to solve, I changed the module name in the Makefile to
xfs_module.

2. I see a strange behaviour when a directory is created, populated
and deleted immediately afterwards. This happened while I was trying
to build egcs. I boiled it down to the following sequence of commands:

mkdir tmpcopy
if [ xlibgcc1.null != x ];                      \
then (cd tmpcopy; ar x ../libgcc1.null);        \
else true;                                      \
fi
(cd tmpcopy; chmod +w * > /dev/null 2>&1)
(cd tmpcopy; ar x ../libgcc2.a)
(cd tmpcopy; ar rc ../tmplibgcc.a *.o)
rm -rf tmpcopy
sleep 3
rm -rf tmpcopy

The first rm -rf comes back with:
rm: Unable to remove tmpcopy: File exists

The second rm -rf works.

The appended system call trace shows that in the first case, reading
the directory does not give all filenames.

I'm afraid I don't know what's going on. I append the debugging output
of xfs and arlad.

Thanks a lot for your efforts.

Rainer Schöpf


--mpGOjiqzet
Content-Type: text/plain
Content-Description: System call trace with truss -a -vgetdents64 -f
Content-Disposition: inline;
	filename="hugo1"
Content-Transfer-Encoding: 7bit

12019:	execve("/tmp/test-bug", 0x08047C04, 0x08047C10)  argc = 2
12019:	 argv: /bin/sh /tmp/test-bug

[ -- cut -- ]


--mpGOjiqzet
Content-Type: text/plain
Content-Description: Syslog debugging output of xfs (all) and arlad (almost-all)
Content-Disposition: inline;
	filename="messages"
Content-Transfer-Encoding: 7bit

Jun  9 14:21:22 harvest.zdv.uni-mainz.de arla[345]: worker 0: processing
Jun  9 14:21:22 harvest.zdv.uni-mainz.de arla[345]: Rec message: opcode = 22 (pioctl), size = 2096

[ -- cut -- ]


--mpGOjiqzet
Content-Type: text/plain; charset=iso-8859-1
Content-Description: .signature
Content-Transfer-Encoding: 8bit


-- 
   Rainer Schöpf
   Zentrum für Datenverarbeitung           A point of view can be a dangerous
    der Universität Mainz                  luxury when substituted for insight
   Anselm-Franz-von-Bentzel-Weg 12         and understanding.
   D-55099 Mainz
   Germany                                  Herbert Marshall McLuhan:
   <Schoepf@Uni-Mainz.DE>                          The Gutenberg Galaxy

--mpGOjiqzet--


home help back first fref pref prev next nref lref last post