What happened is I rebuild gv on a machine with Open3d
installed, and moved the binaries over the ones where
Open3d is not installed. And now all is OK. Not sure
exactly what/where the problem is/was, but at least it
is gone, repeated process on a couple of machines, and
gv is working OK on all.
Troubleshooting:
Used: $export _RLD_ARGS="-trace -log trace.out"
ran gv and got the following output on the machine
which had problems (I just chopped off one line):
9483:/usr/local/bin/gv: /sbin/loader: Error:
Unresolved symbol in /usr/local/bin/gv: threeDClassRec
And this is the output from working machine:
9699:/usr/local/bin/gv: /sbin/loader: resolve_symbol:
threeDClassRec in /usr/shlib/libXaw3d.so for Main at
0x3fffffeb2e8
Thanks very much for your support.
ONE Educative Response might help explain the problem:
--- "Dr. Thomas.Blinn_at_Compaq.com"
<tpb_at_doctor.zk3.dec.com> wrote:
Content-Disposition: inline
I am only guessing, but I strongly suspect that the
/usr/include/X11/Xaw3d.h
header file is part of the Open3D software product.
I searched on my V4.0G
system in the standard base OS inventory (I don't
have Open3D installed as
I have a simple 2D graphics card in the system) and
I don't see evidence of
that file. You can confirm which subset is
delivering any files with the
name in question with a command like this:
grep -i Xaw3d /usr/.smdb./*.inv
My somewhat informed guess is that the missing
symbol, threeDClassRec, is
something that USED to be defined in Compaq's Open3D
programming libraries
but no longer is.
If so, you may be hosed. Look for and read the
Open3D release notes and
see if there is any explicit mention of the removal
of the interface; if
not, we may have screwed up the Open3D software for
the V4.0G release.
Sorry I can't be more definitive -- but that's my
educated guess. I wish
I could tell you to whom you should complain, but
I'm not sure who really
"owns" the Open3D programming libraries these days,
or how we would repair
a defect in the released libraries for V4.0G, if in
fact we introduced a
significant functional regression (which is what I
think you might have
uncovered).
Tom
Dr. Thomas P. Blinn + UNIX Software Group + Compaq
Computer Corporation
110 Spit Brook Road, MS ZKO3-2/W17 Nashua, New
Hampshire 03062-2698
Technology Partnership Engineering
Phone: (603) 884-0646
Internet: tpb_at_zk3.dec.com Compaq's
Easynet: alpha::tpb
ACM Member: tpblinn_at_acm.org PC_at_Home:
tom_at_felines.mv.net
ORIGINAL POST:
Upgrade our machines, some from 4.0D to 4.0G,
others
from 4.0f 4.0G. All those upgraded, so far I have
come
across one problem. Application "gv" is not
working.
Tried re-installing the application, it didn't
make a
difference. I diff 'ed the "ThreeD.h" files it
seems
to be complaining about, from one that works and
one
that doesn't, and its all the same. Moved the
binary
gv from machine where it works to the other, same
problem. On both systems, the threeD class is in
/usr/include/X11/Xaw3d.
From installation instruction on gv, it says:
3. Make sure you have the Xaw3d widget set
(*including* its header
files) installed properly (see config.Unix
again).
I will attach config.Unix file just to see what I
am
missing. A result of tracing gv is shown below.
Any changes to the location of libraries between
4.0F
to 4.0G that you are aware of?
Thanks in advance.
xxx_at_/usr1/users/mollel/gv2/gv-3.5.8/source$
/usr/local/sbin/trace gv
gv
Tracing process /proc/11166
11166:./gv: /sbin/loader: Error: Unresolved symbol
in
./gv: threeDClassRec
11166:./gv: /sbin/loader: Fatal Error: this
executable
has unresolvable symbols
PIOCWSTOP: No such file or directory
xxx_at_/usr1/users/mollel/gv2/gv-3.5.8/source$
=====
Richard F. Mollel
Systems Administrator
SAIC
Greenbelt, MD
Experience is what allows you to recognize
a mistake the second time you make it.
__________________________________________________
Do You Yahoo!?
Yahoo! Shopping - Thousands of Stores. Millions of Products.
http://shopping.yahoo.com/
Received on Tue Dec 26 2000 - 15:30:17 NZDT