comp.lang.idl-pvwave archive
Messages from Usenet group comp.lang.idl-pvwave, compiled by Paulo Penteado

Home » Public Forums » archive » SGI 6.5 and CALL_EXT Problem
Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend 
Switch to threaded view of this topic Create a new topic Submit Reply
SGI 6.5 and CALL_EXT Problem [message #15641] Thu, 03 June 1999 00:00 Go to next message
rmlongfield is currently offline  rmlongfield
Messages: 68
Registered: August 1998
Member
Hi all,
I need help with something that used to work and now doesn't.
My SGI O2 has just been upgraded to 6.5. Prior to this I have used
CALL_EXTERNAL successfully with C and Fortran programs.

The line in my code is:--------------------------------

result = CALL_EXTERNAL('idl_calc_pos.so', 'idl_calc_pos',
yydddhhmm_double)

--------------------------------------------------------
Here is the error message:

% CALL_EXTERNAL: Error loading sharable executable.
Symbol: idl_calc_pos, File = idl_calc_pos.so
7509:/d/rsi/idl_5.2/bin/bin.sgi/idl: rld: Fatal Error:
Cannot Successfully map soname
'idl_calc_pos.so' under any of the filenames

./idl_calc_pos.so:/d/rsi/idl_5.2/bin/bin.sgi/idl_calc_pos.so :/usr/lib/idl_calc_pos.so:/usr

/lib/internal/idl_calc_pos.so:/lib/idl_calc_pos.so:/lib/cmpl rs/cc/idl_calc_pos.so:/usr/lib
/cmplrs/cc/idl_calc_pos.so:/opt/lib/idl_calc_pos.so:
% Execution halted at: IDL_CALC_POS 31


I am also having trouble with the new f77 compiler, independent of IDL
and maybe this is related somehow.
Hope someone can help!

Rose



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
Re: SGI 6.5 and CALL_EXT Problem [message #15687 is a reply to message #15641] Mon, 07 June 1999 00:00 Go to previous messageGo to next message
rmlongfield is currently offline  rmlongfield
Messages: 68
Registered: August 1998
Member
In article <3757A4B0.86F86BAE@ion.le.ac.uk>,
Nigel Wade <nmw@ion.le.ac.uk> wrote:
> There are various reasons why a DSO failed to load.
...(see previous thread)

Thanks Nigel, for the information. I was told today that eventually
IRIX will no longer use o32. I hope IDL will be prepared to deal with
that when it happens.

Rose


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
Re: SGI 6.5 and CALL_EXT Problem [message #15716 is a reply to message #15641] Fri, 04 June 1999 00:00 Go to previous messageGo to next message
Nigel Wade is currently offline  Nigel Wade
Messages: 286
Registered: March 1998
Senior Member
rmlongfield@my-deja.com wrote:
>
> Hi all,
> I need help with something that used to work and now doesn't.
> My SGI O2 has just been upgraded to 6.5. Prior to this I have used
> CALL_EXTERNAL successfully with C and Fortran programs.
>

There are various reasons why a DSO failed to load. The simplest is that
IDL
really couldn't find it, either because it is incorrectly named or
because it
isn't on the run-time library search path. Another reason is that IDL
requires
the DSO to be built using the O32 ABI under IRIX. A third reason is that
it may have
undefined externals if you link against library which has changed
between OS
releases.

You can check for the first by copying the DSO to the directory where
you are
running IDL from, the current directory should be searched for DSOs.

You can check for the second reason by using the 'file' command:
% file idl_calc_pos.so
will tell you what type of library has been created.
IDL needs mips-2 dynamic libs. If it is a mips-3 dynamic lib it was
built using
the N32 ABI and will need to be re-compiled.

If the compiler defaults have changed this could happen. See the man
page for the
compiler and ABI(5) for more details. If the installation has changed
the compiler
defaults to N32 then you will need to compile and link with the -o32
option
added to the command line.

Finally, to check for undefined externals use the nm command. There will
be many
instances (e.g. any calls to routines in the C library). But if you
eliminate these
you should be able to find which external is not satisfied.

--
-----------------------------------------------------------
Nigel Wade, System Administrator, Space Plasma Physics Group,
University of Leicester, Leicester, LE1 7RH, UK
E-mail : nmw@ion.le.ac.uk
Phone : +44 (0)116 2523568, Fax : +44 (0)116 2523555
Re: SGI 6.5 and CALL_EXT Problem [message #15719 is a reply to message #15641] Fri, 04 June 1999 00:00 Go to previous messageGo to next message
rmlongfield is currently offline  rmlongfield
Messages: 68
Registered: August 1998
Member
In article <3756D843.B9F2646A@ssec.wisc.edu>,
Liam Gumley <Liam.Gumley@ssec.wisc.edu> wrote:

> Did you recompile your shared object library under Irix 6.5? If not,
I'd
> try that first. Also, make sure your FORTRAN compiler was upgraded to
> the right version for Irix 6.5.

Hi Liam (and anyone else who might encounter this problem),

Yes, I recompiled everything and still no go. However, I wandered over
to your Web page and had a look at some of your SGI FAQ guides. I was
just scrolling down the page and I spotted in an error discussion the
line:

(-32 may be omitted on non Irix 6 systems) AHA!

Well, yesterday, when I was frantically searching around for a solution,
I found that my new system is n32. I tried returning to o32 but did not
know that the compiling AND ld commands had to have this -32 flag.
This I saw on the FAQ.

Everything seems to work now, except some #84 warnings:
definition of main preempts that in /usr/lib/libftn.so
I hope they aren't serious.

Thanks !
Rose



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
Re: SGI 6.5 and CALL_EXT Problem [message #15727 is a reply to message #15641] Thu, 03 June 1999 00:00 Go to previous messageGo to next message
Liam Gumley is currently offline  Liam Gumley
Messages: 473
Registered: November 1994
Senior Member
rmlongfield@my-deja.com wrote:
> I need help with something that used to work and now doesn't.
> My SGI O2 has just been upgraded to 6.5. Prior to this I have used
> CALL_EXTERNAL successfully with C and Fortran programs.
[error messages removed]
> I am also having trouble with the new f77 compiler, independent of IDL
> and maybe this is related somehow.

Rose,

Did you recompile your shared object library under Irix 6.5? If not, I'd
try that first. Also, make sure your FORTRAN compiler was upgraded to
the right version for Irix 6.5.

Cheers,
Liam.

--
Liam E. Gumley
Space Science and Engineering Center, UW-Madison
http://cimss.ssec.wisc.edu/~gumley
Re: SGI 6.5 and CALL_EXT Problem [message #15729 is a reply to message #15641] Thu, 10 June 1999 00:00 Go to previous messageGo to next message
Vapuser is currently offline  Vapuser
Messages: 63
Registered: November 1998
Member
rmlongfield@my-deja.com writes:



> In article <3757A4B0.86F86BAE@ion.le.ac.uk>,
> Nigel Wade <nmw@ion.le.ac.uk> wrote:
>> There are various reasons why a DSO failed to load.
> ...(see previous thread)
>
> Thanks Nigel, for the information. I was told today that eventually
> IRIX will no longer use o32.

Do you know the time frame for this? There are linkimage routines that I
CANNOT do without. I certainly don't want to program them in IDL.

> I hope IDL will be prepared to deal with
> that when it happens.
>
> Rose
>
>

Me too, or there'll be violence! Or, at least a lot of screaming.
You'd better be listening RSI!


William

--
William Daffer: 818-354-0161: daffer@catspaw.jpl.nasa.gov
Re: SGI 6.5 and CALL_EXT Problem [message #15785 is a reply to message #15641] Mon, 14 June 1999 00:00 Go to previous message
Nigel Wade is currently offline  Nigel Wade
Messages: 286
Registered: March 1998
Senior Member
Vapuser wrote:
>
> rmlongfield@my-deja.com writes:
>
>
>> In article <3757A4B0.86F86BAE@ion.le.ac.uk>,
>> Nigel Wade <nmw@ion.le.ac.uk> wrote:
>>> There are various reasons why a DSO failed to load.
>> ...(see previous thread)
>>
>> Thanks Nigel, for the information. I was told today that eventually
>> IRIX will no longer use o32.
>
> Do you know the time frame for this? There are linkimage routines that I
> CANNOT do without. I certainly don't want to program them in IDL.
>
>> I hope IDL will be prepared to deal with
>> that when it happens.
>>
>> Rose
>>
>>
>
> Me too, or there'll be violence! Or, at least a lot of screaming.
> You'd better be listening RSI!
>
> William
>
> --
> William Daffer: 818-354-0161: daffer@catspaw.jpl.nasa.gov

I guess that eventually they will have to switch to n32, I don't suppose
SGI will
provide o32 forever!

I asked this very question today and was told that the next release
(5.3) will be
in o32 only. n32 or 64 would be classed as a different operating system
requiring
an effective port and complete verification.

This is really beginning to bite us here as people convert to n32 and
have to keep
two copies of libraries to be able to link to their n32 programs and the
o32 IDL.

The FORTRAN90 compiler won't produce o32 code...

--
-----------------------------------------------------------
Nigel Wade, System Administrator, Space Plasma Physics Group,
University of Leicester, Leicester, LE1 7RH, UK
E-mail : nmw@ion.le.ac.uk
Phone : +44 (0)116 2523568, Fax : +44 (0)116 2523555
Re: SGI 6.5 and CALL_EXT Problem [message #15823 is a reply to message #15641] Fri, 11 June 1999 00:00 Go to previous message
rmlongfield is currently offline  rmlongfield
Messages: 68
Registered: August 1998
Member
Hi William,
I do not know when the change will be made. I just happened to be
chatting with someone from SGI who was fixing my machine (with different
problems). I told him that some of my colleagues had libraries in o32
which they did not want to re-compile. This isn't only IDL, it is C and
Fortran compiler stuff.
I agree, it is a bit scary. I worked hard to get the C programs to
work from IDL.
Rose



Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
  Switch to threaded view of this topic Create a new topic Submit Reply
Previous Topic: Re: ascii templates & also IDL from emacs
Next Topic: I' want "Hough Transform" idl source..please..

-=] Back to Top [=-
[ Syndicate this forum (XML) ] [ RSS ] [ PDF ]

Current Time: Fri Oct 10 06:46:29 PDT 2025

Total time taken to generate the page: 2.39568 seconds