help-octave
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: error: no matching function with octave-3.2.4/46bit/gcc-4.3.3


From: Loris Bennett
Subject: Re: error: no matching function with octave-3.2.4/46bit/gcc-4.3.3
Date: Mon, 12 Jul 2010 16:51:20 +0200
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/23.1 (gnu/linux)

"John W. Eaton" <address@hidden> writes:

> On 12-Jul-2010, Loris Bennett wrote:
>
> | Jordi Gutiérrez Hermoso <address@hidden> writes:
> | 
> | > I'll give WAG of what's going on.
> | >
> | > gcc 4.3 *should* have trigged the CXX_ISO_COMPLIANT_LIBRARY macro
> | > during the run of the configure script. For whatever reason, it
> | > didn't. You can try manually modifying config.h and define this macro,
> | > although the wiser solution would be to investigate why it wasn't
> | > defined.
> | 
> | Well WAGged. Tweaking config.h did the trick.
> | 
> | However, doing a fresh configure showed that CXX_ISO_COMPLIANT_LIBRARY
> | is in princple set correctly. I suspect that the endless other tweaking
> | I have been doing to get Octave to compile on AIX had got me into some
> | strange configuration mess.
>
> What tweaking do you need?  We would like to make Octave build out of
> the box on most current systems if possible.  But I don't think any of
> the current developers uses AIX regularly if at all.  So if the people
> who do use AIX (or other uncommon systems) don't report specifically
> the problems they have, it's unlikely that the problems will be fixed.
>
> jwe

The main problem is working out which flags the AIX loader requires,
e.g. -Wl,-brtl to allow libraries with the .so suffix to be found, or
-Wl,-bbigtoc to increase TOC space. The process is rather labourious as
the whole configure-make cycle takes such a long time. Thus, if part of
the make fails at some point and I find that some flag is missing, I
sometimes just rerun the command which failed and then run make
again. Another problem was that setting the option -Wl,-G to avoid a C++
problem caused the next configure run to produce completely spurious
results and subsequently a whole lot of other compilation errors.

Of course, I am solely to blame for shooting myself in the foot by
trying to cut corners and save time. However, I would like to find the
correct environment settings that would allow make to run without error
and would be happy to document this on the Octave wiki.

Thanks for the help and interest

Loris

-- 
Dr. Loris Bennett
ZEDAT Computer Centre
Freie Universität Berlin
Berlin, Germany



reply via email to

[Prev in Thread] Current Thread [Next in Thread]