<< Compiling Octave 3.6.4 in CentOS 5.8 >>

[I. Camps]

Hello all,

I am trying to compile/install Octave in my system.

After the configurations step (which only report the needs for text2dvi and OpenGL libraries ), I tried to compile and get the error bellow:

compilation aborted for DLD-FUNCTIONS/gcd.cc (code 2)
make[3]: *** [DLD-FUNCTIONS/gcd.lo] Error 1
make[3]: Leaving directory `/home/icamps/temp/Octave364/octave-3.6.4/src'
make[2]: *** [all] Error 2
make[2]: Leaving directory `/home/icamps/temp/Octave364/octave-3.6.4/src'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/icamps/temp/Octave364/octave-3.6.4'
make: *** [all] Error 2

I am trying to compile with Intel compilers version 11.0.

I appreciate any help/tips.

Best regards,

Camps

_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave

[Alexander Hansen-2]

On 3/1/13 1:09 PM, I. Camps wrote:

> Hello all,
>
> I am trying to compile/install Octave in my system.
>
> After the configurations step (which only report the needs for text2dvi
> and OpenGL libraries ), I tried to compile and get the error bellow:
>
> compilation aborted for DLD-FUNCTIONS/gcd.cc (code 2)
> make[3]: *** [DLD-FUNCTIONS/gcd.lo] Error 1
> make[3]: Leaving directory `/home/icamps/temp/Octave364/octave-3.6.4/src'
> make[2]: *** [all] Error 2
> make[2]: Leaving directory `/home/icamps/temp/Octave364/octave-3.6.4/src'
> make[1]: *** [all-recursive] Error 1
> make[1]: Leaving directory `/home/icamps/temp/Octave364/octave-3.6.4'
> make: *** [all] Error 2
>
> I am trying to compile with Intel compilers version 11.0.
>
> I appreciate any help/tips.
>
> Best regards,
>
> Camps
>
>
>

You're going to need to provide more context about your error--that's a
very general error message.

Try going back a few lines to the _start_ of the error messages and post
from there.

--AKH
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave

[Jordi Gutiérrez Hermoso-2]

On 1 March 2013 15:09, I. Camps <[hidden email]> wrote:
> I am trying to compile with Intel compilers version 11.0.

We don't generally recommend using non-free compilers to compile
Octave. Try to see if the problem goes away when you compile with gcc.

- Jordi G. H.
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave

[I. Camps]

Dear Alexander & Jordi,

The output from the make command is attached.

We don't generally recommend using non-free compilers to compile
Octave. Try to see if the problem goes away when you compile with gcc.

The Intel compiler plus the MKL mathematical libraries are free for the Linux environment (without technical support).

Thank you.

Regards,

Camps

_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave

[Jordi Gutiérrez Hermoso-2]

On 4 March 2013 14:00, I. Camps <[hidden email]> wrote:
> Dear Alexander & Jordi,
>
> The output from the make command is attached.
>
>> We don't generally recommend using non-free compilers to compile
>> Octave. Try to see if the problem goes away when you compile with gcc.
>
> The Intel compiler plus the MKL mathematical libraries are free for the
> Linux environment (without technical support).

No, it's not free:

    http://www.gnu.org/philosophy/free-sw.html

- Jordi G. H.
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave

[Jordi Gutiérrez Hermoso-2]

On 4 March 2013 14:00, I. Camps <[hidden email]> wrote:
> Dear Alexander & Jordi,
>
> The output from the make command is attached.

Wow, that's quite the output.

I think the error that icc is whining about is this:

    error: no suitable conversion function from "const Complex" to
"double" exists

However, this is happening during a template instantiation. This
should be SFINAE:

    http://en.wikipedia.org/wiki/SFINAE

I don't think this is a bug in the Octave source code. It seems to me
like a bug in icc for not properly implementing SFINAE. Do you agree?

If the compiler were free, you would be free to fix it yourself.
You'll have to beg Intel to fix this instead.

- Jordi G. H.

>
>> We don't generally recommend using non-free compilers to compile
>> Octave. Try to see if the problem goes away when you compile with gcc.
>
> The Intel compiler plus the MKL mathematical libraries are free for the
> Linux environment (without technical support).
>
> Thank you.
>
> Regards,
>
> Camps
>

_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave

[Benjamin Lewis]

Le 2013-03-04 à 14:00, "I. Camps" <[hidden email]> a écrit :

> Dear Alexander & Jordi,
>
> The output from the make command is attached.
>
> We don't generally recommend using non-free compilers to compile
> Octave. Try to see if the problem goes away when you compile with gcc.
> The Intel compiler plus the MKL mathematical libraries are free for the Linux environment (without technical support).
>
> Thank you.
>
> Regards,
>
> Camps

They're not talking about "free as in beer"; there's no cost to use icc on Linux, but it is still non-free as in its license does not allow the investigation of its source nor its modification.  You are required to purchase a commercial license if you intend to sell software compiled with icc.

Regards,

Ben Lewis
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave