From bug-octave-request at bevo dot che dot wisc dot edu Tue Nov  5 14:15:22 2002
Subject: Re: Compiling problems on HP-UX 11.11
From: "G.F.Duivesteijn" <gilbert at dutlhs7 dot lr dot tudelft dot nl>
To: "John W. Eaton" <jwe at bevo dot che dot wisc dot edu>
Cc: bug-octave at bevo dot che dot wisc dot edu
Date: Tue, 5 Nov 2002 21:12:16 +0100

Tomorrow, I'll start the compilation again and I will include all 
output from gcc and configure. Many thanks for the quick response!

Gilbert




On Tuesday, Nov 5, 2002, at 18:20 Europe/Amsterdam, John W. Eaton wrote:

> On  5-Nov-2002, G.F. Duivesteijn <gilbert at dutlhs7 dot lr dot tudelft dot nl> wrote:
>
> | Here is some information. If more information is needed, please don't
> | hesitate to ask.
> |
> | - GNU Make version 3.79.1
> | - gcc (GCC) 3.2
> | - HP-UX 11.11 on a HP J5600
> |
> |  > ./configure --prefix=/diskq/apl/octave-2.1.37
> |  > gmake
>
>> From the messages below, it looks like you are actually compiling
> 2.1.39.
>
> | ---<snip>---
>
> You snipped too early.  You need to include the command that generated
> the errors.  It's probably important what options were used.
>
> | /usr/local/include/c++/3.2/bits/stl_threads.h:115: warning: aggregate
> | has a
> |     partly bracketed initializer
> | /usr/local/include/c++/3.2/bits/stl_threads.h:115: warning: aggregate
> | has a
> |     partly bracketed initializer
> | /usr/local/include/c++/3.2/bits/stl_threads.h:115: warning: aggregate
> | has a
> |     partly bracketed initializer
> | /usr/local/include/c++/3.2/bits/stl_threads.h:115: warning: aggregate
> | has a
> |     partly bracketed initializer
>
> These messages look like bugs in the include files, but I don't recall
> seeing them with g++ 3.2 on a Debian system.
>
> | Array.h: In constructor `Array<T>::Array() [with T = int]':
> | DASPK-opts.h:18:   instantiated from here
> | Array.h:128: call of overloaded `operator new(long unsigned int)' is
> | ambiguous
> | <internal>:128: candidates are: void* operator new(long unsigned int)
> | /usr/local/include/c++/3.2/new:89:                 void* operator
> | new(long
> |     unsigned int, ...)
> | Array.h: In constructor `Array<T>::Array(T*, int) [with T = double]':
> | MArray.h:47:   instantiated from `MArray<T>::MArray(T*, int) [with T 
> =
> | double]'
> | dColVector.h:100:   instantiated from here
> | Array.h:115: call of overloaded `operator new(long unsigned int)' is
> | ambiguous
> | <internal>:115: candidates are: void* operator new(long unsigned int)
> | /usr/local/include/c++/3.2/new:89:                 void* operator
> | new(long
> |     unsigned int, ...)
>
> I don't see these problems on a Debian system when compiling with g++
> 3.2, and I can't tell why the call would be ambiguous.
>
> In my sources, the context surrounding line 128 Array.h is:
>
>     Array (void)
>       {
> 	rep = new typename Array<T>::ArrayRep ();    // line 128
>
>   #ifdef HEAVYWEIGHT_INDEXING
> 	idx = 0;
> 	max_indices = 1;
> 	idx_count = 0;
>   #endif
>       }
>
> jwe
>



-------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:  http://www.octave.org
How to fund new projects:  http://www.octave.org/funding.html
Subscription information:  http://www.octave.org/archive.html
-------------------------------------------------------------