[ghemical-devel] ghemical-1.90 build failure on 64-bit arches

Michael Banck mbanck at debian.org
Sat Dec 10 20:40:36 EST 2005


[oops, saw that I sent this one privately in error, was meant to go to
the list really]

On Sat, Oct 08, 2005 at 06:17:47PM +0300, Tommi Hassinen wrote:
> On Thu, 6 Oct 2005, Tommi Hassinen wrote:
> >this has a good idea how to solve the problem. I think it's a bit 
> >incomplete however, so I'll finish it and let you know when it's ready.
> 
> Ok, it should be ready now, and it's in CVS already.
> 
> I haven't made a separate patch since changes happened in libghemical as 
> well ; I decided to permanently store the generated atom/object 
> identifiers instead of constantly looking for them from a container (that 
> could degrade graphics performance).
> 
> The idea goes as in the original patch so that no pointers are any more 
> pushed into the OpenGL name stack, but identifiers or "keys" instead. 
> These "keys" have been registered before use, and the "key"/pointer pairs 
> are stored into a container for quick search. The "keys" are pushed into 
> the OpenGL name stack, and in selections-related operations the keys that 
> are obtained from OpenGL are used for searching a corresponding pointer 
> from the container.
> 
> According to my testing everything works as before ; please test on the 
> 64-bit platforms then...

We have another report of a similar issue, this time in src/views.cpp:

./views.cpp: In function 'void* convert_cset_to_plotting_udata(project*, i32s)':
./views.cpp:260: error: cast from 'void*' to 'i32u' loses precision

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=341798


cheers,

Michael

-- 
Michael Banck
Debian Developer
mbanck at debian.org
http://www.advogato.org/person/mbanck/diary.html


More information about the ghemical-devel mailing list