[csw-maintainers] RFC: idea to unify 32/64 bit headers

Dagobert Michelsen dam at opencsw.org
Fri Sep 9 09:36:28 CEST 2011


Hi Maciej,

Am 09.09.2011 um 09:31 schrieb Maciej Bliziński:
>>> I'm definitely for it.  Our libgmp build can serve as a testbed.  I'd
>>> be also willing to write some code to facilitate this.  How would you
>>> envision the implementation, Dago?
>> 
>> I just tried this out in the shell..
>> 
>> igalic at unstable10x ...ar/pkg/hwloc/trunk/work/solaris9-i386 % for f in $( find install-isa-i386/opt/csw/include/ -type f ) ; do
>> x86="${f}"
>> amd64="${f/install-isa-i386/install-isa-amd64}"
>> diff -D __amd64 "${x86}" "${amd64}"
>> done
>> 
>> And found some rather disturbing things
>> 
>> /* Defined to 1 if you have the <stdint.h> header file. */
>> #ifndef __amd64
>> /* #  undef HWLOC_HAVE_STDINT_H */
>> #else /* __amd64 */
>> #  define HWLOC_HAVE_STDINT_H 1
>> #endif /* __amd64 */
> 
> That might be related to Solaris Studio magic.

What is the problem with that? From my understanding the above headers should have never
been made it to the installed files. Apart from that it looks good to me. However, I use
a slightly different thing to accomodate for a more complex conditional:
  http://sourceforge.net/apps/trac/gar/browser/csw/mgar/pkg/libcurl4/trunk/Makefile#L157
In a GAR built-in definition I would separate the amd64 and sparcv9 conditionals to the
respective GARCH of course.


Best regards

  -- Dago

-- 
"You don't become great by trying to be great, you become great by wanting to do something,
and then doing it so hard that you become great in the process." - xkcd #896



More information about the maintainers mailing list