fix16_div bug due to __builtin_clzl about libfixmath HOT 4 CLOSED

Hmm, isn't this a duplicate of issue #17, which is already fixed? The macro 
should be based on sizeof(long) so that bitcount is taken into account:
https://code.google.com/p/libfixmath/source/browse/trunk/libfixmath/fix16.c#278

Running the attached test.cc on my computer:

petteri@oddish:~/libfixmath/unittests$ uname -a
Linux oddish 3.11.0-12-generic #19-Ubuntu SMP Wed Oct 9 16:20:46 UTC 2013 
x86_64 x86_64 x86_64 GNU/Linux
petteri@oddish:~/libfixmath/unittests$ g++ -I ../libfixmath -o test test.cc 
../libfixmath/fix16.c
petteri@oddish:~/libfixmath/unittests$ ./test
sizeof(int) = 4
sizeof(long) = 8
9.09087

Ah, just saw the revision number r64. Upgrade to newest svn and it should work.

flatmush: maybe we should put a new release on the downloads tab some time? :)

I would second that suggestion. If this is a known bug that has been fixed then 
maybe the default suggested download shouldn't be a version of the library 
which still contains this bug.

flatmush: I took the liberty of marking the downloads as deprecated. Google 
Code will discontinue the supports for downloads soon anyway, so maybe it is 
better to only have svn.