__APPLE_CC__ defined on head of mainline

[Mike Stump]

On Saturday, January 1, 2005, at 07:22  PM, Timothy J.Wood wrote:
> On Dec 31, 2004, at 11:09 AM, Devang Patel wrote:
>
>> Try __APPLE_ALTIVEC__
>
>   Apple's compiler doesn't seem to define this macro:

Newer ones will... :-)

> % touch /tmp/foo.c; cc -faltivec -dM -E /tmp/foo.c | grep ALTIVEC

I get:

#define __ALTIVEC__ 1
#define __APPLE_ALTIVEC__ 1

:-)

> % uname -a
> Darwin tjw 7.7.0 Darwin Kernel Version 7.7.0: Sun Nov  7 16:06:51 PST 
> 2004; root:xnu/xnu-517.9.5.obj~1/RELEASE_PPC  Power Macintosh powerpc

uname -a is rather uninteresting when talking about compilers, though, 
gcc --version is useful.

> So, my search for a way to determine the correct vector syntax at 
> compile time continues.

Try:

touch /tmp/foo.c; cc -faltivec -dM -E /tmp/foo.c | grep APPLE
#define __APPLE_CC__ 4027
#define __APPLE__ 1
#define __APPLE_ALTIVEC__ 1

:-)  You can check the value of the first, in FSF land, it is just == 1.

And as to why it is useful to check the value Devang pointed out:

mrs $ touch /tmp/foo.c; cc -maltivec -dM -E /tmp/foo.c | grep APPLE
#define __APPLE_CC__ 4027
#define __APPLE__ 1

Notice how differently it behaves.  == 1 behaves one way, < C1 another 
way, and > C2 yet another....