MMX and SIMD

[llewelly@xmission.com]

"William M. Quarles" <walrus@bellsouth.net> writes:

> Are MMX and SIMD instruction capabilities included in GCC now?

Yes. There are both intrinsics and also optimization flags to enable
    automatic generation of them.

See http://xrl.us/b43h and search for -mfpmath and -mmmx for
    optimization flags, http://xrl.us/b43i for vector intrinsics, and
    http://xrl.us/b43k for platform specific x86 builtins, including
    mmx and sse[2] builtins.