On Feb 14, 2012, at 3:10 AM, Jan Seiffert wrote: > At least for the GCC side of things i don't see it as that important, > because normally GCC detects the classic shift as a byteswap (as long > as it unsigned) and uses the right instruction, when some -Ox is > given. Ok, then I'll leave it as is and not use the built-in for Z_SOLO. Mark