s390/bitops: Provide optimized arch_test_bit()

Provide an optimized arch_test_bit() implementation which makes use of
flag output constraint. This generates slightly better code:

bloat-o-meter:
add/remove: 51/19 grow/shrink: 450/2444 up/down: 25198/-49136 (-23938)

Acked-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>