Micro-optimize pop_lsb() for 64bit case
authorMarco Costalba <mcostalba@gmail.com>
Thu, 1 Nov 2012 08:30:03 +0000 (09:30 +0100)
committerMarco Costalba <mcostalba@gmail.com>
Fri, 2 Nov 2012 11:11:49 +0000 (12:11 +0100)
commit94ecdef8ac7855e7a44efd7890c7e8f8d5819397
treeded5515e90693e3cb5d44ca9cfc0bd7cb589ef79
parente3b03278129e4d7f130826aef209668b3268aa94
Micro-optimize pop_lsb() for 64bit case

On Intel, perhaps due to 'lea' instruction this way of
zeroing the lsb of *b seems faster than a shift+negate.

On perft (where any speed difference is magnified) I
got a 6% speed up on my Intel i5 64bit.

Suggested by Hongzhi Cheng.

No functional change.
src/bitboard.h