Marco Costalba
94ecdef8ac
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.
2012-11-02 12:11:49 +01:00
..
2012-10-24 15:01:39 +02:00
2012-09-30 11:35:14 +02:00
2012-10-21 11:16:21 +02:00
2012-11-02 12:11:49 +01:00
2012-10-06 13:01:44 +02:00
2012-09-22 11:19:10 +02:00
2012-09-16 10:32:57 +02:00
2012-10-28 10:17:57 +01:00
2012-10-28 10:17:57 +01:00
2012-10-21 11:16:21 +02:00
2012-10-21 09:12:02 +02:00
2012-10-21 11:16:21 +02:00
2012-08-29 19:11:44 +02:00
2012-10-16 00:13:16 -07:00
2012-10-21 11:16:21 +02:00
2012-10-21 11:16:21 +02:00
2012-10-08 11:43:47 +02:00
2012-09-04 09:38:51 +02:00
2012-09-16 14:14:55 +02:00
2012-08-19 11:01:46 +01:00
2012-10-22 10:03:53 +02:00
2012-09-09 10:24:40 +02:00
2012-10-27 13:31:23 +02:00
2012-09-14 09:57:13 +02:00
2012-10-21 11:16:21 +02:00
2012-10-21 11:16:21 +02:00
2012-10-20 10:48:26 +02:00
2012-11-02 11:43:23 +01:00
2012-10-30 20:03:35 +01:00
2012-10-21 11:16:21 +02:00
2011-12-30 13:52:16 +01:00
2012-11-01 18:45:38 +01:00
2012-10-24 15:01:39 +02:00
2012-10-24 15:01:39 +02:00
2012-10-01 10:44:04 +02:00
2012-03-31 09:16:09 +01:00
2012-03-27 14:26:58 +01:00
2012-07-08 09:36:40 +01:00
2012-03-31 19:07:11 +01:00
2012-10-22 00:38:12 +02:00
2012-10-28 11:28:14 +01:00
2012-10-13 14:49:01 +02:00
2012-08-18 12:26:22 +01:00