Ykkrosh
d252e245ef
Optimise fixed multiplies on MSVC x86
...
On MSVC, casting i32 to i64 then multiplying gets compiled into a call
to __allmul, which is slow. Use the __emul intrinsic instead.
GCC 4.6+ appears optimise this case automatically and doesn't need any
special handling.
This reduces the cost of ComputeShortPath by about 50% (testing AI vs AI
on Oasis 01).
This was SVN commit r13873.
2013-09-20 20:17:54 +00:00
..
2013-06-23 00:50:24 +00:00
2013-09-11 20:41:53 +00:00
2013-03-25 05:01:36 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-15 14:03:53 +00:00
2013-09-11 20:41:53 +00:00
2012-08-08 05:18:20 +00:00
2013-02-22 23:20:59 +00:00
2013-07-07 22:44:47 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-20 20:17:54 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-10 14:17:04 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-09-11 20:41:53 +00:00
2013-08-18 09:27:11 +00:00
2013-03-05 22:52:48 +00:00
2013-03-05 22:52:48 +00:00
2013-09-10 14:17:04 +00:00
2013-03-13 20:44:48 +00:00
2013-07-07 22:44:47 +00:00
2013-07-07 22:44:47 +00:00
2012-08-09 19:06:01 +00:00
2012-08-08 05:18:20 +00:00
2012-08-08 05:18:20 +00:00
2013-02-22 23:20:59 +00:00
2013-02-22 23:20:59 +00:00
2013-07-07 22:44:47 +00:00
2013-03-23 17:59:54 +00:00
2013-07-16 03:46:30 +00:00
2012-05-19 23:07:41 +00:00
2012-05-19 23:07:41 +00:00
2012-05-24 18:25:31 +00:00
2012-05-24 18:25:31 +00:00
2013-08-03 19:20:20 +00:00
2013-09-11 20:41:53 +00:00
2012-07-19 04:45:01 +00:00
2012-07-30 21:06:54 +00:00
2013-03-05 22:52:48 +00:00
2013-03-05 22:52:48 +00:00
2013-07-17 05:42:16 +00:00
2012-12-06 19:46:13 +00:00
2012-12-06 19:46:13 +00:00
2013-02-03 02:08:20 +00:00
2013-02-03 02:08:20 +00:00