Gael Guennebaud e6b217b8dd bug #1652: implements a much more accurate version of vectorized sin/cos. This new version achieve same speed for SSE/AVX, and is slightly faster with FMA. Guarantees are as follows:
- no FMA: 1ULP up to 3pi, 2ULP up to sin(25966) and cos(18838), fallback to std::sin/cos for larger inputs
  - FMA: 1ULP up to sin(117435.992) and cos(71476.0625), fallback to std::sin/cos for larger inputs
2019-01-09 15:25:17 +01:00
..
2018-08-01 21:34:47 -04:00
2016-05-18 14:03:03 +02:00
2018-11-23 15:37:09 +01:00
2018-11-08 18:40:22 +01:00
2018-12-08 00:05:03 +01:00
2018-10-07 23:29:51 +02:00
2015-08-04 16:12:16 +02:00
2018-08-01 21:34:47 -04:00
2018-11-23 15:12:06 +01:00
2018-10-10 08:23:40 +02:00