1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2025-02-04 06:08:26 +02:00
FFmpeg/libavfilter
Clément Bœsch 43d16aef63 lavfi/nlmeans: make compute_safe_ssd_integral_image_c faster
before:  ssd_integral_image_c: 49204.6
after:   ssd_integral_image_c: 44272.8

Unrolling by 4 made the biggest difference on odroid-c2 (aarch64);
unrolling by 2 or 8 both raised 46k cycles vs 44k for 4.

Additionally, this is a much better reference when writing SIMD (SIMD
vectorization will just target 16 instead of 4).
2018-05-08 10:28:06 +02:00
..
2017-05-12 10:48:13 +02:00
2017-11-26 14:05:25 +01:00
2018-05-04 14:24:56 +00:00
2018-03-31 23:26:30 +01:00
2018-03-31 23:26:30 +01:00
2018-05-05 21:50:22 +02:00
2017-06-26 19:43:41 +02:00
2017-07-08 11:49:54 +02:00
2018-05-04 14:24:56 +00:00
2018-04-29 13:24:50 +02:00
2017-12-27 11:54:22 +01:00
2018-04-21 11:34:49 +02:00
2018-01-03 19:45:01 +01:00
2017-08-05 21:05:22 +02:00
2018-03-08 11:23:34 +01:00
2018-01-05 03:35:48 +01:00
2018-05-04 23:17:53 +02:00
2017-11-25 09:51:33 +01:00
2017-12-06 15:42:23 +01:00
2018-05-04 14:24:56 +00:00
2017-08-15 15:27:42 +02:00
2018-04-13 08:41:51 +02:00
2017-12-23 10:48:14 +01:00