John Crispin
f45132d1d4
Codel: avoid a nul rec_inv_sqrt ...
One condition before codel_Newton_step() was not good if
we never left the dropping state for a flow. As a result
rec_inv_sqrt was 0, instead of the ~0 initial value.
codel control law was then set to a very aggressive mode, dropping
many packets before reaching 'target' and recovering from this problem.
Brought over from 3.5-stable
Signed-off-by: Dave Taht <dave.taht@bufferbloat.net>
SVN-Revision: 32950
2012-08-02 13:39:50 +00:00
2012-06-05 06:57:51 +00:00
2012-06-05 06:57:51 +00:00
2012-06-05 06:57:51 +00:00
2012-06-05 06:57:51 +00:00
2012-08-02 11:25:49 +00:00
2012-07-29 13:20:51 +00:00
2012-07-26 17:06:55 +00:00
2012-07-17 16:47:57 +00:00
2012-07-25 15:50:26 +00:00
2012-07-26 15:42:18 +00:00
2012-07-31 14:27:47 +00:00
2012-07-26 17:03:08 +00:00
2012-07-25 08:01:38 +00:00
2012-06-05 06:57:51 +00:00
2012-07-25 19:36:19 +00:00
2012-05-15 10:58:52 +00:00
2012-07-26 17:03:02 +00:00
2012-05-15 10:58:52 +00:00
2012-06-05 06:57:51 +00:00
2012-08-02 13:39:50 +00:00
2012-05-15 10:58:52 +00:00
2012-05-15 10:58:52 +00:00
2012-06-05 06:57:51 +00:00
2012-07-25 15:50:23 +00:00
2012-07-15 14:25:30 +00:00
2012-08-02 13:29:49 +00:00
2012-05-15 10:58:52 +00:00
2012-06-05 06:57:51 +00:00
2012-08-01 22:09:13 +00:00
2012-06-05 06:57:51 +00:00
2012-06-05 06:57:51 +00:00
2012-06-05 06:57:51 +00:00
2012-07-25 15:50:13 +00:00
2012-06-05 06:57:51 +00:00
2012-07-29 13:20:39 +00:00
2012-05-15 10:58:52 +00:00
2012-06-05 06:57:51 +00:00
2012-07-25 15:50:45 +00:00
2012-06-05 06:57:51 +00:00
2012-05-15 10:58:52 +00:00
2012-06-05 06:57:51 +00:00
2012-04-21 15:18:54 +00:00
2012-07-25 15:50:42 +00:00
2012-06-05 06:57:51 +00:00
2012-06-16 16:45:28 +00:00
2012-07-27 15:35:15 +00:00
2012-05-15 10:58:52 +00:00
2012-06-05 06:57:51 +00:00
2012-04-29 15:00:32 +00:00
2012-05-15 10:58:52 +00:00
2012-06-05 06:57:51 +00:00
2012-07-25 19:36:19 +00:00
2012-06-20 19:04:35 +00:00
2010-08-06 21:36:35 +00:00
f45132d1d4