Commit Graph

6 Commits (master)

Author SHA1 Message Date
Miroslav Lichvar 5d6c3433f7 linreg: use sample weight.
Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
2015-03-27 10:58:04 +01:00
Miroslav Lichvar f0b0c1116a servo: add support for weighted samples.
Add weight parameter to the sample function. Samples with smaller weight
are less reliable, they can be ignored by the servo or the adjustments
of the clock can be smaller.

Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
2015-03-27 10:58:04 +01:00
Miroslav Lichvar 37459fd1ad linreg: fix servo resetting
The stats for the maximum size were not reset, which caused the
the servo to reuse old data instead of returning with unlocked
state.

Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
2014-11-20 20:15:34 +01:00
Miroslav Lichvar 97f72ceecc linreg: Handle leap second gracefully.
When leap second is inserted/deleted to the reference time, adjust the
reference point as if the clock was stepped in the opposite direction to
keep the slope and error statistics valid and correct the offset
gracefully.

Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
2014-06-22 09:30:31 +02:00
Miroslav Lichvar 61de819d56 Include clock rate ratio in delay calculation.
With the new linreg servo the frequency offset and time offset are
controlled separately. The ratio between master's frequency and the
current frequency of the local clock is known and can be used when
calculating delay or peer delay to improve their accuracy.

This greatly improves the stability of the delay when the servo is
correcting a large offset.

Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
2014-03-14 09:52:40 +01:00
Miroslav Lichvar a5890cbe98 Add an adaptive servo based on linear regression.
This servo uses linear regression to estimate current time and
frequency error. The number of points used in the regression is
variable (from 4 to 64 in powers of 2) and is selected by a long-term
statistic of the prediction error.

Future improvements could include tracking of sudden frequency changes
(e.g. due to temperature variations), better stability of the error
statistic when a large offset is corrected, options to set the speed of
the adaptation, minimum and maximum number of points, or an option to
prefer frequency accuracy over time accuracy.

Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
2014-03-14 09:52:40 +01:00