On Thu, Jan 02, 2020 at 07:42:11AM -0800, Richard Cochran wrote: > On Thu, Jan 02, 2020 at 04:00:48PM +0100, Miroslav Lichvar wrote: > > learned to not trust it and test the code both ways to make sure it's > > correct, which I remember quite clearly doing with this patch. I'm not > > sure what went wrong. > > The severity of this bug is not large when the frequency offsets are > small, as is usually the case.
I think the servos should be able to recover even with very large frequency offsets (e.g. when correcting a large phase offset with disabled steps), but the clockcheck may prevent the servo from correcting the clock if the measured frequency is larger than the frequency limit. In the default configuration that could happen with PHCs with maximum adjustment of about 20% or more. -- Miroslav Lichvar _______________________________________________ Linuxptp-devel mailing list Linuxptp-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
