On Tue, 2019-08-20 at 16:38 +0530, Alexander Sylvester wrote: On Wed, 7 Aug 2019 at 19:16, Richard Cochran wrote: > > On Wed, Aug 07, 2019 at 09:30:58AM +0200, Miroslav Lichvar wrote: > > > This doesn't look right to me. Either the HW/driver is broken, or > > the > > > master is correcting a large offset too. > > > > +1 > > > > Please tell us what HW is running on the slave. > > > > Thanks, > > Richard > Hello Richard, > > At this point I cannot reveal the hardware name, but it is using > openIL OS.
To mirror what Richard said, without knowing the hardware or driver we can't really help you at this point. It is almost certain that the fault lies somewhere in the programming logic of the driver, with a small chance that the fault is in HW design. Because so much of the logic for handling this is in driver, there's nothing we can do on this list if it's not public. My last suggestion is: use a tool like phc_ctl to do individual set, get, adjust, and set frequency options. Make sure that all of those have valid semantics (i.e. that set followed by a get 5 seconds later makes sense, similar to adjust. Also make sure that a frequency adjustment translates to the equivalent increase or decrease in clock rate). Best of luck! Regards, Jake _______________________________________________ Linuxptp-users mailing list Linuxptp-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-users
