On July 7, 2015 4:18:51 PM CDT, Jan Sommer <soja-li...@aries.uberspace.de> wrote: >Am Dienstag, 7. Juli 2015, 16:06:11 schrieb Joel Sherrill: >> Thanks! >> >> That applied perfectly and I just pushed it. >> >> Hopefully it corrected the anomalies you saw. >> > >Forgot to mention it. I tested it with the example from earlier and the >delays worked fine then. > >> FWIW I just noticed the Pi clock driver does not have support >> for getting the nanoseconds since the last tick so timestamps >> will have the granularity of the clock tick. >> > >Well, for me it's less about nanoseconds accuracy, it's just what the >Ada runtimes seems to use.
It simply wasn't honoring the user's configuration request. Thanks for fixing it. >> On 7/7/2015 3:52 PM, Jan Sommer wrote: >> > Hi, >> > >> > I hope that's how this works. Here is a patch for the raspberrypi >clock driver to use the correct value for the timer counter based on >rtems_configuration_get_microseconds_per_tick() as discussed in the >user-list. >> > >> > Cheers, >> > >> > Jan >> > ------------------------------------------------------------- >> > >> >> --joel _______________________________________________ devel mailing list devel@rtems.org http://lists.rtems.org/mailman/listinfo/devel
