Hello, Aug 6, 2019, 11:07 AM by [email protected] <mailto:[email protected]>:
> Hi! > > My old machinekit installation runs on a BBB under debian jessie with a > xenomai kernel. > I start the axis gui via an SSH session using indirect OpenGL rendering. > For normal CAM jobs this is reasonable fast and the gui is responsive. > Only huge jobs with e.g. 3D cam operations take some time to load but then > they are fine as well. > I want to switch to a more recent version because I can not get the XHC-HB04 > pendant working under jessie > (see https://github.com/machinekit/machinekit-hal/issues/209) > > So I configured a second BBB using debian stretch. > I tested various kernels but all seem to be way slower than the old kernel > coming with my debian jessie image. > Some kernels export the wrong cpu frequency settings and do not use the > maximum clock but even those who do are still slow. > Opening a very basic CAM jobs in axis loads for a minute or more. > > Something seems wrong... When I open a second console via ssh > while axis is running the console is laggy and unresponsive as well. > > Any hints? > I am not completely sure if you are talking about the latency of HAL instance or the general responsiveness of Linux OS. However, to use Debian Buster/Stretch with modern kernel version, you pretty much have to use Preempt_RT or Vanilla. The Xenomai currently in Machinekit is EOLed 2.6.something version running in tandem with 3.something Linux kernel. To use 4.19.something or 5.2.something kernel, you would need the Xenomai3 or EVL Core, for which support is currently not finished. Cern. -- website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit --- You received this message because you are subscribed to the Google Groups "Machinekit" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/machinekit/Llbidjv--3-1%40tuta.io.
