[GitHub] [incubator-nuttx] patacongo commented on pull request #1726: nRF52: extend systimer support

[GitBox <gi...@apache.org>]

patacongo commented on pull request #1726:
URL: https://github.com/apache/incubator-nuttx/pull/1726#issuecomment-687900378


   
   > @patacongo There is something I'm unsure about: as you can see, at the end of `up_timer_initialize()` I called `nxsched_alarm_expiration()`. If I do not do this, the 200ms scheduling alarm will not start until I for example do "sleep 1" in the console. This means that until I do so, the scheduler thinks no time has passed and thus gives a wrong value for the first alarm to up_alarm_start(). I have observed this behavior previously (and I think it is not specific to TICKLESS_ALARM) where I see the 200ms timer only starting after an initial task is launched. I'm confused how this should work. In fact, is it even necessary to keep the 200ms scheduling alarm ON even once we go back to having just one task?
   
   Well, that k,ind of makes sense.  In Tickless mode, the OS sets the interval timer, then waits for the timer expiration which calls  `nxsched_alarm_expiration()` to process the timer expiration.  Since it does need an event to get started.  But I don't recall having done that with other architectures.
   
   Let me look at STM32.
   


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org