Hi,
We are using V10.4.3.
Your idea of insert sanity check for limiting the xTicksToJump to xExpectedIdleTime - 1 is a great idea - minimum code changes and only in my hook for vPortSuppressTicksAndSleep.
I will test it.
Hi,
Just want to keep you updated 
First, I discovered your code at No-Drift FreeRTOS tick/tickless for STM32 via LPTIM · GitHub
where the code comments where very helpful… althoght my implementation is quite different. e.g. I’m using the ARRM interrupt and not the CMP, the CMP is initialized with ARR value-1 and using to check if CNT was wrap around and I’m using the ST HAL LPTIMER functions, so the silicon bugs handled there.
Secondly, since time drift was very large, I’m now recognizing an interrupt burst, exit sleep mode (portSUPPRESS_TICKS_AND_SLEEP is skipped), and then returning after X quiet mSec.
My systick time drift now is less then 2% (30 seconds in 30 minutes) which is greate.
Thanks for your support,
Dagan
1 Like
Thank you for reporting back.