Dear All,

Let me say I'm very enthusiastic about what SystemView potentially has to offer, if it wasn't for an annoying bug.

First let me clarify some circumstances:

I'm using a STM32G0 microprocessor in a project I'm currently working on.
To make development easier I decided to implement an alternative RTOS which I'm more familiar with github.com/weston-embedded/uC-OS3
Because it's a rather complex and time critical application I also decided to implement a trace/viewer segger.com/downloads/systemview/systemview_target_src

Reviewed all of the documentation involving the porting of code to the platform and was relatively successful, except for one very annoying thing.



Somehow the timeline/timing information is incorrect and I cannot get around it.
SysTick is follow the programmed timer interval but as actual runtime not as interval ?

My first approach to debug this issue is to check if all timers and clocks have been configures correctly, by toggling a pin based on the system tick intervals and can confirm it's spot on correct.
Next checked if the correct frequency information are supplied to SystemView at startup, which reading from the System window are correctly supplied.

So last few days I started reviewing supplied code, moreover the trace functionality implementation, but that looks OK (although the Weston Embedded "maintained original" code is very different from the UCOS3 version Silabs is currently using)?

So I'm lost and I'm open for suggestions from fellow coders that have come across the same issue or are able to provide some next clue's.

Hello Johannes,

Can confirm your statement

"A huge gap/jump in the timestamps happens when there is an overflow in the timestamp delta, i.e. the previous event had a higher timestamp than the current one"

is the root cause of the problem, more over the SEGGER_SYSVIEW_X_GetTimestamp() was reading values from Cortex M3 style registers and not M0+

So implemented my own routine et voila issue solved > thanks for the tip!