osal_start_reload_timer stops reloading timeout

design_software

Other Parts Discussed in Thread: CC2541

I am calling osal_start_reload_timer for an event to regularly happen, let's say every 10 seconds.

setReloadTimer( LOCATION_UPDATE, 10000 );

Unexpectedly, at some point, the event flag is not being raised.

I have a modified Keyfodemo code in a custom board.

The chip is cc2541. Power saving is disabled. Serial interface is used.

Are there any other events or setting that can trigger this issue?

over 10 years ago

0 design_software over 10 years ago

Prodigy 70 points

Sorry it should be
osal_start_reload_timer( serialInterface_TaskID, LOCATION_UPDATE, 10000 );

0 Igor Sherer over 10 years ago in reply to design_software

Guru 31280 points

Hi,

Is this issue can be easily reproduced?

For the sake of experiment, what happens if you begin with

osal_start_timerEx( serialInterface_TaskID, LOCATION_UPDATE, 10000 ); inside init()

and then every time the LOCATION_UPDATE event is invoked you start the same timer

osal_start_timerEx( serialInterface_TaskID, LOCATION_UPDATE, 10000);

Is it behaves the same as with the reload_timer?

0 design_software over 10 years ago in reply to Igor Sherer

Prodigy 70 points

Hi Igor,

Since the issue with osal_start_reload_timer happens after several hours, I will be testing my device overnight implementing osal_start_timerEx instead. I'll post the result here tomorrow.

But do you already have an idea of what is causing this?

Thanks!

Regards,
Ronnel

0 Igor Sherer over 10 years ago in reply to design_software

Guru 31280 points

well, somehow the reload_timer halts, or the Event isn't cleared properly, not sure yet, this test, hopefully, will shed some light

0 design_software over 10 years ago in reply to Igor Sherer

Prodigy 70 points

Ok. BTW, I am also having problems with osal_start_timerEx that's why I opt to use osal_start_reaload timer. I am sometimes getting NO_TIMER_AVAILABLE with the former which I think caused by too much timer. I even cancel the timer before re-starting it. And I tried this reload timer to even minimize creating of timers. Then this issue occurred.

0 Igor Sherer over 10 years ago in reply to design_software

Guru 31280 points

Do you use much of memory heap in your application?
lack of available memory for OSAL in the heap may cause such things.
When you open the timer, ensure it was really opened by OSAL.
Another thing, is it possible that you have a memory lick, which, at some point, breaks timers' normal behavior

0 design_software over 10 years ago in reply to Igor Sherer

Prodigy 70 points

Sorry but how do I know if I use much of heap memory? Is it in the logs after compiling?
My application is based on the Keyfobdemo so I assume that the heap used is almost the same.
Am I right that the heap memory is being used when calling malloc? And I'm not sure if I am calling other api functions that are calling it well aside from osal timer.

0 Igor Sherer over 10 years ago in reply to design_software

Guru 31280 points

heap is part of the ram, so if you are using large arrays, or many timers you are taking this memory from available heap,

0 design_software over 10 years ago in reply to Igor Sherer

Prodigy 70 points

Okay. There are at least 3 arrays in my application with size of 150+. But if they were successfully created and as well as the reloading timers, how come at some point they fail?

Thanks,
Ronnel

0 Igor Sherer over 10 years ago in reply to design_software

Guru 31280 points

arrays of what type? uint8, uint32, custom types (structs)?

0 design_software over 10 years ago in reply to Igor Sherer

Prodigy 70 points

char array. Upon checking there are two 150 and one 250 sized char array.

Btw, seems like the problem is in osal_start_reload_timer. I still have to test it further to check as I put a system reset if osal_start_timerEx returns NO_TIMER_AVAILABLE.

Thanks.

Bluetooth®︎

Bluetooth forum

osal_start_reload_timer stops reloading timeout