CCS/AWR6843: Debugging Problem with synchronized Cores: DSP is halted but Cortex doesn't halt

Hi Frank,

I tried this with my IWR6843ISK. The cross-triggering between the two cores seemed to work (a halt of the DSP with trigger the ARM to halt also).

Instead of using a breakpoint, if you manually halt the DSP with both are running, does the ARM halt?

Also, what version of CCS are you using?

Thanks

ki

Hi Ki,

I am using CCS Version: 10.1.0.00010.

If I normally halt the cores it works. But with breakpoints not.

I also get this error when synchronising the cores:

Cortex_R4_0: Trouble configuring cross triggering: Error enabling this function: There is no AET resource to support this job.

Thank you very much.

Regards,

Frank

Hello,

I found a suggestion here:

https://e2e.ti.com/support/tools/ccs/f/81/t/335425?Error-enabling-this-function-There-is-no-AET-resource-to-support-this-job

"Hardware breakpoints use an emulator resource called AET. There are a limited amount of AET resources. Other things that use AET are watchpoints, the profile clock, etc. When you are out of AET resources, you need to release some by disabling some of the other actions that are using them. That means disabling other hardware breakpoints. Note that there many "hidden" breakpoints that are automatically set by the debugger. Some of theses are breakpoints set at the exit point of the application or your start label (main). CIO will also use breakpoints. You can disable many of these in the debugger options."

But this isn't working too.

Regards,

Frank

Frank Kuchenbrod said:

I also get this error when synchronising the cores:

Cortex_R4_0: Trouble configuring cross triggering: Error enabling this function: There is no AET resource to support this job.

Do you get this error every time you try to enable a sync group?

I have been experimenting with breakpoints. In my example, I have a program with a simple loop. When I set a breakpoint for the DSP and run both cores, the ARM does not halt as you said. But when I resume execution and the breakpoint is hit again, I do see both ARM and DSP halt. The behavior I see is that the triggering works every other time. Is that along the lines of what you see?

Hello,

it seems to be the same.

I'm working with mmwave_sdk:

In my case the cores are synchronised in software and if the core not stops it runs unfortunately in a trap.

Regards,

Frank