STLink failure - too many breakpoints

[squin]

Posted on August 02, 2016 at 10:49

I have a couple of STM32L476 mBed boards talking to each other. The development environment is Keil UVision. After some time one of the boards stops operating as expected. I then try to halt execution which seems to stop but no execution point is shown in the code. Pressing the next line, or jump out of buttons produces a message saying ''Debugger - Cortex-M Error'', ''This target device does not support all the defined breakpoints! Please reduce the number of breakpoints and start again''. I have tried this by pressing the ''Kill all breakpoints'' button and re-trying. The result is the same. The STLink reports hardware version V2-1 and firmware version V2J25M14. Any suggestions as to how I might get the debugger to function properly?

#stlink-mbed-debugger

[Tesla DeLorean]

Posted on August 02, 2016 at 17:10

That isn't the most up to date firmware, pull the ST-LINK Utilities (v3.9.0 w/V2.J27.S6) and update with those. Make sure Keil is using the most current .DLL

Sounds more like a connectivity issue. Does you device go into low power modes? Do you have USART output? Can you have a Hard Fault Handler that outputs usable diagnostic information if it gets stuck in there.

What version of Keil?

What exactly do you see in the registers and disassembly window *before* you step into the abyss?

[squin]

Posted on August 03, 2016 at 11:13

I only just checked this thread as I did not get an email saying anyone had replied (if these are supposed to be sent out).

I had updated the various drivers to the latest available from the website and repeated the error. After a while I stopped being stupid and looked at the disassembly window which showed the code looping in a section of flash which contained a BMP file for a splash screen. The error only occurred after around 20 minutes of run time so I was unable to single step into the abyss. I re-ran the test a few more times and found that a hard fault had occurred which allowed me to look at the stack and find the function call before the crash. This allowed me to find the code error. I imagine when the code runs wild like that it could cause issues with breakpoints (although I don't have a good understanding of the ST breakpoint mechanism).

The Keil UVision tool chain is V5.18a and the STLINK firmware now reports V2J27M15.

Thanks for getting back to me.