[possible bug in STM32CubeIDE 1.3.0] SWV doesn't work for some STM32H7 MCU

[PHlin.1]

here is the original thread:

https://community.st.com/s/question/0D73W000000PWhW/debug-with-swo-and-swv-in-stm32cubeide-130

I try both - built-in ST-LINK and external one, in both cases I got the same problem, could somebody (community member or ST employee) confirm that SWV is working with STM32CubeIDE 1.3.0 and STM32H7 MCU? (STM32H7A3ZI in my case, but reported is also H743ZI)

Thanks,

Petr

[Pavel A.]

For me it works with H750 and H753.

But overall, debugging in CubeIDE is less robust vs. Atollic, IMHO

--pa

[PHlin.1]

Could you please confirm that you tested it with STM32CubeIDE 1.3.0? It works for me in 1.2.0.

Thanks,

Petr

[Pavel A.]

@PHlin.1​ You're correct, I tested it on v 1.2.0. Now updated to 1.3.1 and indeed it no longer works. :thumbs_down:

The failure seems to be caused by SWV. Even if it is enabled in debug config, but the "SWV ITM data console" red button NOT activated, debugger works normally.

As soon as the red button is clicked, debugger immediately fails.

The board is Nucleo H743.

Target is not responding, retrying...

Debugger connection lost.

Shutting down...

-- pa

[MEdgerton]

I have just started having the same issue. I upgraded the debugger firmware and after a few power cycles it worked and all the test programs worked.

Then it stopped. Now some work sometimes. Then others do. Then none.

This is on H743zi2 Nucleo dev board.

Previously I was using an F746 and this didn't have this issue. Along with the fact lwip actually works out the box (I've got it going on H743 now but it's not obvious what to do as it doesn't generate the code the same way as the F7).

The H7 is faster and better hardware, but with this debug and the older version of lwip together with it all being a real effort to get it to the same code level of the F7 makes me worry how well supported and how reliable the code generated by the IDE will be.

I may try the slower processor see if I can get enough. I digress though... back.to fixing the debug console issue.

Anyway any updates on resolving the debug issuse would be great. I'm going to try some other things to fix it over the next few days.

Mark