[SOLVED] SES v7.32 cannot talk to nordic target board

This site uses cookies. By continuing to browse this site, you are agreeing to our Cookie Policy.

  • [SOLVED] SES v7.32 cannot talk to nordic target board

    After upgrading to 7.32 the target connection has become unstable. Occasionally I can burn flash from SES (i.e. debug GO) but after a couple iterations the target connection no longer works. The only fix is to reboot the Mac.

    Downgrade back to SES v7.12 fixes the problem.

    My environment:

    MacOS 14.0
    JLINK v7.92j
    Target is custom board based on nRF54840
  • Screenshot of error attached.

    Support report is attached (support.txt)
    Images
    • Screenshot 2023-10-20 at 5.09.35 PM.png

      26.69 kB, 401×183, viewed 179 times
    • Screenshot 2023-10-20 at 5.09.58 PM.png

      181.34 kB, 821×806, viewed 189 times
    Files
    • support.txt

      (613.3 kB, downloaded 181 times, last: )
  • Hello,

    Thank you for your inquiry.
    The "low on memory" error from J-Link usually happens if a debug session is not closed gracefully and target connection is simply dropped during debugging.
    This might be due to low power modes, the target application itself or some incorrect debug setting.

    How exactly are you stopping the debug session usually?
    Do you lose target connection during debugging sometimes?
    Are you using the Nordic soft device?
    Did you make sure to use Monitor Mode Debugging, as the MCU may not be halted in the conventional way as long as the soft device is running?
    segger.com/products/debug-prob…y/monitor-mode-debugging/
    wiki.segger.com/nRF52_Series_D…Debugging_on_Nordic_nRF52


    Best regards,
    Nino
    Please read the forum rules before posting.

    Keep in mind, this is *not* a support forum.
    Our engineers will try to answer your questions between their projects if possible but this can be delayed by longer periods of time.
    Should you be entitled to support you can contact us via our support system: segger.com/ticket/

    Or you can contact us via e-mail.
  • I ran a very simple test where I used SES to build and load the app to the target, start the app, then hit the red stop button, then add some whitespace to main.c and repeat the sequence again. No breakpoints or stepping or anything like that.

    The first time I tried to load the app I got the error "failed to download application. Undocumented error". See screen shot.

    So I hit GO again and this time the app started running normally. I hit stop, adding some whitespace, and hit GO again. After a couple tries I got a new error "JLINK ERROR" and the console showed the error "T-bit of XPSR is 0 but should be 1. Changed to 1". Never saw that before.

    Then I rebooted the target board and SES and tried again. This time I was able to get 4 iterations without any errors.

    Note that this time I used the universal installer. Normally I used the "apple silicon installer" because I have the M1 chip. Not sure if that makes a difference.

    Answers to your questions:

    How exactly are you stopping the debug session usually? I hit the red stop button on SES debugger.

    Do you lose target connection during debugging sometimes? Yes for sure sometimes I have a bug in the code that causes the target to go off into the weeds. Rare but it happens.

    Are you using the Nordic soft device? Yes but I don't use breakpoints when the SD is running. If I need to use breakpoints I disable the SD.

    Did you make sure to use Monitor Mode Debugging, as the MCU may not be halted in the conventional way as long as the soft device is running?
    No I do not use MMD, instead I just turn off the SD for debugging.
    Images
    • Screenshot 2023-11-08 at 3.07.07 PM.png

      23.92 kB, 395×219, viewed 101 times
    • Screenshot 2023-11-08 at 3.08.37 PM.png

      19.97 kB, 398×182, viewed 125 times
    • Screenshot 2023-11-08 at 3.08.56 PM.png

      35.14 kB, 485×249, viewed 103 times
  • Hello,

    This is the first time we hear from such issue. So I assume it is some issue with the setup which worked on accident before with V7.12.

    Is the issue reproducible for you on an eval board? If yes, could you provide a reproducer so we can try to replicate the behaviour in house?

    Best regards,
    Nino
    Please read the forum rules before posting.

    Keep in mind, this is *not* a support forum.
    Our engineers will try to answer your questions between their projects if possible but this can be delayed by longer periods of time.
    Should you be entitled to support you can contact us via our support system: segger.com/ticket/

    Or you can contact us via e-mail.
  • A footnote for anyone interested..

    I upgraded to SES 8.10b and I no longer see this problem, in fact the target communications have never been better. I have no need to use 7.32 or any other 7x build so for me this issue is resolved.