[SOLVED] Cannot quit Ozone after J-Link lost connection to target

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

  • [SOLVED] Cannot quit Ozone after J-Link lost connection to target

    Hi,

    when I am connected to a target and debugging in Ozone via J-Link Pro, and then the target is shut down or the connection is lost for some other reason, Ozone is stuck and cannot be used anymore.

    - Debug.Stop(); has no effect
    - trying to quit the whole application results in a dialog box asking whether the session should be closed. Both choices offered, Yes and No do not have any effect. I can only ask Windows 10 to terminate the application. Is this a known problem?

    Thanks and regards
  • Hello,

    Thank you for your inquiry.
    Such an issue is not known to us.
    Could you provide more information about your setup?
    What version of Ozone are you using? Does it work when using the latest one?
    What host OS are you running Ozone on?
    Does the issue appear only with a specific target device or with any?

    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.
  • Hi,

    I am running Ozone V3.10 (which is the most recent one) on Windows 10 Pro and this happened to me with a custom board (stm32f412ce) as well as with the blue pill (stm32f103c8t6).

    Both boards have an exteral power supply and if that goes away, Ozone gets stuck in the way I described.

    Well, I just tried to reproduce this and failed when:
    - disconnecting the power supply from the board
    - sending the PC to sleep while the debugging session was active

    In both cases, Ozone behaved correctly. ... But I can observe the other behaviour regularily; I am just not sure how to reproduce it properly.

    In one case I got the message that the J-Link was started in bootloader mode and needed its firmware to be re-flashed. ... Weird, but that's another issue.

    Kind regards

    The post was edited 4 times, last by icwiener ().

  • Hello,

    FYI, the currently latest version of Ozone is V3.10c
    We tried to reproduce the reported behaviour but Ozone was shutting down correctly every time power was cut from the target board.
    Please understand that we can't put more time into this without a reliable reproducer.
    Should you find a reliable way to reproduce the reported behaviour feel free to provide us with that information.

    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.
  • Hi,

    thanks for trying. Yes, I found the V3.10c shortly after writing and saw the issue with it as well. Surprised me that a one month old installation was outdated by several versions already. :)

    I have problems reproducing the behaviour quickly but leaving the setup in a debugging session running over night by just Locking the screen (and Windows going to hibernate later) gives me a good chance of seeing this.

    However, this might be a very rare (and not advisable) thing to do, so I will just try to avoid it. :)

    If I find a way to reproduce this more easily, I will report back.

    Cheers
  • Hi,

    now using 3.10d I just had this happeing to me right in the middle of a debugging session.
    No lost connection, no Windows hibernate.
    Ozone just stopped sampling data and now I can neither stop debugging nor quit the application.

    There is no output on the Console when it stops sampling.

    I'll add a screen shot of the status bar, maybe it's helpful.

    Best regards
    Images
    • ozone_status.PNG

      2.42 kB, 571×21, viewed 479 times
  • Hello,

    Does your target application have any watchdogs, low power modes or similar running that might interfere with the debug interface?
    You seem to record quite a chunk of data samples, does the freeze happen around the same samples number each time or different ones?
    In Ozone under Tools->Preferences->General->Data Sampling Window->Data Limit, what size do you have set here? Does it make a difference if you increase this value significantly?

    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.
  • Hi Nino,

    there is a watchdog but it did not hit. And we do not use any low-power modes.

    it does not happen regularly enough to answer if it happens around the same sampling number. I will check next time.

    The Data Limit configured is 16000

    Best regards
  • Hi,

    I just observed that after around 420 000 samples the Timeline view and the status bar (when Data Sampling or Timeline are selected) become rather slow and even stuck. The Data Sampling view is showing data just normal.

    I am running at a sampling frequenzy of 10kHz just for provoking this behaviour and have 15 expressions in the Data Sampling table, some of which are updated quite often (probably close to the 10kHz).

    Seems the Timeline feature is at its limits here and maybe this leads to the observes behaviour of not being able to close the application (althou I am currently not running into this).

    Best regards
  • Hi,

    a few more observations.

    It is easiest to trigger the stuck timeline at 10kHz. When reaching around 420 000 samples, it becomes stuck here.

    If I then set the Sampling Freq to 100Hz, the graph moves again.

    If I set the Sampling Freq to 500Hz however, the graph starts to move in the wrong direction. The time also counts backwards at that point.

    Set back to 100Hz the graph moves again but it does not show current values but values from the past. I can drag the graph forward and see more values but it snaps back to the wrong position in time as soon as I release the mouse button.

    Best regards

    The post was edited 1 time, last by icwiener ().

  • Hi,

    this still hits me several times a week in Ozone v3.20a.

    I do not know how to cleanly reproduce but it always happens when a debugging session runs for an hour or two.



    Maybe you are able to check your code for the case when the debugging session is not stoppable anymore.


    Edit: It happened again, this time with a different Sampling Frequency. In the video I had set it to 1kHz and now at 100Hz.
    Both times it foze at around 400 000ms



    Cheers

    The post was edited 2 times, last by icwiener ().

  • Hi,

    I tried to reproduce the behaviour with the additional information but everything is working fine for me even after letting it run over 2 hours.
    It appears to be setup specific.
    Could you provide an example project that would run on some eval board with which the behaviour was reproducible with?
    Is the behaviour reproducible more often if you decrease the data buffer size for the data graph?

    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.
  • Hi,

    since I can only debug this at work and I am usually busy doing other stuff, I find it hard to create a minimal example project.

    Would it be possible for you to send me some debug version of Ozone which writes some log file?

    This still happens to me two to four times a day, which is quite inconvenient.

    Cheers
  • Hi,

    we received your request via our support system in parallel. To avoid duplicate or even lost information between channels this thread will be closed now and continued via our support system.

    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.