[ABANDONED] Unable to capture trace data

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

  • [ABANDONED] Unable to capture trace data

    I'm using a J-Trace Cortex-M3 probe (not a Pro) and can't capture any trace data. In the interest of determining the issue, here's the simplified system I'm using to get trace working:

    Probe: Segger J-Trace Cortex-M Rev.3, HW V3.10, updated with latest firmware (Nov 23, 2017)
    Debugger: TrueStudio 9.1.0 (Ozone doesn't support trace on non-Pro J-Trace probes)
    Board: Segger Cortex-M Trace Reference Board, executing a very simple loop
    Signal Integrity: I confirmed proper output levels and Setup/Hold at the probe ribbon cable with a 500 MHz realtime DSO. There's little or no signal skew that I can detect.
    Triggers: (1) Trace Start Trigger and (1) Trace Stop Trigger in the main() loop

    I can confirm the 40 MHz trace clock and (apparently) trace packets, based on the logic activity on the 4 trace data lines, shown on the DSO. I say 'apparently' as I haven't captured the data stream manually, nor do I know how to interpret it.

    The software and debug probe seem to be operating correctly, as it appears the probe is able to capture packets, as shown in the RAWTrace image below.

    No matter what I try, single stepping / running without breakpoints, I cannot get the probe / debug software to capture/decode any instructions from the 4-bit ETM interface into the trace buffer.

    Any ideas?
    Images
    • LiveTrace.jpg

      141.54 kB, 838×550, viewed 503 times
    • RAWTrace.jpg

      148.69 kB, 836×551, viewed 518 times
  • Hello,

    Thank you for your inquiry.
    We just tried to reproduce the behaviour with an old 3.2 probe but everything was working as expected with the tutorial project and Ozone:
    segger.com/downloads/jlink/J-Trace_PRO_CortexM_Tutorial.zip

    Could you give the tutorial project a try? While you might get a pop up in Ozone that no valid license is detected it is fine to use for evaluation.
    Is the tutorial project working?

    Generally we recommend upgrading your trace probe to a J-Trace PRO as the J-Trace CM V3 is both out of support and maintenence.
    Here are some benefits listed for upgrading:
    wiki.segger.com/J-Trace_vs._J-Trace_PRO


    You can also use our trade in program so you get a discount on the new probe:

    segger.com/purchase/trade-in-program/

    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.
  • Yes, the tutorial program worked fine with the Cortex Reference Board under Ozone. All other aspects of the tutorial worked very well with the 3.1 probe - it's a nice compact demonstration and I commend Segger for such an elegant solution to getting customers started with trace. However, I could not capture an ETM trace with Ozone and the Reference Board. I abandoned the tutorial project and Ozone after I read the following in the Ozone User Manual

    5.13.3 Debug Probe Requirements

    - ETM trace requires a J-Trace PRO model to be employed


    Perhaps you've got an internal version of Ozone that doesn't require a newer probe? :whistling:

    I am exploring upgrading to the Pro model, as the code profiling would be a helpful capability and I prefer to avoid Keil if possible. The effort with the 3.10 probe began as an exercise to ensure the signals on my application board were sufficiently clean to support trace.

    Since the 3.1 probe is out of support, I was hoping some user might be able to suggest how the probe could capture and decode the ETM frame data...but not capture a trace. The 3.1 probe seems to function properly in every other way.
  • Hello,

    RMS_STM32 wrote:

    Perhaps you've got an internal version of Ozone that doesn't require a newer probe?
    Unfortunately it is not gonna happen, Ozone was created after the J-Trace CM 3.x series got discontinued so it would be a significant support hassle for a probe that is no longer produced.

    RMS_STM32 wrote:

    I am exploring upgrading to the Pro model, as the code profiling would be a helpful capability and I prefer to avoid Keil if possible. The effort with the 3.10 probe began as an exercise to ensure the signals on my application board were sufficiently clean to support trace.
    Great to hear. You can find a list with tested devices and trace examples for Ozone here:
    segger.com/products/debug-prob…echnology/tested-devices/


    RMS_STM32 wrote:

    Since the 3.1 probe is out of support, I was hoping some user might be able to suggest how the probe could capture and decode the ETM frame data...but not capture a trace. The 3.1 probe seems to function properly in every other way.
    What would be the use case here?
    If you are looking to write your own trace interpreter, logger the J-Link SDK can be used for this:
    segger.com/products/debug-prob…nk/technology/j-link-sdk/

    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.