• State TI Thinks Resolved
  • Locked Locked
  • Replies 12 replies
  • Answers 1 answer
  • Subscribers 84 subscribers
  • Views 347 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TMS320F28P550SJ: How to debug two 28p55 dsp that are connected in dasiy-chain using only one xds100v3 isolated debug probe?

songmingbo
Prodigy 165 points
Part Number: TMS320F28P550SJ


Tool/software:

Hi, TI's experts.

    I have a question about debugging two 28p55, they are connected with only one xds100v3 debug probe in dasiy-chain topology. Now I can download the firmware to the respective flash memory of each DSP. However, I'm having trouble setting breakpoints in the code for the slave DSP. The console is showing an error, and I can't set the breakpoints.  Please help me solve this problem. Following is a screenshot of my ccxml file.

  • 0
    TI__Genius 13035 points

    Hello, 

    Can you share the console error message? What version of CCS are you using?

    There is a limitation of two hardware breakpoints when working flash. Can you try not having any breakpoints in the master DSP and see whether you can set breakpoint in the slave DSP?

    Best,

    Matt

  • 0 in reply to Matt Kukucka
    Prodigy 165 points

    Hello,

    Yes, I can show you the message. Following the first one is the screenshot of the console and I use the CCS v12.8. You can see I put only one HW breakpoint in the slave DSP but it was wrong. However I found that I can put a breakpoint into the interrupt service routine of the slave DSP but I can not put it in the task.

    When I put the same breakpoint into the master DSP, it works successfully. Let me show you in the second screenshoot.

  • 0 in reply to songmingbo
    TI__Genius 13035 points

    Hello,

    That is unusual. I have a few clarifying questions:

    1. Are you using any compiler optimizations?
    2. Did you load the symbols for the program on the slave DSP before setting the breakpoint?

    Best,

    Matt

  • 0 in reply to Matt Kukucka
    Prodigy 165 points

    Hi,

    1. Yes, I am using optimization level-2;

    2. No, I just load the program and click run it.

    Additionally, I found that I can put software breakpoint in the slave DSP but I can not put hardware breakpoint in the slave DSP. I am using FreeRTOS. When I try to set breakpoints in a task, it fails. This is because tasks generally run in flash memory. However, setting breakpoints in functions located in the ramfuncs section works successfully. So I want to know if the daisy-chain topology supports setting hardware breakpoints in the code of the slave DSP?

    Best,

    Mingbo Song

  • 0 in reply to songmingbo
    TI__Genius 13035 points

    Hello,

    Hardware breakpoints are supported in daisy-chain topologies. 

    songmingbo said:
    1. Yes, I am using optimization level-2;

    Can you try turning compiler optimizations off?

    Also, please try removing potential "hidden" breakpoints: https://software-dl.ti.com/ccs/esd/documents/troubleshooting_no-aet-resources.html#resolution

    Best,

    Matt

  • 0 in reply to Matt Kukucka
    Prodigy 165 points

    Hello.

    Sure, Now I have turned off the optimizations and removed all of the breakpoints, but I'm quite sure this still isn't working. 

    Could you please take the board you have on hand and conduct an experiment to see if it is really possible to set hardware breakpoints in the flash code of the slave DSP?  If it's possible, thank you very much.

  • 0 in reply to songmingbo
    TI__Genius 13035 points

    Hello,

    I will attempt this test tomorrow. In the meantime, I'll loop in the CCS experts to see if they can provide additional assistance.

    Best,

    Matt

  • 0 in reply to Matt Kukucka
    TI__Guru**** 471801 points

    There is a known issue with hardware breakpoints on the second device of a dasiy-chained target environment:

    https://sir.ext.ti.com/jira/browse/EXT_EP-10378

    While the information in the link above specifically mentions Tiva devices, it is quite possible it impacts other diasy-chained devices also.

    Is the breakpoint in question a hardware breakpoint? Do software breakpints work?

    Thanks

    ki

  • 0 in reply to Ki
    TI__Genius 13035 points

      ,

    Yes, this happens when they are attempting to set hardware breakpoints (in the flash). Software breakpoints (in the ramfuncs) do not have this issue and work as intended.

    Best,

    Matt

  • 0 in reply to Matt Kukucka
    TI__Guru**** 471801 points

    Matt - thanks for trying this out. It is likely related to that other ticket I mentioned. I will mention this to engineering to see if we can get an update.

  • 0 in reply to Ki
    Prodigy 165 points

    Hello, Ti expert.

    Has a solution been found for this problem?
    I am looking forward to this problem being solved.
    Best,
    Mingbo Song.
  • 0 in reply to songmingbo
    TI__Guru**** 471801 points

    I discussed this with engineering and they will look to address this with the CCS 20.4.0 release scheduled for the end of the year.

    UPDATE/EDIT: Unfortunately the resolution for this issue is more involved than anticipated and it will not be in the CCS 20.4.0 release. It has been pushed out to 20.5.0. My apologies.