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.

CCS/OMAP-L138: Can I send a run command to both arm9 and c674x cores when both cores are grouped?

Part Number: OMAP-L138
Other Parts Discussed in Thread: OMAPL138, TMDSLCDK138

Tool/software: Code Composer Studio

Hi Champs,

I grouped both arm9 and c674x cores and loaded simple sys/bios examples into the cores. Then I excuted a Resume command and got an exception error on arm9 as below.
But I could run the examples individually.

So, is it arrowed to group the cores and run them at the same time?



Regards,
j-breeze



  • Hi,

    So, is it arrowed to group the cores and run them at the same time?


    It should be possible.. I've done this goruping two dsp (tms320c66x) cores.

    Best Regards,
    Yordan
  • Hi Yordan,

    Thank you for your reply. I also have done the grouping dsp cores in c66x multi core devices(C6657 and  C6678) and run the dsp cores at the same time.
    How about the grouping arm9 core and c674x core in OMAP-L138? It is a heterogeneous multi core device. Is it arrowed to group the arm9 core and c764x core and run them at the same time?  

    Regards,
    j-breeze

  • I haven't tested this. Lets wait for the RTOS team feedback.

    Best Regards,
    Yordan
  • Hi j-breeze,

    I tested this and witnessed similar behavior. My guess is that it's due to a limitation of the JTAG debugger. What is the reason for needing to run both cores simultaneously?
  • Hi Sahin,

    I appreciate your cooperation.

    Actually, I don't want to do the grouping cores and run both cores simultaneously.
    What I really worry about is whether the error occurs or not when my program is running, because I am not sure the root cause now.

    Could you please let me clarify if you have any?

    Regards,
    j-breeze

  • Hi j-breeze,

    Can you share the program you're using so I can try reproducing this on my end?

    Also, which emulator are you using?
  • j-breeze said:
    I also have done the grouping dsp cores in c66x multi core devices(C6657 and  C6678) and run the dsp cores at the same time.
    How about the grouping arm9 core and c674x core in OMAP-L138? It is a heterogeneous multi core device. Is it arrowed to group the arm9 core and c764x core and run them at the same time?  

    Note that the behavior of the fixed group can vary depending on the types of cores that are grouped together. For example, the multi-core C66x devices support true synchronous execution behavior and any execution commands would be executed synchronously. 

    http://processors.wiki.ti.com/index.php/Multi-Core_Debug_with_CCS#Fixed_Groups

    I don't believe the same behavior would apply to the L138. You can still group the cores and it would still send the execution command to both cores but no true synchronous behavior and no guarantee which core gets the command first

    As for the behavior you are seeing, if there is a dependency for one of the cores to execute first, then that could explain the issue.

    Hope this helps explain this

    ki

  • Hi,

    As Ki mentioned, the OMAPL138 core is unable to have a tight synchronization, but the execution of grouped cores should work fine, provided they don't clash memory and peripherals. I can verify this with multiple versions of CCS (v6.2, v7.3, v8.0) and multiple JTAG debug probes (XDS100v2, XDS200, XDS560v2), therefore this is certainly an old and outstanding bug.

    In this case, I filed today the bug number CCBT-2263  - in a few hours you can check its status in the link SDOWP in my signature below.

    I apologize for the inconvenience,

    Rafael


  • Hi,

    Thank you so much for your support.

    The error that Rafael showed me in his previous post was different from the error I got when I run my program. It was a SYS/BIOS Abort.
    So, as Sahin suggested to me, I am considering asking you to share the program.
    I've already ported the program to TMDSLCDK138 from my custom board and confirmed that the same error occurred.
    However, I'm concerned about the sharing because the program includes our copyright codes and I don't want to attach it to this thread.

    So, could you please let me know how to share it?

    Regards,
    j-breeze

  • Hi j-breeze
    Please share it via Sugawara-san. He is in touch with us on this issue.

    Regards
    Mukul
  • Hi Mukul,

    Thank you for your reply.

    I heard from Sugawara-san about what I had to do for the sharing, and I've decided not to share the program.
    Because I'm not sure whether I can reproduce this issue or not after the code modification and the bug report has been issued.

    So, I'd like to consider the sharing again after the bug fix.

    I really appreciate all your help.

    Regards,
    j-breeze
  • Hi Rafael,

    I can not see the bug number CCBT-2263 in the link SDOWP. Could you please check the status?

    Regards,
    j-breeze

  • j-breeze,

    The bug was found to be originated from a different component (emulation protocol stack), therefore it was migrated and reassigned a different number: DBGTRC-3954. It is publicly available at the link SDOWP in my signature below.

    Regards,
    Rafael
  • Hi Rafael,

    Thank you for your support.
    I can see the DBGTRC-3954 now.

    Regards,
    j-breeze