• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 779 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.

Original question:

LP-AM243: Unable to perform SOC initialization using CCS scripting

LP-AM243: Initialization script is not working

Isaac Lucas de Lima Yuki
Intellectual 315 points
Part Number: LP-AM243


Hello all,

I'm unable to debug with the script load_dmsc_hsfs. 

Sometimes I get the behaviour which is described in the related post, and it seems to load the sciclient initialization SBL. 

But also sometimes it doesn'?  What is going on? I'm using the same GEL files, the same sci client, the same scripts. 

The only thing I alterated was the SBL itself, changing the version to debug instead of release using the example.

```

Connecting to MCU Cortex_R5_0!
Writing While(1) for R5F
Running the board configuration initialization from R5!  //Gets stuck here, no more console log, no loading of app
com.ti.debug.engine.scripting.Target.run(): Timed out after 20000ms (C:\ti\mcu_plus_sdk_am64x_09_01_00_41\tools\ccs_load\am64x\load_dmsc_hsfs.js#117)

```

Regards,

Isaac Yuki

  • 0
    TI__Guru* 75731 points

    Hello Isaac,

    As long as the `sciclient_ccs_init` example is getting loaded and run, you can ignore that timeout from the `load_dmsc_hsfs.js` script. When this timeout occurs, are you not getting any logs from the `sciclient_ccs_init` example like shown

    Regards,

    Prashant

  • +1 in reply to Prashant Shivhare
    Intellectual 315 points

    Hello Prashant,

    This exactly was the problem I was facing and that's also why I couldn't load any example into the DDR. The SBL was not run, when I loaded the script.

    Now I found out why:

    Starting the load_dmsc_script and loading the sciclient_css_init example is only really ensured when you start the debugger over your target config.

    It seems that load_dmsc_script does not properly load the active/default script from CCS and this at least explains the past behavior.

    Now that I deleted all Target Configs and start the debugger over a user-defined target config, things behave as expected.

    This is, I think, a big issue to address when handling the CCS debugger. This cost me 4 full days of frustration and I would hope that TI could at least document this issue somewhere and maybe give the exact reasons why.

    Thank you for your fast responses, Prashant, it really means a lot.

    Regards,

    Isaac

  • 0 in reply to Isaac Lucas de Lima Yuki
    TI__Guru* 75731 points

    Hello Isaac,

    Isaac Lucas de Lima Yuki said:
    Starting the load_dmsc_script and loading the sciclient_css_init example is only really ensured when you start the debugger over your target config.

    If you are not first launching the target configuration then that probably explains the issue. Actually, the target configurations needs to be launched first before running the `load_dmsc_hsfs.js` script. This is also stated out in the documentation here

    https://software-dl.ti.com/mcu-plus-sdk/esd/AM64X/09_01_00_41/exports/docs/api_guide_am64x/HSFS_MIGRATION_GUIDE.html#CCS_BOOT_HS_FS

    ------

    Steps to do DMSC initialization from CCS

    • Set the board in DEV-BOOT mode (mandatory)
    • Power ON the EVM/board
    • Launch the CCS target configuration file
    • Run the load_dmsc_hs_fs.js script from the scripting console

    ------

    On another note, you actually don't really need to run this JS script. It is more of a convenience that necessity. All this script does is reset the R5F core then load and run the `sciclient_ccs_init` example. This you can do manually as well. I have attached below the screen recording showing the same:

    Regards,

    Prashant

  • 0 in reply to Prashant Shivhare
    Intellectual 315 points

    Yeah, got that. Still you don't think on this, when you're searching for the reasons on the code itself. Anyways, this issue can be marked as resolved.