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.

cannot launch project on L138 LCDK any more

Other Parts Discussed in Thread: OMAP-L138, OMAPL138

Hi!

I am working on a project on the OMAP-L138 LCDK and got stuck with a downloading problem:

Environment:

Win7 64bit
CCS5.5
SYS/BIOS 6.35
XDS100v2
LCDK OMAPL138
Programming the DSP of the OMAP with CCS5.5

The behavior of the system:

Until today I downloaded my code to the LCDK and started it after successful download in the debug window of CCS.
Usually the DSP is in suspend mode after download.
If not (happened every now and then), I suspend it, perform a CPU reset and a restart (all in the debug window).
Then I see a statement like "enty point reached" and I can start my code.

But today the behavior changed:

Now the DSP is not in suspend mode after downloading but running and if I reset it and click on restart it runs right away (at an address of 0x00712148 with an idle instruction).
I could not figure out how I can get back to the expected behavior described above.

Thanks for your suggestions,

Alex

  • Today it worked for about half an hour, now its again not possible to run code from ccs5.5 on the DSP of the LCDK.

    I could not find anything on this error in documents or on the e2e, and as I am relatively new to the ccs framework I got completely stuck with this problem.

    Alex

  • Alex,

    Unfortunately I don't have the same board as you, but a few threads discuss similar scenarios with tips that will allow you to further debug this scenario. I highlight the one below, which is very comprehensive:

    http://e2e.ti.com/support/development_tools/code_composer_studio/f/81/p/272191/952094.aspx#952094

    Are you running Linux on the ARM by any chance? If so, it may be interfering on the proper connection to the DSP (this post talks a bit about it).

    Hope this helps,

    Rafael

     

  • Hi Rafael! thanks for your help!

    I already found the first thread you linked. The difference to my situation is, that I cannot run any program on the DSP from CCS.
    I tried several example programs that can be downloaded (in the SYS/BIOS packet) and none of them works (although they were fine before).
    As the configuration of these examples is exactly for the LCDK L138 board I think it is not a GEL-problem.

    The LCDK comes with a linux-system on an SD card which runs perfectly if I insert the card and boot it.

    Before the run-problem occured, I successfully worked on the DSP without booting an ARM system.
    And as the LCDK was not changed I furthermore think it is unlikely to be a problem of the LCDK.

    To make it even more curious one of the demos worked today, but only once.
    A few minutes later the same demo did not. Certainly with NO changes of settings/code.

    I think I get me a bag to hit ;-/

    Alex

  • Alex,

    I am really unsure what may be happening in your case, but I would like to confirm one detail: are you using a GEL file with your board? If so, are you sure you are running the exact same initialization procedure every time?

    The reason of my question is that the LCDK has its GEL associated with the ARM9 core, therefore the device initialization only runs when the ARM core is connected first. If you connect straight to the DSP this will not run and therefore the behavior will be different at times.

    Also, if the issue does not happen if you load/run code straight from a cold boot (power up or using the board's reset button), it is possible this may be missing some important initialization steps. I would then try to reset the core before loading the code (or perhaps even power cycling the board) to get a consistent initial hardware status.

    Unfortunately I am running out of ideas... Apart from that I am not sure what else may be happening.

    Regards,

    Rafael

  • Hello Rafael,

    I do not have my own GEL file, I just used the configuration as it is in the demos.

    Although it worked really fine the last few days, I will change the LCDK board and try to reproduce the behavior today.

    Best regards

    Alex

  • Hello!

    I found out that the behavior was influenced by mechanical impact of the LCDK board.

    Although I am not 100% sure I think that it was a bad solder joint at one of the boot-cfg dip switch.

    After resoldering I did not face the above issues any more.

    Thanks for your help,

    Alexander