• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 99 subscribers
  • Views 1549 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.

xloader with ccsv5 on omap4430

divya ojha
Prodigy 90 points


Hi,

I am trying to compile and run xloader, using ccsv5 on omap4430, in linux. I am able to load the coad into the RAM but program counter does not jump to the loaded location. Could i get any input on how to resolve this.

  • 0
    TI__Intellectual 2505 points

    What board are you using? Does it have an SD card slot, or any other type of persistent memory? If so, you do not need CCS to compile and load x-loader. See the appropriate sections in one of the TI software release notes, such as http://omappedia.org/wiki/4AJ.1.1_OMAP4_Jelly_Bean_Release_Notes.

    Regards,

    Magdalena

  • 0 in reply to Magdalena Iovescu
    Prodigy 90 points

    I am using OMAP4430 rev A3, it has SD card slot and i have been able to boot through a card ( compiling the source as directed in the llink shared by you ).

    But i wanted to debug using ccs and also wanted to see if compiling and loading is possible using ccs.

    I have a doubt because ROM bootloader expects an xloader binary at serial/sd card/nand slots. I wanted to know if jtag interface can be used for boot using ccs. The binary is getting loaded but the program counter is not jumping to the start of the binary.

  • 0 in reply to divya ojha
    TI__Expert 8055 points

    Divya,

    To load the x-loader through JTAG, I believe you need to use the x-load file, not the MLO file that you would use for MMC booting.  The x-load file should be generated when you build the x-loader normally.  Have you placed the x-loader in RAM in the location as shown in section 27.4.2.2 of the OMAP4 TRM ("RAM Memory Map")?  I believe you will need to explicitly set the PC to this address; otherwise, the ROM code will be traversing the booting device list set by the SYSBOOT pins looking for a valid bootloader.

    Regards,
    Gina 

  • 0 in reply to Gina Glaser
    Prodigy 90 points

    Thanks for your inputs.

    In the present execution, the binary( x-load ) is not getting loaded in the RAM. I am using the default x-load.map and the gel files.

    The board is ES2.2. I attached here the gel file. It would be great if i could get to know if these files are doing the right kind of initialization.

    8233.omap4430_startup_common.gel0257.omap4430_cortexa9_cpu0_startup.gel

    Regards,

    Divya

  • 0 in reply to divya ojha
    TI__Expert 8055 points

    Divya,

    Earlier you mentioned that you are able to load the code into the RAM, but the PC doesn't jump to the loaded location.  Now you mention that the x-load is not getting loaded in the RAM.  Can you please clarify which the issue is or what you are doing differently in the two scenarios?

    Thank you,
    Gina 

  • 0 in reply to Gina Glaser
    Prodigy 90 points

    I am sorry, since I am new to ccs, I was not able to figure out what was different in two scenarios, later i realized that my gel files were improper in the second case.

    I saw this link, following which, i could execute u-boot step by step. - http://processors.wiki.ti.com/index.php/Sitara_Linux_Training:_uboot_linux_debug_with_ccsv5

    I tried do the same with x-loader, but with x-loader, two executable files get generated. I don't know which of these should be taken for load symbol option described as step 2 in
    "Perform SPL debug"

    Could i get any guidance on this, so that i be able to see step wise execution of x-loader on the board.

    Regards

    Divya

  • 0 in reply to divya ojha
    TI__Intellectual 2505 points

    Divya,

    We typically use Lauterbach to debug bootloaders on OMAP4, not CCS. For CCS specific instructions, please ask in the CCS forum, they will be able to better help you.

    Regards,

    Magdalena