• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 97 subscribers
  • Views 646 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.

EZSDK on 8148 based custom board.

Gilles Chanteperdrix
Prodigy 200 points

Other Parts Discussed in Thread: TVP7002, SYSBIOS

Hi,

I am trying to boot the EZSDK file system on a custom board using a 8148 processor. This custom board has 512MB DDR, so, I followed the instructions here:

http://e2e.ti.com/support/dsp/davinci_digital_media_processors/f/717/p/131933/488024.aspx#488024

To setup the memory map, that is, I modified the configuration of the LISA registers in the u-boot port for our custom board, and modified the root file system to pass a memory map binary to the "firmware_loader" command (to be completely clear, where 0xB3D00000 is replaced with 0xAB000000).

The EZSDK boot scripts fail when trying to load the VPSS firmware. Adding some traces, it seems firmware_loader is blocked line 216 when calling:

status = Ipc_control (procId, Ipc_CONTROLCMD_STARTCALLBACK, NULL);

Enabling GT_trace in the syslink module, I see that syslink is looping, with an occasional message:

Ipc_attach: Ipc_procSyncStart failed!

The exact same root file system works on the 8148 evm board, at least up to the point where the "matrix-gui" application starts and images are displayed on the HDMI output.

What could be the cause of this issue?

Thanks in advance.

 

Gilles

  • 0
    Prodigy 200 points

    Hi again,

    I see some posts with apparently similar issues:

    http://e2e.ti.com/support/dsp/davinci_digital_media_processors/f/717/p/132288/474584.aspx#474584

    http://e2e.ti.com/support/dsp/davinci_digital_media_processors/f/717/p/98209/343367.aspx#343367

    Both posts seem to suggest that the firmware loaded on DM8168 may be accessing hardware only present on the EVM board, could we have the same issue on DM8148?

    Does the EZSDK contain the sources of this firmware, which would allow us to re-generate a firmware?

     

  • 0
    TI__Intellectual 2910 points

    The HDVPSS firmware seems to have asserted and stopped. That is why A8 is trying to Ipc_procSyncStart  with HDVPSS slave core in a tight loop from which it will break out when a Ipc_attach with A8 occurs on the HDVPSS slave core.

     We do have a function in the firmware which returns an error when the recommended I2C slave devices (video encoders/ decoders as in TVP7002/Si9022A) are not located at the addresses as expected. But it is not supposed to assert, then sends out a warning which says that the Expansion card is not connected.

    If you have an emulator (XDS-510/560v2), you can confirm the state of HDVPSS-M3 (ISS-M3 in the CCS debug terminology) by connecting to the core and loading symbols from the executable that gets loaded on the hdvpss core.

     

    Archith 

     

  • 0 in reply to Archith John Bency
    Prodigy 200 points

    Hi,

    I attached the JTAG and loaded the symbols. The JTAG interrupts the processor in what seems to be the OS idle loop.  The stack trace being:

    ti_sysbios_utils_Load_idleFxn__E()

    ti_sysbios_knl_Idle_loop__E()

    ti_sysbios_knl_Task_exit__E()

    ti_sysbios_utils_Load_idleFxn__E()

    0xBEBEBEBE

    Are we right to assume that the prebuilt binary for hdvpss found at board-support/prebuilt-images/dm814x-firmware_05_02_00_14 in the EZSDK may be loaded on custom hardware?

    Thanks for the help.

    Regards.

     

    Gilles