• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 100 subscribers
  • Views 148 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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:

AM5729: TI PDK SBLLibCPUReset function stuck in while loop

AM5729: AM5729 vs AM5729-E: Utils_eveBoot Infinite Loop

Jared Silbermann
Prodigy 155 points
Part Number: AM5729

Tool/software:

Hi,

I am having some problems with ti-opencl monitor firmware execution which I believe may be related to the related question on this post. I have a variety of different AM5729 chips:

Those with a DIE_ID_2 (Part Number Identifier) starting with 0x4E all seem to get stuck in an infinite loop somewhere with Utils_eveBoot function (related posts suggests it is related to SBLLibCPUReset  / PMHALResetRelease ).

Those with a DIE_ID_2 starting with 0x3E all have no problems and opencl firmware can execute correctly.

From looking into the ti-opencl source code there are checks if the board type starts with 0x3E or 0x4E as such:

static int GetNumEVEDevices()
{
uint32_t board_type = ( *((uint32_t *) CTRL_WKUP_STD_FUSE_DIE_ID_2)
& 0xFF000000) >> 24;
int num_eves = 0;
if ( board_type == 0x3E || // AM5729-E (EtherCat)
board_type == 0x4E) // AM5729

This is where I noticed different names in the comments for these chips. I can see online the AM5729-E  refers to the chip with all industrial protocols enabled. Are there differences with the enabling of the industrial protocols  that may be contributing to the issues I have having. I am using SDK 6.0.032 and the main problem is I am unable to run any OpenCL related examples. I am quite suspicious as to why all of my AM5729-E boards work with OpenCL but AM5729 boards do not, and would love to know the different between these Part Numbers.

I have made another post regarding this issue for reference but made another post on the suggestion of the TI support member who was helping me (https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1346538/am5729-ti-opencl-firmware-execution-problem---stuck-at-booting-eve-s---tiocl-fatal-failed-to-open-eve-message-queue).

Regards,
Jared

  • 0
    TI__Mastermind 25209 points

    Hello Jared,

    I will need to confer internally with regards to these differences. Unsure as to what could be the correlation. 

    From the datasheet, on both AM5729 and AM5729-E boards, the EVE cores are supported.

    The -E devices only enable a feature related with industrial protocols. 
       -e2e.ti.com/.../am5726-industrial-protocols-support

    Just to confirm, are all of your boards AM5729x?

    -Josue

  • 0 in reply to Josue Zamitiz-Ayala
    Prodigy 155 points

    Hi Josue,

    Yes they are AM5729 boards.

  • 0 in reply to Jared Silbermann
    TI__Mastermind 25209 points

    Hello Jared,

    I have confirmed internally with one of our HW engineers, this should not affect the functionality of the EVE cores. The check is for corresponding e-fuse register and should only affect the mentioned Industrial Protocol , and the number of eve cores depending on the device type.

    Datasheet exerpt:

    Some further Questions:

    • What is the exact rate of fault occurrence and how many devices?
    • Can you please share exact part numbers?
    • Are these custom boards or TI EVMs?

    I do not see a correlation so far.

    Best,

    Josue