• State Resolved
  • Locked Locked
  • Replies 10 replies
  • Answers 1 answer
  • Subscribers 101 subscribers
  • Views 329 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.

AM69A: AM69A: Intermittent Boot Failures from SPL -> U-Boot

Ernest Van Hoecke
Prodigy 200 points

Part Number: AM69A
Other Parts Discussed in Thread: AM69, SK-AM69

Hi,

This is a continuation of https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1522868/am69a-intermittent-boot-failures---frequency-handshake-timeout-and-spl-hang

The first issue described there was resolved.

The second issue however is still open.

I recreated the setup but this time with the TI AM69 EVM and the 6.12 SDK. Here, I gathered the following statistics on how often the issue occurred: Total Successes: 14740, Failures: 20

This test setup monitors the UART output and decides a boot is a fail every time the U-Boot header is not found in the output. I confirmed that in those cases, the execution is stuck and nothing is happening anymore. In case the U-Boot version print happened, it immediately powers off  the whole board for a second and then restores power (I used POWERDOWNz for this on the automation header). 

I have not gathered the traces again in the manner that Parth did earlier, but since this was done with the default SDK and EVM, it should be possible to recreate on your side. When rebooting immediately upon successfully reaching U-Boot, it should not take many hours to trap the bug as it is expected to happen 1 in 750 power cycles and only takes a few seconds per iteration. Please let me know if it is helpful anyways to enable more tracing and what to pay attention to, and I will get back to you soon.

Have there been any developments from your side on this problem? Thanks in advance!

Kind regards,
Ernest 

  • 0 in reply to Keerthy J
    Intellectual 725 points

      

    no, that was not tried, but given that this is reproducible with your own board I would appreciate if you (TI) could take full ownership of this debugging task, starting with the confirmation that you can reproduce it on your desk and on your own HW.

  • 0
    TI__Mastermind 41380 points

    Hi,

    Ernest Van Hoecke said:
    I recreated the setup but this time with the TI AM69 EVM and the 6.12 SDK

    I noticed the above statement and not sure if maybe the SDK version is incorrect. 

    Can you clarify which SDK you are referring to?  

      , do you know which SDK is being referenced above?

    Regards,
    Kevin

  • 0 in reply to Kevin S
    Prodigy 200 points

    Hi  

    I meant the SK-AM69 instead of the EVK, apologies for the confusion. I used the PROCESSOR-SDK-LINUX-AM69 software release v11.00.10.01, which uses Linux 6.12. 

    Does that clear things up?

    Thanks and kind regards,

    Ernest 

  • 0 in reply to Ernest Van Hoecke
    TI__Mastermind 41380 points

    Hi,

    Ernest Van Hoecke said:
    I used the PROCESSOR-SDK-LINUX-AM69 software release v11.00.10.01, which uses Linux 6.12

    Yes, this makes more sense. I was looking for the SDK version as opposed to the Linux version, but didn't see an SDK version earlier than 08.xx for AM69.  

    Note, I see SDK Linux version 11.00.00.08. Will check with Keerthy if something newer has been released.

    Regards,
    Kevin

  • 0 in reply to Kevin S
    TI__Guru**** 167700 points

    Hi,

    Thanks for the details. Can you share the failing logs?

    Also share the corresponding passing logs. 

    Best Regards,

    Keerthy 

  • +1 in reply to Keerthy J
    TI__Guru**** 167700 points

    Hello,

    As per the 2nd failure:

    U-Boot SPL 2024.04-34430-gf8c18ba41a3-dirty (May 16 2025 - 08:41:36 +0200)
SYSFW ABI: 4.0 (firmware rev 0x000b '11.0.9--v11.00.09+ (Fancy Rat)')
Initialized 4 DRAM controllers
SPL initial stack usage: 13456 bytes
HW CFG: 0x 0
Trying to boot from MMC1
Authentication passed
Authentication passed
Authentication passed
Loading Environment from nowhere... OK
init_env from device 17 not supported!
Authentication passed
Authentication passed
Starting ATF on ARM64 core...

NOTICE:  BL31: v2.11.0(release):v2.11.0-906-g58b25570c9-dirty
NOTICE:  BL31: Built : 04:20:32, Nov  1 2024
I/TC: 
I/TC: OP-TEE version: 4.4.0-dev (gcc version 13.3.0 (GCC)) #1 Fri Oct 18 17:45:27 UTC 2024 aarch64
I/TC: WARNING: This OP-TEE configuration might be insecure!
I/TC: WARNING: Please check https://optee.readthedocs.io/en/latest/architecture/porting_guidelines.html
I/TC: Primary CPU initializing
I/TC: GIC redistributor base address not provided
I/TC: Assuming default GIC group status and modifier
I/TC: SYSFW ABI: 4.0 (firmware rev 0x000b '11.0.9--v11.00.09+ (Fancy Rat)')
I/TC: Activated SA2UL device
I/TC: Enabled firewalls for SA2UL TRNG device
I/TC: SA2UL TRNG initialized
I/TC: SA2UL Drivers initialized
I/TC: HUK Initialized
I/TC: Primary CPU switching to normal world boot

U-Boot SPL 2024.04-34430-gf8c18ba41a3-dirty (May 16 2025 - 08:42:40 +0200)
SYSFW ABI: 4.0 (firmware rev 0x000b '11.0.9--v11.00.09+ (Fancy Rat)')
HW CFG: 0x 0
Trying to boot from MMC1
Authentication passed
Authentication passed

------- Hangs here -------

    The failure is in A72 SPL stage. We have seen similar failure in A72 SPL stage during OSPI boot.

    The fix: https://lore.kernel.org/all/20250325053420.4wuenxmlnqsojuyt@prasanth-server/T/#u)

    Could you try the above patch and let us know if it helps with MMC1 boot as well.

    - Keerthy

  • 0 in reply to Keerthy J
    Prodigy 200 points

    Hi  ,

    I can happily confirm that this patch solved the issue, thanks! I will comment on the mailing list as well to ask about the upstream status of this.

    Kind regards,
    Ernest

  • 0 in reply to Ernest Van Hoecke
    Intellectual 725 points

     : can you please help getting this fixed in mainline and in your U-Boot branch?

  • 0 in reply to fd
    TI__Guru**** 167700 points

    Yes. The development team is already working on it and hence you see the upstream U-boot post. I do not know when it will get in but team is working on it and should get in the next SDK.

    Closing this as you are unblocked.

    - Keerthy