AM263P4: TCMB RBL Log - Persisting over power cycle

Part Number: AM263P4

If the ROM Bootloader fails to load the Secondary Bootloader and records the fault in volatile memory (e.g., TCMB), but the unit is subsequently power-cycled, how is that boot failure log retained or retrieved on the next startup, given that volatile memory contents may be lost?

  • Hi Addison,

    To analyze the ROM failure, failure dump is taken in the same session without power cycling the device. The failure log will be present in the TCMB region as explained in the 5.9.2 Logger section of AM263P TRM.

    - May I know if you are you able to observe that the TCMB is retaining the logs persistently even after power cycling? 

    - May I know which is the configured boot mode?

    -  Also, are you performing a complete POR or just a warm reset.

    Thanks & Regards,

    Aswin

  • How is the failure dump retrieved from the device if there is a failure in the RBL?

    In the AM263P TRM it states "Any failures detected by R5 or HSM while booting, will lead to SoC warm reset issued by WDT (watchdog timer) after 180 sec. From the ROM perspective, cold reset and warm reset are the same as far as boot flow is considered." Would the logs persist over the warm reset? How are they accessed in this case?

    I am configured in secure boot mode.

  • How is the failure dump retrieved from the device if there is a failure in the RBL?

    The system will persist in this failure state until the WDG is triggered. WDG trigger happens at t=180sec. The ROM log i.e the failure log from TCMB can be taken during this time before the WDG issues warm reset.

    I am configured in secure boot mode.

    Can you tell me which is the boot media, i.e OSPI boot mode, QSPI boot mode, UART boot mode etc?

  • Are the ROM failure logs retrieved using JTAG during that time period?

  • Hi Addision, 

    Yes, The TCMB log can be taken by connecting the debugger and then by taking the memory dump.

  • Does this hold true that you can connect to the debugger and download the logs before the device enters warm reset after a failure in the RBL, if the device is in HS-SE mode (when in that mode - JTAG connection to either M3_0 or R5_0 is prohibited). 

    Also just to confirm - after the warm reset triggered by the WDG - the logs in TCMB will no longer be available? 

  • Also just to confirm - after the warm reset triggered by the WDG - the logs in TCMB will no longer be available?

    After both warm reset and POR the TCMB logs won't be available. 

    Regarding the JTAG connection in secure devices upon SBL boot failure, I need to check this.

  • Hi Addison,

    Regarding JTAG connection, the SBL needs to be understood by the RBL to make the JTAG available for connecting the debugger. So in HSSE device where SBL boot itself failed, this would not be possible. You can do the same procedure with a HSFS device and obtain the ROM logs.

  • Hi Addison,

    Adding to my previous reply, if the SBL boot by RBL failed before certificate validation then JTAG will be locked. But if the fail occurred after this phase, then JTAG would be unlocked. This is because JTAG lock/unlock is decided by the RBL based on SBL header.