• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 98 subscribers
  • Views 369 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.

TDA4VM: TDA4 SOC hang up ( PSDKL 8.1 )

zmite
Prodigy 230 points
Part Number: TDA4VM


Hi experts,

    our board sometimes hang up when boot up.

    A72 and MCU1_0 communicate with dmsc failed.

    please check whether SYSFW run into abnormal status ?

    thanks.

This is A72 LOG:

[2022-07-05 19:45:03.646] [    0.000000] cma: Failed to reserve 512 MiB
[2022-07-05 19:45:04.805] [    1.247709] ti-sci 44083000.dmsc: Mbox timedout in resp(caller: ti_sci_cmd_clk_is_on+0x48/0xa8)
[2022-07-05 19:45:04.805] [    1.256400] ti-sci 44083000.dmsc: Mbox send fail -110
[2022-07-05 19:45:04.805] [    1.261439] ti-sci-clk 44083000.dmsc:clocks: is_prepared failed for dev=293, clk=2, ret=-110
[2022-07-05 19:45:05.796] [    2.271687] ti-sci 44083000.dmsc: Mbox timedout in resp(caller: ti_sci_cmd_put_device+0x18/0x28)
[2022-07-05 19:45:05.839] [    2.280458] ti-sci 44083000.dmsc: Mbox timedout in resp(caller: ti_sci_cmd_put_device+0x18/0x28)
[2022-07-05 19:45:05.839] [    2.280468] ti-sci 44083000.dmsc: Mbox timedout in resp(caller: ti_sci_cmd_clk_is_on+0x48/0xa8)
[2022-07-05 19:45:05.839] [    2.289229] ti-sci 44083000.dmsc: Mbox send fail -110
[2022-07-05 19:45:05.840] [    2.297901] ti-sci 44083000.dmsc: Mbox send fail -110
[2022-07-05 19:45:05.840] [    2.302936] ti-sci 44083000.dmsc: Mbox send fail -110

This is MCU1_0 LOG:


[2022-07-05 19:45:03.077] mcu_uart0 inited
[2022-07-05 19:58:23.183]     -->Sciclient_serviceSecureProxy(949): failed
[2022-07-05 19:58:23.214]     -->Sciclient_service(334): failed
[2022-07-05 19:58:23.245]     -->Sciserver_UserProcessMsg(593): failed
[2022-07-05 19:58:23.309]     -->Sciserver_processtask(270): failed
[2022-07-05 19:58:23.341] Failed to process message and failed to send nak response
[2022-07-05 19:58:23.404] Failed to process message and failed to send nak response
[2022-07-05 19:58:23.468] Failed to process message and failed to send nak response

  • 0
    TI__Guru**** 166330 points

    Hi,

    Is this a custom board? Or is this TI EVM?
    Are you also loading remote cores like R5Fs, C6x, C7x?
    Can you share the complete bootlog starting from the first line of SPL/SBL?

    - Keerthy

  • 0 in reply to Keerthy J
    Prodigy 230 points

    Hi experts,

        This is a custom board, not EVM.  and this issue is not easy to be reproduced,  about every 6 seconds to do a reset, and almost one or some days to be reproduced.

        the bootloader is SPL, SPL load R5Fs/C66/C7x, then Linux Kernel. I did not keep the full log, but the log is the same before kernel start. as kernel is quiet, you may refer :

    U-Boot SPL 2021.01 (Jul 05 2022 - 13:07:39 +0000)
    SYSFW ABI: 3.1 (firmware rev 0x0015 '21.9.1--v2021.09a (Terrific Lla')
    Trying to boot from SPI
    Starting ATF on ARM64 core...

    NOTICE:  BL31: v2.5(release):08.00.00.002-dirty
    NOTICE:  BL31: Built : 07:59:17, Jan 16 2022

    U-Boot SPL 2021.01 (Jul 05 2022 - 13:08:10 +0000)
    SYSFW ABI: 3.1 (firmware rev 0x0015 '21.9.1--v2021.09a (Terrific Lla')
    load rproc: ok
    [    0.000000] cma: Failed to reserve 512 MiB

      Thanks.

  • 0 in reply to Keerthy J
    Prodigy 230 points

    addtional:   we add a reset script in systemd's service script.  when linux boot up, it will execute the reset script . the reset script will delay 4 seconds and call reboot command.  this make the whole system reboot about every 6 seconds.

  • 0 in reply to zmite
    TI__Guru**** 166330 points

    Hi,

    In the bootargs can you set the 'boot' variable to reset. So we can test continuous reset from the U-Boot side.
    Does that work well on your board?
    It works for 100s of resets on the TDA4 EVM.

    Also Linux boot takes 20+ seconds using the SD boot mode how are you able to reset in 6 seconds. Can you
    help reproduce the issue on TDA4-EVM with your sequence?

    - Keerthy

  • 0 in reply to Keerthy J
    Prodigy 230 points

    we have optimized the bootup time, so that A72 will execute systemd quickly. during these days, this issue is reproduced 5 times. I found that the issue log almost show access clk 293 / 294 / 295 failed.  according to the device tree, I think clk 293/294/295 is used by serdes_wiz.

    100 times reset is not enough to reproduced this issue,  in my enviroment, it will take thounds times. and uboot does not bring up enough devices, maybe it's hard for uboot to reproduce this issue. but I have not try it yet.

  • 0 in reply to zmite
    TI__Guru**** 166330 points

    Hi,

    I do not have the setup to reproduce with optimized boot. I will try to reproduce with reboots on the default SDK.

    zmite said:

        please check whether SYSFW run into abnormal status ?

    It is not sysfw but the Power management & resource management is done by DM device manager firmware running on MCU1_0.
    If you have access to CCS please connect and check if you can connect to MCU1_0.

    - Keerthy