Jetson box poweroff failed with error "mail tmo TX timeout"

I using a customer board, flashed with the latest JP5.0.2. JP was flashed to EMMC. there is a DP monitor connected to this device, when I finished all setting, and click “power off…” button, the DP monitor was off, but I found that power LED was not change(in previous version of JP, power LED will switch to red if power off success). and about 2-3 minites later, the device was power on automatically! it boot to login in interface again.

Then I unplug power and plug in power supply again, connect a debug port to check if it was really power off. try again and get such logs

t@ubuntu:~$ poweroff
Failed to set wall message, ignoring: Interactive authentication required.
Failed to power off system via logind: Interactive authentication required.
Failed to open initctl fifo: Permission denied
Failed to talk to init daemon.
t@ubuntu:~$ sudo poweroff
[sudo] password for t:
[   94.417081] Trying to unregister non-registered hwtime source
▒▒WARNING @ [platform/drivers/mailbox/mail_routing_layer/mail_routing_layer.c]: mail rx timeout
WARNING @ [platform/drivers/mailbox/tmo_link_provider/mail_tmo.c]: mail tmo TX timeout
WARNING @ [platform/drivers/mailbox/tmo_link_provider/mail_tmo.c]: mail tmo TX timeout
WARNING @ [platform/drivers/mailbox/tmo_link_provider/mail_tmo.c]: mail tmo TX timeout
WARNING @ [platform/drivers/mailbox/tmo_link_provider/mail_tmo.c]: mail tmo TX timeout
WARNING @ [platform/drivers/mailbox/tmo_link_provider/mail_tmo.c]: mail tmo TX timeout
[0000.024] W> RATCHET: MB1 binary ratchet value 4 is larger than ratchet level 2 from HW fuses.
[0000.032] I> MB1 (prd-version:
[0000.038] I> Boot-mode: Coldboot
[0000.041] I> Platform: Silicon
[0000.043] I> Chip revision : A02P
[0000.047] I> Bootrom patch version : 15 (correctly patched)
[0000.052] I> ATE fuse revision : 0x200
[0000.055] I> Ram repair fuse : 0x0
[0000.058] I> Ram Code : 0x0
[0000.061] I> rst_source: 0x2, rst_level: 0x1
[0000.066] I> Boot-device: QSPI (instance: 0)
[0000.070] I> Qspi flash params source = brbct
[0000.074] I> Qspi clock source : pllp
[0000.077] I> Qspi-0 initialized successfully

I will attach the debug port logs, dmesg log. and folder of /var/log later at attachment.
it looks like my power off was failed. then I try connect no DP monitor, only login in via debug port. execute “poweroff” command- on debug port, but the error still occurred. Is there any help for this issue?

This error not occurred on JP4.6. (1.3 MB)

log_debug_port.txt (458.4 KB)
log_dmesg.txt (104.0 KB)


Please remove all peripherals and leave only the uart cable to dump the uart log. Try to reproduce issue under such situation.

Thanks for your replies,

It reproduced without any device connected, this is the uart log.

log_debug_port_no_device_connected.txt (151.6 KB)

Is this issue able to reproduced on devkit?

we can only test on devkit tomorrow because of devkit was broken, new develop kit was in transit, will arrive tomorrow

not reproduce on devkit

Hi, if it can’t repro on devkit then there should be some design difference between custom board and devkit carrier. Please check that first to find out the difference based on P3509 schematic and Design Guide doc.

thanks for your response, we will check design guide doc and schematic.
On devkit, execute poweroff, it can find that “SHUTDOWN_REQ” from pin-233 has signal, while on our device, we didn’t detected this signal, we will try to find why this signal not send out.

hello wayne,

there is 3 lan connected to our pcie switch, and found one of switch can not remove by command cd :/sys/bus/pci/devices/0005:03:00.0 , echo 1 > remove. we thought this lan cause the error. when power off, the debug log:

WARNING @ [platform/drivers/mailbox/mail_routing_layer/mail_routing_layer.c]: mail rx timeout
WARNING @ [platform/drivers/mailbox/tmo_link_provider/mail_tmo.c]: mail tmo TX timeout

We can not found the file of platform/drivers/mailbox/tmo_link_provider/mail_tmo.c from source code of kernel. my first question was where can I found this file, from any patch?

You cannot find this file. This is bpmp log and bpmp is not open source.

Check your power sequence first. Remove all peripherals and test. See which one is causing issue.

Update for this issue

it’s not related to our power sequence or design, it was because of wol not supported on JP5.0.1DP, disable all wake-on-lan for each ethernet device.

may be wol will supported in fulture


@Cruise_Tang Thanks for your replay. Yes, in my test bed on JP5.02 GA, also I found such issue. If enabled ethernet wol by ethtool command, during system power off, it will print error message as “mail tmo TX timeout”.

@WayneWWW Does it mean JP5.02 can’t support WOL on ethernet, right? Thanks.

Sorry that this issue has been 3 weeks ago so I don’t know what you are talking about.

If you are asking the Realtek PHY on Jetson NX module supports WoL or not, then rel-35.1 does not.

But we have already fixed it. Next release would have it.
Or if you need a patch, we could share it too. Please file a new topic for it.

@WayneWWW Thanks for your quick reply. Just filed another BUG as follows:

please help to check. Thanks.

1 Like