Xavier NX can not boot after shut down in J5.0.2

Autonomous Machines Jetson & Embedded Systems Jetson Xavier NX
1

Hi,

Platform : Xavier nx
Jetpack : 5.0.2

I’m using our carrier board and can not boot after shut down

The power light is off but the uart log still print error message and it can not boot by push the power button

After a few minute, the system will reboot automatically

This situation similar with this topic, but seems no solution yet

The error message :

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
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
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
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

And this error not occurred on JP4.6.2 on our carrier board

As the article mentioned above, the mail_tmo.c is not found in kernel source because of that’s a bpmp log which is not open source

And the power sequence is okay since it was work on j4.6.2

The wol also check and the patches supply by WayneWWW are already in j5.0.2 jetpack

Is there any suggestion to check the route cause?

3

Does this behavior reproduce on devkit?

4

No, devkit has no mail error

7

Hi,

To clarify if this issue is related to WoL. Could we disable ethernet node in kernel first and see if this could bypass error?

8

Hi WayneWWW,

Sorry late for reply

I have tried the default image with disabling the ethernet(I210) node pcie@141a0000

The error is not occurred and device can boot and shutdown normally

9

Hi,

For this issue, we may only share the workaround.
The official solution needs to ask from Intel as this is vendor drivers issue.

10

Okay, thanks for your great help

11

Hi,

The workaround is comment out the shutdown handler in our pcie driver (kernel/kernel-5.10/drivers/pci/controller/dwc/pcie-tegra194.c)

.shutdown = tegra_pcie_dw_shutdown,

2 Likes
13

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.