Nano fails to boot

Autonomous Machines Jetson & Embedded Systems Jetson Nano
#1

First time booting a nano and the boot process is taking forever, seems to hang at two places and does not complete initial setup. Nano eventually reboots and the process continues. I’m attached with usb keyboard, mouse and HDMI display as well as monitoring the serial port on a laptop. I’m using the 5v 4amp power supply. At the end of this post I’ve included two boot logs using the serial cable method. The first boot log starts from the beginning until we appear to stall for about 300 seconds:

[ 1.096115] tegradc tegradc.1: dpd enable lookup fail:-19
[ 1.262720] imx219 6-0010: imx219_board_setup: error during i2c read probe (-121)
[ 1.262748] imx219 6-0010: board setup failed

Next the Nano seems to struggle to boot and eventually cycles through 1, 2, 3, 4 and 5 of 5 start jobs for quite a while. This seems to fail and we end up with:

[ 1088.629883] INFO: task systemd-tmpfile:3794 blocked for more than 120 seconds.
[ 1088.638827] Not tainted 4.9.140-tegra #1
[ 1088.645045] “echo 0 > /proc/sys/kernel/hung_task_timeout_secs” disables this message.
[ 1088.654747] Kernel panic - not syncing: hung_task: blocked tasks
[ 1088.662530] CPU: 0 PID: 584 Comm: khungtaskd Not tainted 4.9.140-tegra #1
[ 1088.671103] Hardware name: NVIDIA Jetson Nano Developer Kit (DT)
[ 1088.678884] Call trace:
[ 1088.683087] [] dump_backtrace+0x0/0x198
[ 1088.690263] [] show_stack+0x24/0x30
[ 1088.697089] [] dump_stack+0x98/0xc0
[ 1088.703900] [] panic+0x11c/0x298
[ 1088.710443] [] watchdog+0x300/0x3b8
[ 1088.717242] [] kthread+0xec/0xf0
[ 1088.723772] [] ret_from_fork+0x10/0x40
[ 1088.730831] SMP: stopping secondary CPUs
[ 1088.736566] Kernel Offset: disabled
[ 1088.741798] Memory Limit: none
[ 1088.752949] Rebooting in 5 seconds…

And the process starts over. Attached are more complete boot logs.
nano_boot_log.txt (10.3 KB)
nano_boot_log_initial.txt (18.9 KB)
nano_boot_log_full.txt (62.2 KB)

Jetson nano can't boot up successfully
#2

Your log is being truncated by the forum. Please use a attachment.

Also, logs below are not fatal error. It is shown even with normal boot.

[ 1.096115] tegradc tegradc.1: dpd enable lookup fail:-19
[ 1.262720] imx219 6-0010: imx219_board_setup: error during i2c read probe (-121)
[ 1.262748] imx219 6-0010: board setup failed

This log is the one that causes problem. What release are you using? Are you using a nvidia devkit?

[ 1088.629883] INFO: task systemd-tmpfile:3794 blocked for more than 120 seconds.
[ 1088.638827] Not tainted 4.9.140-tegra #1
[ 1088.645045] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1088.654747] Kernel panic - not syncing: hung_task: blocked tasks
[ 1088.662530] CPU: 0 PID: 584 Comm: khungtaskd Not tainted 4.9.140-tegra #1
[ 1088.671103] Hardware name: NVIDIA Jetson Nano Developer Kit (DT)
[ 1088.678884] Call trace:
[ 1088.683087] [<ffffff800808bdb8>] dump_backtrace+0x0/0x198
[ 1088.690263] [<ffffff800808c37c>] show_stack+0x24/0x30
[ 1088.697089] [<ffffff800845d820>] dump_stack+0x98/0xc0
[ 1088.703900] [<ffffff80081c2198>] panic+0x11c/0x298
[ 1088.710443] [<ffffff8008181f90>] watchdog+0x300/0x3b8
[ 1088.717242] [<ffffff80080dbee4>] kthread+0xec/0xf0
[ 1088.723772] [<ffffff8008083850>] ret_from_fork+0x10/0x40
[ 1088.730831] SMP: stopping secondary CPUs
[ 1088.736566] Kernel Offset: disabled
[ 1088.741798] Memory Limit: none
[ 1088.752949] Rebooting in 5 seconds..
#3

Thanks for the reply. I have moved the boot logs to an attachment to the original post. Additionally, I added one here which is probably the better one to look at. Its a full boot cycle with multiple forced reboots captured.

I am using the latest image from the NVIDIA Jetson Nano website. I’ve burned the image to two different SDCards. One from Linux, one from MacOS. Both with the same results.

The link to the SDcard image I use is here: https://developer.nvidia.com/jetson-nano-sd-card-image
nano_boot_log_full.txt (62.2 KB)

#4

i got a problem like this and i don’t know why

1.096115] tegradc tegradc.1: dpd enable lookup fail:-19
[ 1.262720] imx219 6-0010: imx219_board_setup: error during i2c read probe (-121)
[ 1.262748] imx219 6-0010: board setup failed
not writing through dangling symlink ‘etc/resolv.conf’

[FAILED] failed to start load/save Random seed

[FAILED] failed to start udev kernel device manager

then completes installations in another lines, display goes off then repeat the same steps with the same errors

although i installed the sd card image and removed it many times

#5

Maybe you could use sdkmanager to install it instead.
https://developer.nvidia.com/embedded/jetpack

There are lots of possible reason that cause sdcard boot failure. But sdkmanager should be the most stable one.

#6 
not writing through dangling symlink 'etc/resolv.conf'
[FAILED] failed to start load/save Random seed
[FAILED] failed to start udev kernel device manager

If your device keeps booting again and again, probably also hits kernel panic. You could remove “quiet” keyword in your rootfs/boot/extlinux/extlinux.conf and the serial console will dump more verbose log.

However, I think the fastest way to resolve this issue is try sdkmanager.

#7

Following your suggestion I’m trying to use SDK Manager to flash my Nano. It appears, in order for SDK Manager to fully flash the Nano, the Nano needs to boot into Ubuntu. The problem is, this Nano never makes it to the login screen in Ubuntu.

The flashing appears to complete to the point of rebooting the nano…and the SDK Manager is at the step “SDK Manager is about to install SDK components on your Jetson Nano (Developer Kit version)” although there are a lot of watchdog errors.

I can’t use automatic mode as that requires a working Linux system on the Nano and manual mode results in the same boot issue as before.

Here is a log: https://gist.github.com/jmhebert/be617a99d799b500279608c09114561a

#8

Hi Jeffrey,

We need to enable more log. Current log is not enough to debug.

You could use either way in below.

method 1. Remove the “quiet” keyword in extlinux.conf under your Linux_for_Tegra/rootfs/boot/extlinux and re-flash your board. Please do not use sdkmanager to flash in this step. Just use the flash.sh. If you don’t know how to use it, please refer to quick start guide on download center.

method 2. Plug the micro usb cable on the port and unplug any monitor cable (HDMI/DP). Use oem-config to configure your user account through UART with interface /dev/ttyACM0.

https://docs.nvidia.com/jetson/l4t/index.html#page/Tegra%20Linux%20Driver%20Package%20Development%20Guide%2Fflashing.html%23wwpID0E0JD0HA

-> Headless Mode Flow in oem-config

You should be able to access the UART or SSH to device after this configuration. Then, plug the HDMI cable the dump the kernel log with command dmesg.

#9

I was not able to complete all of your steps below to get to oem-config.

I used method 1. I edited the extlinux.conf file and removed quiet by commenting out the line so that it looks like this now:

# APPEND ${cbootargs} quiet

I then flashed the nano using the flash.sh script. The log of the command line for that flash is here: https://gist.github.com/jmhebert/f1cf849c48ba610ef55b317c786f1814

I had the UART capturing the boot sequence on the Serial Debug Console on J44. I did not command the reboot - it just happened, and the device continued to reboot continually after about 2640 seconds. I’ve captured the boot sequence: https://gist.github.com/jmhebert/681ae906cc5d4c724b85e2ea4200003a

What I can’t seem to figure out is how to get oem-config to run.

Do you have any more detailed steps as to how oem-config can get started? Am I able to use oem-config from the serial debug console on J44 or do I have to use the microUSB cable? Specificaly, does Linux need to be running on the nano for oem-config to be running? I don’t think I am getting to the point where Linux is running. If I can ssh into the device, how do I determine the IP address?

Thanks
Jeff

#10

Hi,

Please do not remove the whole APPEND in extlinux.conf. Cbootargs is also important for boot up. Just need to delete the quiet.

I had the UART capturing the boot sequence on the Serial Debug Console on J44. I did not command the reboot - it just happened, and the device continued to reboot continually after about 2640 seconds. I’ve captured the boot sequence: https://gist.github.com/jmhebert/681ae906cc5d4c724b85e2ea4200003a

I am not sure if this error comes from the missing of cbootargs or not. Does your nano alwasy reboot before we start these test methods?

Do you have any more detailed steps as to how oem-config can get started? Am I able to use oem-config from the serial debug console on J44 or do I have to use the microUSB cable? Specificaly, does Linux need to be running on the nano for oem-config to be running? I don’t think I am getting to the point where Linux is running. If I can ssh into the device, how do I determine the IP address?

It should be the micro usb port. And yes, Linux needs to be running on nano for oem-config to run. It cannot run under such case in which kernel panic causes reboot. That is why you need to reply my previous question first. If this issue was already there, then I would suggest you to change another SDard to test.

#11

Thank you. I fixed the APPEND statement. I am linking the log from the flashing procedure here:

https://gist.github.com/jmhebert/45ea00a2a9975744b510a581071f9bdf

and the boot sequence from the Serial Debug Console here:

https://gist.github.com/jmhebert/9baec075e8addb3b7e6ff3475e5f40bd

After flashing my nano always boots itself automatically. If you notice the boot results in a kernel panic at 1088 and a subsequent reboot (partially captured reboot). This pattern continues indefinitely.

I will try another SDcard and will identify any different results.

#12

Using a different SDcard from a different manufacturer worked. I have two Kingston 32GB SDCards - both achieve the same (failed) results. Switched to a 16GB Samsung SDCard and booted into Linux successfully, and was able to configure the environment with oem-config

#13

I have used 2 PNY microSD 64G from the Nvidia re-seller and Sandisk 32G and all of them failed to boot after few days of working , its very frustrated to know that this happens a lot , what is the reason , is sumsung is better then the others ??(I am using Jetpack 4.3 for nano image from the official site)

#14

did you solve this? i am running into this same issue. every few days the it wont boot up.

#15

Hi, i am using SanDisk SD card

#16

Same Sandisk extreme 64gb uhs-3