Install a Yocto image on a Jetson Nano production module

Hi everyone,

I am currently trying to install a Yocto image on a production module. The module is installed on a “homecooked” board with a UART debug port wich I used to work with the Jetson board.

Board model: P3448 180-13448-DAAA-B01
JetPack version: JetPack_4.5.1

I have a problem with flashing the eMMC, so I have tried to flash the standard system (with JetPack) but it is not working too. I am not certain that my board is in recovery mod, you can see the result of the lsusb command:

Bus 001 Device 001: ID 1d6b:0002
Bus 002 Device 001: ID 1d6b:0003
Bus 001 Device 002: ID 0424:7500

In addition, when I tried to flash the board with the flash script, I got the error:

Error: Invalid target board - jetson-nano-emmc.

During my research to resolve my problem, I have try to plug the production board on a SDK Board I had (P3450 180-13449-DAAF-B01) but the Jetson Nano freeze when trying to boot.

I am a bit confused on what to try next so I hope someone here can help me to see more clearly how to preceed!

Thanks for reading me,
Best.

The board name is just *.conf file under your Linux_for_Tegra. Just share what you got there.

1 Like

I have realized my setup was incomplete on the board so I have copied the *. conf in the Linux_for_Tegra.
But that lead to a new problem:

readlink: invalid option – ‘e’
BusyBox v1.28.0 (2018-08-17 15:38:46 -00) multi-call binary.

Usage: readlink [-fnv] FILE

Display the value of a symlink

   -f      Canonicalize by following all symlinks
   -n      Don't add newline
   -v      Verbose

###############################################################################

L4T BSP Information:

Unknown Release

#############################################################################
Error: tegrarcm is missing.

I still doubt that I am in recovering mode, but I don’t know how to go into that mode, should I add some hardware on the home made board?

Best,
Clément

No, your problem is it looks like there are lots of files missing in your Linux_for_Tegra.

For example, you should not copy the config file from somewhere else to the Linux_for_Tegra… it should be there from the beginning.

Where is the original driver package Linux_for_Tegra coming from? What content was in it?

I recover the Linux_for_Tegra from the Nvidia SDK. Then I transfer it to my Jetson Nano with UART transfert.
Should I process differently?

Hi,

What did you transfer to jetson nano?? Not understand what you are doing.

Hi,

I feel like I don’t understand too !

  • I download the Linux_for_Tegra from Nvidia SDK (jetpack) for Jetson Nano.
  • I transfer the Linux_for_Tegra folder with UART on the Jetson Nano
  • I try to install with the flash.sh

Can you tell me if I am doing something wrong ?
I admit, I am lost on this one. My final goal is to flash my own generate image with Yocto and I was hoping to get a better understanding of how this board works by using the “classical” way.

Thanks for your help.

SDKmanager only able to run on x64 host machine. So the driver package Linux_for_Tegra will only run on x64 host machine too.

It will not work if you move it to jetson nano… the binaries only support x64 arch. The jetson nano is ARM based cpu which is arm64 architecthre…

1 Like

Refer to the quick start for understanding how to run the flash.sh on your x64 ubuntu host correctly.

Oh, I miss that information! Thanks to your help I have succeeded to flash the standard image, now I will try to flash my Yocto image!
Maybe you have some hints on this subject too?

Thanks a lot again for your help.

No, I have no experienc of Yocto image. Maybe you can check the posts shared by other users.