FLASHING FAILED - Xavier - JetsPack 4.4 DP - SDKManager

Hello,

I just spent a day trying to flash the XAVIER to put the 4.4 but at the same time doing a factory original restoration. I can’t do it. On the SDK Manager, I always have the same error at the time of FLASH.
I confirm that I respect the documentation:
OFF power
connecting USB to the host
Click on Force Recovery
Click Power and Force Receive
and hold 2 buttons.

I have to do it wrong, where is my mistake?
It is the first time that I do this, because the first time it was summer 2019 I had just received it, and it worked perfectly.

Can you confirm that the procedure formats Xavier well because I would also like to leave on a clean basis.

Logs:

  • 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : exec_command: /tmp/tmp_NV_L4T_FLASH_XAVIER_WITH_OS_IMAGE_COMP.sh
  • 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : sudo ./flash.sh jetson-xavier mmcblk0p1
  • 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : ###############################################################################
  • 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : # L4T BSP Information:
  • 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : # R32 , REVISION: 4.2
  • 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : ###############################################################################
  • 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : # Target Board Information:
  • 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : # Name: jetson-xavier, Board Family: t186ref, SoC: Tegra 194,
  • 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : # OpMode: production, Boot Authentication: NS,
  • 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : ###############################################################################
  • 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : copying soft_fuses(/home/hergo/nvidia/nvidia_sdk/JetPack_4.4_DP_Linux_DP_JETSON_AGX_XAVIER/Linux_for_Tegra/bootloader/t186ref/BCT/tegra194-mb1-soft-fuses-l4t.cfg)… done.
  • 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : ./tegraflash.py --chip 0x19 --applet “/home/hergo/nvidia/nvidia_sdk/JetPack_4.4_DP_Linux_DP_JETSON_AGX_XAVIER/Linux_for_Tegra/bootloader/mb1_t194_prod.bin” --skipuid --soft_fuses tegra194-mb1-soft-fuses-l4t.cfg --bins “mb2_applet nvtboot_applet_t194.bin” --cmd “dump eeprom boardinfo cvm.bin;reboot recovery”
    *** 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : Reading board information failed.**
    *** 11:20:52 ERROR : Flash Jetson AGX Xavier 16GB/32GB : /usr/bin/env:**
    *** 11:20:52 ERROR : Flash Jetson AGX Xavier 16GB/32GB : ‘python’: No such file or directory**
    *** 11:20:52 ERROR : Flash Jetson AGX Xavier 16GB/32GB : [exec_command]: /bin/bash -c /tmp/tmp_NV_L4T_FLASH_XAVIER_WITH_OS_IMAGE_COMP.sh; [error]: exit status 1**
    *** 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : [ Component Install Finished with Error ]**
    *** 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : [ 12.00 KB used. Disk Avail: 43.52 GB ]**
    *** 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB : [ NV_L4T_FLASH_XAVIER_WITH_OS_IMAGE_COMP Install took 0s ]**
    *** 11:20:52 ERROR : Flash Jetson AGX Xavier 16GB/32GB : [error]: Run commands failed at step Install: command /tmp/tmp_NV_L4T_FLASH_XAVIER_WITH_OS_IMAGE_COMP.sh finished with error**
    *** 11:20:52 INFO : Flash Jetson AGX Xavier 16GB/32GB :**
    *** 11:20:52 ERROR : Flash Jetson AGX Xavier 16GB/32GB : command terminated with error**
    *** 11:20:52 ERROR : Flash Jetson AGX Xavier 16GB/32GB : Install ‘Flash Jetson AGX Xavier 16GB/32GB’ failure, command < using adapter to install NV_L4T_FLASH_XAVIER_WITH_OS_IMAGE_COMP@JETSON_AGX_XAVIER to /home/hergo/nvidia/nvidia_sdk/JetPack_4.4_DP_Linux_DP_JETSON_AGX_XAVIER > terminated with error.**
  • 11:20:52 INFO : Device Mode Host Setup in Target SDK : Install ‘Device Mode Host Setup in Target SDK’ Skipped
  • 11:20:52 INFO : DateTime Target Setup : Install ‘DateTime Target Setup’ Skipped
  • 11:20:52 INFO : CUDA Toolkit for L4T : Install ‘CUDA Toolkit for L4T’ Skipped
  • 11:20:52 INFO : cuDNN on Target : Install ‘cuDNN on Target’ Skipped
  • 11:20:52 INFO : TensorRT on Target : Install ‘TensorRT on Target’ Skipped
  • 11:20:52 INFO : OpenCV on Target : Install ‘OpenCV on Target’ Skipped
  • 11:20:52 INFO : VisionWorks on Target : Install ‘VisionWorks on Target’ Skipped
  • 11:20:52 INFO : VPI on Target : Install ‘VPI on Target’ Skipped
  • 11:20:52 INFO : NVIDIA Container Runtime with Docker integration (Beta) : Install ‘NVIDIA Container Runtime with Docker integration (Beta)’ Skipped
  • 11:20:52 INFO : Multimedia API : Install ‘Multimedia API’ Skipped
  • 11:20:52 INFO : DeepStream : Install ‘DeepStream’ Skipped
  • 11:20:53 INFO : All done!

Does this install python?

sudo apt update
sudo apt-get install python

If so, then try flashing again.

I had a few issues trying to upgrade from pre jetpack 4.3 to 4.4 using sdk manager. It seemed easier to upgrade to jetpack 4.3 first, then use apt to upgrade to 4.4 from the jetson:

Upgrading JetPack
https://docs.nvidia.com/jetson/jetpack/install-jetpack/index.html#upgrade-jetpack

##COPIED FROM DOC##
Starting with JetPack 4.4, upgrading from previous JetPack versions to the latest JetPack version can be achieved using a package management tool like apt*. Follow the below steps to perform the upgrade:

To upgrade to new minor release

1.Open the apt source configuration file in a text editor, for example:

$ sudo vi /etc/apt/sources.list.d/nvidia-l4t-apt-source.list

  1. Change the repository name and download URL in the deb commands.

The original commands are:

deb https://repo.download.nvidia.com/jetson/common r32 main

deb Index r32 main

Where is identifies the platform’s processor:

•t186 for Jetson TX2 series

•t194 for Jetson AGX Xavier series or Jetson Xavier NX

•t210 for Jetson Nano or Jetson TX1

  1. Change the repository name from r32 to r32.4, and to the appropriate processor name for your platform. If your platform is Jetson Xavier NX, for example:

deb https://repo.download.nvidia.com/jetson/common r32.4 main

deb Index t194 r32.4 main

  1. Save and close the source configuration file.

5.Enter the commands:

$ sudo apt update

$ sudo apt dist-upgrade

If apt prompts you to choose a configuration file, reply Y for yes (to use the NVIDIA updated version of the file).

.Reboot your Jetson device when the upgrade is finished.

  1. Once L4T is upgraded, reinstall the nvidia-jetpack metapackage with the following:
    sudo apt-get install nvidia-jetpack

In case the disk space is limited (for example, if you are using Jetson Nano with 16GB microSD card), follow the below steps to upgrade:

  1. Remove all JetPack compute components (the example shown below is for upgrading from JetPack 4.3).
    sudo apt autoremove --purge nvidia-container-csv-cuda libopencv-python libvisionworks-sfm-dev libvisionworks-dev libvisionworks-samples libnvparsers6 libcudnn7-doc libcudnn7-dev libnvinfer-samples libnvinfer-bin nvidia-container-csv-cudnn libvisionworks-tracking-dev vpi-samples tensorrt libopencv libnvinfer-doc libnvparsers-dev libcudnn7 libnvidia-container0 cuda-toolkit-10-0 nvidia-container-csv-visionworks graphsurgeon-tf libopencv-samples python-libnvinfer-dev libnvinfer-plugin-dev libnvinfer-plugin6 nvidia-container-toolkit libnvinfer-dev libvisionworks libopencv-dev nvidia-l4t-jetson-multimedia-api vpi-dev vpi python3-libnvinfer python3-libnvinfer-dev opencv-licenses nvidia-container-csv-tensorrt libnvinfer6 libnvonnxparsers-dev libnvonnxparsers6 uff-converter-tf nvidia-docker2 libvisionworks-sfm libnvidia-container-tools nvidia-container-runtime python-libnvinfer libvisionworks-tracking

  2. If JetPack was installed using SDK Manager, remove the local repo (the example shown below is for upgrading from JetPack 4.3).
    sudo apt purge cuda-repo-l4t-10-0-local-10.0.326 libvisionworks-repo libvisionworks-sfm-repo libvisionworks-tracking-repo

  3. Run apt-clean.
    sudo apt clean

Hello,

Thank all ! All is OK. Restore / Flash / 4.4. OK :)

I would like understand why Nvidia don’t make by defaut OpenCV with GPU Cuda… it’s very strang for machine with GPU ;)

NVIDIA Jetson AGX Xavier - Jetpack 4.4 DP [L4T 32.4.2]

  • Up Time: 0 days 1:49:21 Version: 2.0.4
  • Jetpack: 4.4 DP [L4T 32.4.2] Author: Raffaello Bonghi
  • Board: e-mail: raffaello@rnext.it
    • Board(s): P2888-0001, P2822-0000
    • Code Name: galen
    • GPU-Arch: 7.2
    • SOC: tegra194 - ID:25
    • Type: AGX Xavier
  • Libraries:
    • CUDA: 10.2.89
    • OpenCV: 4.1.1 compiled CUDA: NO
    • TensorRT: 7.1.0.16
    • VisionWorks: 1.6.0.501
    • cuDNN: 8.0.0.145
  • Hostname: xavier
  • Interfaces:
    • eth0: 192.168.0.8

Hi Hergo,

I would like understand why Nvidia don’t make by defaut OpenCV with GPU Cuda… it’s very strang for machine with GPU ;)
Please refer to:
Is there a plan on next Jetpack release for opencv (cuda) support?

Hello, thank you.
But I did it myself. The subject is that I do not understand why NVIDIA does not release a CUDA compatible product when the main purpose of the sale of its product is precisely to exploit this.

NVIDIA Jetson AGX Xavier - Jetpack 4.4 DP [L4T 32.4.2]

 - Up Time:        1 days 0:50:36                                                     Version: 2.0.4
 - Jetpack:        4.4 DP [L4T 32.4.2]                                                 Author: Raffaello Bonghi
 - Board:                                                                              e-mail: raffaello@rnext.it
   * Board(s):     P2888-0001, P2822-0000
   * Code Name:    galen
   * GPU-Arch:     7.2
   * SN:           
   * SOC:          tegra194 - ID:25
   * Type:         AGX Xavier
 - Libraries:
   * CUDA:         10.2.89
   * OpenCV:       4.3.0-dev compiled CUDA: YES
   * TensorRT:     7.1.0.16
   * VisionWorks:  1.6.0.501
   * cuDNN:        8.0.0.145
 - Hostname:       xavier
 - Interfaces:
   * eth0:         
   * docker0:

HI,

Actually we only provide the minimum requirement for those sdk which needs the opencv.
As openCV is not maintained by NV directly, we cannot always make sure their cuda sample is always in good status. Honestly speaking, we once enabled some cuda samples but turned out it had some error.

Exactly the same error when flashing Xavier.

Does Nvidia have an official fix yet?

Hi,

We don’t have any official fix for flash problem.
Actually everyone has different cause of flash error when they hit it. The first comment in this post is lacking of python installed on his host machine. Thus, don’t know what error you are trying to point out here.

I got the same error when I try to flash 4.4DP into a new box.
It looks very strange since the box have python2.7 installed.

As commented by linuxdev, I tried to install python using apt-get but nothing happened because it has been already installed.

I would suggest you could read this page and use flash.sh to do the work.

https://elinux.org/Jetson/General_debug

Honestly speaking, sdkmanager is easy to get failure so I don’t like to use it to debug.