Cuda Setup Failure on Xavier using SDK Manager

Autonomous Machines Jetson & Embedded Systems Jetson AGX Xavier
1

Hello,

I am using Ubuntu 18.04 as a host machine on Jetson Xavier, I had multiple problems flashing the Xavier but solved them and managed to finally flash the jetson successfully after installing python 2.7 on the host machine.

Currently I am trying to install the jetson SDK software on Xavier however, everytime there is an error saying that the local apt repository has some sort of broken packages because of CUDA and that is error it says.

18:40:49 INFO : CUDA Toolkit for L4T : gnupg is already the newest version (2.2.4-1ubuntu1.2).
18:40:49 INFO : CUDA Toolkit for L4T : You might want to run ‘apt --fix-broken install’ to correct these.
18:40:49 INFO : CUDA Toolkit for L4T : The following packages have unmet dependencies:
18:40:49 INFO : CUDA Toolkit for L4T : nvidia-container-csv-cuda : Depends: cuda-toolkit-10-0 (= 10.0.326-1) but it is not going to be installed
18:40:50 INFO : CUDA Toolkit for L4T : E: Unmet dependencies. Try ‘apt --fix-broken install’ with no packages (or specify a solution).
18:40:50 INFO : CUDA Toolkit for L4T : Unhandled error when running sudo apt-get install -y gnupg
18:40:50 INFO : CUDA Toolkit for L4T : exit status 1
18:40:50 INFO : CUDA Toolkit for L4T : [ Package Install Finished with Error ]
18:40:50 INFO : CUDA Toolkit for L4T : [ 32.00 KB used. Disk Avail: 834.01 GB ]
18:40:50 INFO : CUDA Toolkit for L4T : [ NV_L4T_CUDA_TARGET_POST_INSTALL_COMP Install took 54s ]
18:40:50 ERROR : CUDA Toolkit for L4T : Run commands failed at step Install: command scp -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no /tmp/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh flash@192.168.55.1:~; ssh -t -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no flash@192.168.55.1 “~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh && rm -f ~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh” finished with error
18:40:50 INFO : CUDA Toolkit for L4T :
18:40:50 ERROR : CUDA Toolkit for L4T : command terminated with error
18:40:50 ERROR : CUDA Toolkit for L4T : install ‘CUDA Toolkit for L4T’ failure, command < cd ‘/home/mahmoud/nvidia/nvidia_sdk/JetPack_4.2_Linux_P2888’ ; ‘/opt/nvidia/sdkmanager/resources/app/output/installUtils/adapter’ -a=‘install’ -c=‘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’ -i=‘NV_L4T_CUDA_TARGET_POST_INSTALL_COMP’ -v=‘10.0’ -l=‘/home/mahmoud/nvidia/nvidia_sdk/JetPack_4.2_Linux_P2888’ -L=‘/home/mahmoud/.nvsdkm/logs/JetPack_4.2_Linux’ -p=‘2082’ -t=‘P2888’ -e=‘/home/mahmoud/Downloads/nvidia/sdkm_downloads/cuda-repo-l4t-10-0-local-10.0.166_1.0-1_arm64.deb’ > terminated with error
18:40:50 ERROR : Unavailable components detected, please check log
18:40:50 ERROR : cuDNN on Target : cuDNN on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
18:40:50 ERROR : TensorRT on Target : TensorRT on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
18:40:50 ERROR : OpenCV on Target : OpenCV on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
18:40:50 ERROR : VisionWorks on Target : VisionWorks on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
18:40:50 ERROR : Multimedia API : Multimedia API Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,

I am currently stuck and I don’t know what solution should I go with because there are a lot dependencies and currently I removed python 2.7 as I thought it was dependency problem between python 2.7 installed to get the sdk manager to flash and 3.6 as default on the system.

Can anyone suggest what I should do? also, it seems the problem has transferred to the Kit as it shows the same error.

There is also a sign of connection failure between the jetson and the host machine.

2

Try this first on the host PC:

sudo apt update
sudo apt --fix-broken install
sudo apt-get install libgconf-2-4 libcanberra-gtk-module
sudo apt-get install python
3

I tried the commands, did the updates, and all things are up to date and nothing is broken.

The install now previously was failing at 30% now it fails at 40% with the errors below:

12:28:14 INFO : CUDA Toolkit for L4T : Fetched 1,907 kB in 4s (524 kB/s)
12:28:18 INFO : CUDA Toolkit for L4T : Reading package lists…
12:28:18 INFO : CUDA Toolkit for L4T : *begin sudo apt-get install -y gnupg, 5
12:28:18 INFO : CUDA Toolkit for L4T : Reading package lists…
12:28:19 INFO : CUDA Toolkit for L4T : Building dependency tree…
12:28:19 INFO : CUDA Toolkit for L4T : Reading state information…
12:28:19 INFO : CUDA Toolkit for L4T : gnupg is already the newest version (2.2.4-1ubuntu1.2).
12:28:19 INFO : CUDA Toolkit for L4T : You might want to run ‘apt --fix-broken install’ to correct these.
12:28:19 INFO : CUDA Toolkit for L4T : The following packages have unmet dependencies:
12:28:20 INFO : CUDA Toolkit for L4T : nvidia-container-csv-cuda : Depends: cuda-toolkit-10-0 (= 10.0.326-1) but it is not going to be installed
12:28:20 INFO : CUDA Toolkit for L4T : E: Unmet dependencies. Try ‘apt --fix-broken install’ with no packages (or specify a solution).
12:28:20 INFO : CUDA Toolkit for L4T : Unhandled error when running sudo apt-get install -y gnupg
12:28:20 INFO : CUDA Toolkit for L4T : exit status 1
12:28:20 INFO : CUDA Toolkit for L4T : [ Package Install Finished with Error ]
12:28:20 INFO : CUDA Toolkit for L4T : [ 52.00 KB used. Disk Avail: 832.83 GB ]
12:28:20 INFO : CUDA Toolkit for L4T : [ NV_L4T_CUDA_TARGET_POST_INSTALL_COMP Install took 1m36s ]
12:28:20 ERROR : CUDA Toolkit for L4T : Run commands failed at step Install: command scp -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no /tmp/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh flash@192.168.55.1:~; ssh -t -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no flash@192.168.55.1 “~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh && rm -f ~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh” finished with error
12:28:20 INFO : CUDA Toolkit for L4T :
12:28:20 ERROR : CUDA Toolkit for L4T : command terminated with error
12:28:20 ERROR : CUDA Toolkit for L4T : install ‘CUDA Toolkit for L4T’ failure, command < cd ‘/home/mahmoud/nvidia/nvidia_sdk/JetPack_4.2_Linux_P2888’ ; ‘/opt/nvidia/sdkmanager/resources/app/output/installUtils/adapter’ -a=‘install’ -c=‘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’ -i=‘NV_L4T_CUDA_TARGET_POST_INSTALL_COMP’ -v=‘10.0’ -l=‘/home/mahmoud/nvidia/nvidia_sdk/JetPack_4.2_Linux_P2888’ -L=‘/home/mahmoud/.nvsdkm/logs/JetPack_4.2_Linux’ -p=‘14530’ -t=‘P2888’ -e=‘/home/mahmoud/Downloads/nvidia/sdkm_downloads/cuda-repo-l4t-10-0-local-10.0.166_1.0-1_arm64.deb’ > terminated with error
12:28:20 ERROR : Unavailable components detected, please check log
12:28:20 ERROR : cuDNN on Target : cuDNN on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
12:28:20 ERROR : TensorRT on Target : TensorRT on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
12:28:20 ERROR : OpenCV on Target : OpenCV on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
12:28:20 ERROR : VisionWorks on Target : VisionWorks on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
12:28:20 ERROR : Multimedia API : Multimedia API Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,

and still having the connection failure problem on both host and jetson.

4

Can you try those same commands from #2 on the Xavier? After this try just CUDA install to the Xavier without flashing again.

5

the problem is solved now thanks a lot,

so regarding the host I did what you recommended

then at the Xavier side I did the following

sudo apt-get repository universe
sudo apt-get update
sudo apt-get install freeglut3 freeglut3-dev libxi-dev libmu-dev
sudo apt --fix-broken install

while taking care of firewall of the home network …

6

Knowing you all are past your problems I had a similar issue using SDK Manager 1.1.0.6343

I am a little behind just getting into these agx’s

Was able to flash the host but CUDA Toolkit would not install because the time was getting synced wrong, turns out the repos do not like it if your time is off (it just time!)

Ended up changing the script that does the sync to use ntpdate -u [timeserver]

you do have to do the apt-get update && apt-get install ntp ntpdate on the host after flashing (ssh)

CUDA Toolkit installed fine, just an fyi

host files the SDK generated in /tmp in my case:
tmp_NV_DATETIME_SETUP_TARGET_COMP.sh (edited) tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh tmp_NV_L4T_FILE_SYSTEM_AND_OS_T194_COMP.sh tmp_NV_L4T_TENSORRT_DLA_TARGET_POST_INSTALL_COMP.sh
tmp_NV_DEVICE_MODE_SETUP_IMAGE_COMP.sh tmp_NV_L4T_CUDNN_TARGET_POST_INSTALL_COMP.sh tmp_NV_L4T_FLASH_XAVIER_WITH_OS_IMAGE_COMP.sh tmp_NV_L4T_VISIONWORKS_TARGET_POST_INSTALL_COMP.sh
tmp_NV_DEVICE_MODE_SETUP_TARGET_COMP.sh tmp_NV_L4T_DRIVERS_T194_COMP.sh tmp_NV_L4T_OPENCV_TARGET_POST_INSTALL_COMP.sh