ERROR: could not run targetfstool [ERROR:utilities:Could not find /targetfs/root/apt-repos/binary/cuda-repo-ubuntu1804-*_arm64.deb]

Autonomous Vehicles DRIVE AGX Xavier DRIVE AGX Xavier General
1

Please provide the following info (check/uncheck the boxes after creating this topic):
Software Version
DRIVE OS Linux 5.2.6
DRIVE OS Linux 5.2.6 and DriveWorks 4.0
DRIVE OS Linux 5.2.0
DRIVE OS Linux 5.2.0 and DriveWorks 3.5
NVIDIA DRIVE™ Software 10.0 (Linux)
NVIDIA DRIVE™ Software 9.0 (Linux)
other DRIVE OS version
other

Target Operating System
Linux
QNX
other

Hardware Platform
NVIDIA DRIVE™ AGX Xavier DevKit (E3550)
NVIDIA DRIVE™ AGX Pegasus DevKit (E3550)
other

SDK Manager Version
1.9.1.10844
other 2.2.0.12021

Host Machine Version
native Ubuntu 18.04
other

Still trying to build the Drive Software 10.0 (Rev. 2) Linux for Drive AGX Pegasus (target device).

Now getting an error in the NV_DRIVE_PLATFORM_RFS_TARGET_COMP.log is as follows:

07:28:48.584 - Info: INFO:targetfstool_5_1_6_x_avrfs:installing cuda on targetfs
07:28:48.584 - Info: ERROR:utilities:Could not find /targetfs/root/apt-repos/binary/cuda-repo-ubuntu1804-*_arm64.deb

Full logs:

SDKM_logs_DRIVE_Software_10.0_(rev._2)_Linux_for_DRIVE_AGX_Pegasus_XT_2024-11-11_07-24-41.zip (2.0 MB)

Note this command in docker_run_targetfstool.sh:

# Then, run the targetfstool with the supplied tool file and targetfs.
info "Running $tool to targetfs at $targetfs."
#shellcheck disable=SC2046
docker run --privileged --rm \
    --network=host \
    --mount=type=bind,source="$tool_dir",target=/tool_dir \
    --mount=type=bind,source="$targetfs",target=/targetfs \
    "${volume_args[@]}" \
    "$docker_image" \
    /targetfstool/targetfstool.py \
        --loglevel="$( [ $verbose == 0 ] && echo WARN || echo "$loglevel")" \
        $( [ $nologcolors == 1 ] && echo --nologcolors ) \
        --toolfile="/tool_dir/$tool_file" \
        --toolfileargs="$tool_file_args" \
        /targetfs \
        run_full_tool || failure "could not run targetfstool"

The /targetfs is /home/nvidia/nvidia/nvidia_sdk/DRIVE_Software_10.0_Linux_OS_DRIVE_AGX_PEGASUS_XT/DRIVEOS/drive-t186ref-linux/targetfs

Note that root directory does not contain apt-repos/binary/cuda-repo-ubuntu1804-*_arm64.deb as expected:

ls /home/nvidia/nvidia/nvidia_sdk/DRIVE_Software_10.0_Linux_OS_DRIVE_AGX_PEGASUS_XT/DRIVEOS/drive-t186ref-linux/targetfs/root/
drive-setup.sh  move10GbePhy_A.sh  move10GbePhy_B.sh  pdk-scripts  run-once

Instead, the deb file exists in a different path:

$ sudo find / -name cuda-repo-ubuntu1804-*_arm64.deb
/home/nvidia/nvidia/nvidia_sdk/DRIVE_Software_10.0_Linux_OS_DRIVE_AGX_PEGASUS_XT/DRIVEOS/drive-t186ref-cuda/target/cuda-repo-ubuntu1804-10-2-local-10.2.19_1.0-1_arm64.deb
2

@SivaRamaKrishnaNV seeing if you can help me on this one. Note this is different to the similar issue previously solved.

4

Dear @sleff,
Are you using latest sdkmanager to flash the target?

5

Yes 2.2.0.12021

6

Dear @sleff,
In previous topic, I notice you are using sdkmanager inside a docker container. Are you still doing the same? if so, can you try flashing using sdkmanager on ubuntu 18.04 host

7

Hi @SivaRamaKrishnaNV

I have tried from both docker container and native ubuntu 18.04 host.

The error is the same:

11:18:43.483 - Info: INFO:targetfstool_5_1_6_x_avrfs:installing cuda on targetfs
11:18:43.483 - Info: ERROR:utilities:Could not find /targetfs/root/apt-repos/binary/cuda-repo-ubuntu1804-*_arm64.deb
...
11:18:43.557 - Info: ERROR:__main__:
11:18:43.557 - Info: Traceback (most recent call last):
11:18:43.557 - Info: File "/targetfstool/targetfstool.py", line 217, in <module>
11:18:43.557 - Info: main()
11:18:43.557 - Info: File "/targetfstool/targetfstool.py", line 201, in main
11:18:43.557 - Info: args['func'](buildtracker, args['targetfs'], '', args)
11:18:43.557 - Info: File "/targetfstool/targetfstool.py", line 84, in _run_full_tool_func
11:18:43.557 - Info: run_targetfstool(buildtracker, targetfs_dir, targetfs_variant, args)
11:18:43.557 - Info: File "/targetfstool/targetfstool.py", line 110, in run_targetfstool
11:18:43.557 - Info: targetfstool.modify()
11:18:43.557 - Info: File "/targetfstool/targetfstool_base.py", line 66, in modify
11:18:43.557 - Info: self.do_modify()
11:18:43.557 - Info: File "/tool_dir/targetfstool_5_1_6_x_avrfs.py", line 308, in do_modify
11:18:43.557 - Info: self.install_cuda_target()
11:18:43.557 - Info: File "/tool_dir/targetfstool_5_1_6_x_avrfs.py", line 110, in install_cuda_target
11:18:43.557 - Info: cuda_repo_deb_info = get_cuda_repo_deb_info(cuda_repo_deb_glob)
11:18:43.558 - Info: File "/targetfstool/utilities.py", line 28, in get_cuda_repo_deb_info
11:18:43.558 - Info: raise Exception
11:18:43.558 - Info: Exception
11:18:43.558 - Info: ERROR:common:Total log record counts: CRITICAL: 0, ERROR: 2, WARNING: 0
11:18:43.668 - Info: ERROR: could not run targetfstool
8

Hi @SivaRamaKrishnaNV are you able to reproduce this? Thanks

9

Dear @sleff,
Unfortunately, I don’t have ubuntu 18.04 host to verify if there is any change in DRIVE SW 10 installation with recent sdkmanager. I dont expect any such changes to happen.
Does the host installation of packages is showing as successful in sdkmanager? Could you show it in sdkmanager GUI ?