FvbNorFlashStandaloneMm.c(868): ((BOOLEAN)(0==1 after flashing Jetson_Xavier_NX_QSPI_35

Autonomous Machines Jetson & Embedded Systems Jetson Xavier NX
1

First I use jetson-nx-jp461-sd-card-image with success - but I would update to new JP513-xnx-sd-card-image.
Therfore I flashed on a ruuning jp461 xavier NX dev pack the NX_QSPI_35.1 ( jetson-xavier-nx-devkit.spi.img) to /dev/mtd0 and boot with new JP513-xnx-sd-card-imag. First I seems to work
Ubuntu 20.04.6 LTS linux ttyTCU0

linux login: [ 98.352498] using random self ethernet address
[ 98.355263] using random host ethernet address
[ 98.721323] using random self ethernet address
[ 98.724952] using random host ethernet address
[ 110.816297] ramoops reserved-memory:ramoops_carveout: failed to locate DT /reserved-memory resource
[ 110.819985] ramoops: probe of reserved-memory:ramoops_carveout failed with error -22
but on new reboot there will be always the error
ASSERT [FvbNorFlashStandaloneMm] /dvs/git/dirty/git-master_linux/out/nvidia/optee.t194-uefi/StandaloneMmOptee_RELEASE/edk2-nvidia/Silicon/NVIDIA/Drivers/FvbNorFlashDxe/FvbNorFlashStandaloneMm.c(868): ((BOOLEAN)(0==1))

Maybe something in bootloader QSPI /dev/mtd0 (EEPROM ???) went wrong and could not be repaired.
So I think about reflash with JETSTREAM 5.02 - but I work on an ubunbtu 22.04.4 LTS and newest SDK with JETSTREAM 6.x may not work (not supported by the support matrix for XAVIER NX developer board ?)
So my only chance to reflash with docker image sdkmanager-2.0.0.11405-Ubuntu_18.04_docker.tar.gz or sdkmanager-2.0.0.11405-Ubuntu_20.04_docker.tar.gz ?. I did not know what image ist best and if I could repair /dev/mtd0 with this or flash back to former QSPI for jp461 ???

3

The QSPI image is still for 35.1, maybe we forgot to/just did not update it, while you are using an SD card image that is 35.5, and that why the error shows up.
Either re-flash the whole device with SDK Manager to 35.5, or use an SD card image that is 35.1 (JetPack 5.0.2).

4

By the way, it’s JetPack, not JETSTREAM or stuff like that…

5

I use “sdkmanager --cli --action install --login-type devzone --product Jetson --target-os Linux --version 5.0.2 --host --target JETSON_XAVIER_NX_TARGETS --flash --license accept --stay-logged-in true --collect-usage-data enable --exit-on-finish true” with sdkmanager:2.0.0.11405-Ubuntu_18.04. Host install is fine but tardet “JETSON XAVIER NX” fails. Log is :

---2024-03-10 18:46:48 +   Welcome to NVIDIA SDK MANAGER v2.0.0.11405   +

2024-03-10 18:47:56 Sending feedback message...
2024-03-10 18:47:57   ===== INSTALLATION FAILED ===== 
2024-03-10 18:47:57       - CUDA on Host: Up-to-date
2024-03-10 18:47:57       - CUDA Cross Compile Package on Host: Up-to-date
2024-03-10 18:47:57       - NvSci: Up-to-date
2024-03-10 18:47:57       - VPI on Host: Up-to-date
2024-03-10 18:47:57       - Nsight Systems: Up-to-date
2024-03-10 18:47:57       - Nsight Graphics: Up-to-date
2024-03-10 18:47:57       - Nsight Compute: Up-to-date
2024-03-10 18:47:57       - CUPTI: Up-to-date
2024-03-10 18:47:57       - Compute Sanitizer: Up-to-date
2024-03-10 18:47:57       - Drivers for Jetson: Failed
2024-03-10 18:47:57       - File System and OS: DependencyFailure
2024-03-10 18:47:57       - Flash Jetson Xavier NX: DependencyFailure
2024-03-10 18:47:57       - DateTime Target Setup: DependencyFailure
2024-03-10 18:47:57       - CUDA Toolkit for L4T: DependencyFailure
2024-03-10 18:47:57       - cuDNN on Target: DependencyFailure
2024-03-10 18:47:57       - TensorRT on Target: DependencyFailure
2024-03-10 18:47:57       - OpenCV: DependencyFailure
2024-03-10 18:47:57       - VPI on Target: DependencyFailure
2024-03-10 18:47:57       - NVIDIA Container Runtime with Docker integration (Beta): DependencyFailure
2024-03-10 18:47:57       - Multimedia API: DependencyFailure
2024-03-10 18:47:57       - BSP Patch: DependencyFailure
2024-03-10 18:47:57 
2024-03-10 18:47:57   ===== Installation failed - Total 21 components =====
2024-03-10 18:47:57   ===== 0 succeeded, 1 failed, 9 up-to-date, 11 skipped =====
2024-03-10 18:47:57 

2024-03-10 18:49:59 ++++++++++++++++++++++++++++++++++++++++++++++++++
2024-03-10 18:49:59 +   Welcome to NVIDIA SDK MANAGER v2.0.0.11405   +

2024-03-10 18:51:13 Sending feedback message...
2024-03-10 18:51:18   ===== INSTALLATION FAILED ===== 
2024-03-10 18:51:18       - CUDA on Host: Up-to-date
2024-03-10 18:51:18       - CUDA Cross Compile Package on Host: Up-to-date
2024-03-10 18:51:18       - NvSci: Up-to-date
2024-03-10 18:51:18       - VPI on Host: Up-to-date
2024-03-10 18:51:18       - Nsight Systems: Up-to-date
2024-03-10 18:51:18       - Nsight Graphics: Up-to-date
2024-03-10 18:51:18       - Nsight Compute: Up-to-date
2024-03-10 18:51:18       - CUPTI: Up-to-date
2024-03-10 18:51:18       - Compute Sanitizer: Up-to-date
2024-03-10 18:51:18       - Drivers for Jetson: Failed
2024-03-10 18:51:18       - File System and OS: DependencyFailure
2024-03-10 18:51:18       - Flash Jetson Xavier NX: DependencyFailure
2024-03-10 18:51:18       - DateTime Target Setup: DependencyFailure
2024-03-10 18:51:18       - CUDA Toolkit for L4T: DependencyFailure
2024-03-10 18:51:18       - cuDNN on Target: DependencyFailure
2024-03-10 18:51:18       - TensorRT on Target: DependencyFailure
2024-03-10 18:51:18       - OpenCV: DependencyFailure
2024-03-10 18:51:18       - VPI on Target: DependencyFailure
2024-03-10 18:51:18       - NVIDIA Container Runtime with Docker integration (Beta): DependencyFailure
2024-03-10 18:51:18       - Multimedia API: DependencyFailure
2024-03-10 18:51:18       - BSP Patch: DependencyFailure
2024-03-10 18:51:18 
2024-03-10 18:51:18   ===== Installation failed - Total 21 components =====
2024-03-10 18:51:18   ===== 0 succeeded, 1 failed, 9 up-to-date, 11 skipped =====
2024-03-10 18:51:18 
2024-03-10 18:51:18 
--
and console shows always 
I/TC: Reserved shared memory is disabled                                      
I/TC: Dynamic shared memory is enabled                                          
I/TC: Normal World virtualization support is disabled                           
I/TC: Asynchronous notifications are disabled                                   
                                                                                
ASSERT [FvbNorFlashStandaloneMm] /dvs/git/dirty/git-master_linux/out/nvidia/opt)
I wonder about [0003.401] E> Failed to get board id info!                                      
[0003.405] E> Failed: Unknown device 6                                          
[0003.413] I> Found 41 partitions in QSPI_FLASH (instance 0)

Neither SD Image nor flash with SDK Manger works
Only when I remove SD Card the target goes to recovery mode - I have also a NVMe disk inserted.

6

now I have found: JetPack 5.1.3 includes Jetson Linux 35.5.0 with following highlights: at https://developer.nvidia.com/embedded/jetpack-archive. There are the relation from JetPack to Linux (QSPI boot loader). Next I try this with cli to flash whole device with SDK manager. But I did not know how to flash to PCIe NVMe or is it only possible to flash to SD-Card? I hope to recover from ASSERT [FvbNorFlashStandaloneMm]

7

I started the sdk process without failure - but loss of communication leads to skip flashing target. But when I restart container no progress on log could be seen - also no error. On docker console I see a prompt and 2 directories “Download/nvidia” with sdkm_downloads and nvidia with “nvidia_sdk”. But I did not see any progress and so I do not know when to restart the device

8

Jetpack 5.1.3 does not help. I could not flash target. Target components → “Flash Jetson” always skipped and sd-card image JP502-xnx-sd-card-image-b231.zip always leads to ASSERT failure. I plugged in FC REC and GND jumper, than connect power cable and at time to manual flash the board would not be recognized when jumper was removed after instruction from sdkmanger to do this. lsusb shows correct id ID 0955:7e19 NVIDIA Corp. APX when jumper is connected. Also I use ping 192.168.55.1 from docker container with response . 0% packets loss.
I did not now what to do next. With these lot of dependencies it is very hard to overcome the booting problem when one step in the chain fails or is wrong. The QR code for authentication often leads to cut off login by loss of website and login failure when smartphone switch off display and I could not hit final step for login because website was lost.

9

Why do you need Docker when you have a native Ubuntu PC?
Please download the BSP manually and follow these steps:

$ tar xf ${L4T_RELEASE_PACKAGE}
$ sudo tar xpf ${SAMPLE_FS_PACKAGE} -C Linux_for_Tegra/rootfs/
$ cd Linux_for_Tegra/
$ sudo ./apply_binaries.sh
$ sudo ./tools/l4t_flash_prerequisites.sh

https://docs.nvidia.com/jetson/archives/r35.5.0/DeveloperGuide/IN/QuickStart.html
https://developer.nvidia.com/downloads/embedded/l4t/r35_release_v5.0/release/jetson_linux_r35.5.0_aarch64.tbz2
https://developer.nvidia.com/downloads/embedded/l4t/r35_release_v5.0/release/tegra_linux_sample-root-filesystem_r35.5.0_aarch64.tbz2

10

thank you very much - my board is running but ssh connection fails and no Jetson Runtime componets (Cuda Runtime, Cuda X-AI Runtime could be installed). ssh.service: Start request repeated to quick. no update or fix–broken would work.

11

Simply do

sudo apt update
sudo apt install nvidia-jetpack

on the device.

12

ok that seems easy - but I only have root partition APP of 14GB and a UDA Partition of 462GB. Could I resize the UDA partition and mount some space whete nvidia-jetpack would be installed ( ~/nividia or /opt/nvidia[-sdk] ? ).
I know nothing about UDA partition with unknown format and fear to damage the partition.
I think on flashing ssd the partitioning would be gone wrong.

13

Do this:

15

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.