Orin Nano 8GB no display

Hello NVIDIA Team,

I am encountering an issue with an Orin Nano 8GB running JetPack 5.1.1 on a custom board. I have connected a Full HD Display (ACER B233HU) to the Display Port of the carrier board.

Upon powering on the device, the UEFI Screen with the NVIDIA logo is displayed. However, after that, the screen goes black and I am unable to see any further output.

I can establish a connection to the Orin Nano over the USB interface, and once the initial configuration is complete, I can access it via SSH. This indicates that the device boots up properly.

I have tried using multiple different screens, all with the same outcome - the device does not display anything after booting.

For reference, I have attached the dmesg output and Xorg log of the device:

dmesglog.txt (69.4 KB)
Xorg.0.log (11.6 KB)

Interestingly, when using the same custom carrier board, configuration files, etc., but with an Orin NX instead of an Orin Nano 8GB, the display port works perfectly after booting up.

I would greatly appreciate any assistance or insights you can provide regarding this issue.

Thank you all in advance for your support

Best regards,
Olivier

Hi,

sorry for the late reply.
How did you connect the device to monitors? Via HDMI or DP?
Can you please attach the same dmesg and Xorg logs when you can get it work on Orin NX?

Hi Dave,

The display is connected to the DisplayPort connector on the carrier board using an HDMI-DisplayPort adapter.
I have attached the dmesg log and Xorg log of the NX for your reference:
dmesglog_nx.txt (72.2 KB)
Xorg.0.log (18.7 KB)
Upon reviewing the dmesg log, I noticed that ‘nvidia-modeset’ is loaded on line 1001 for the Orin NX, while it is not happening on the Orin Nano (which should occur around line 976).
Could this cause our problem?

Thanks and best regards,
Olivier

Hi,

Just want make sure that are you using exactly the same BSP on both your Orin NX and Orin Nano?
Like, is nvidia-modeset.ko present in your Orin Nano?
What if you manually load this module?

Hi Dave,
Yes, we are using exactly the same BSP on both. Flashed from the same host basically back to back.
nvidia-modeset.ko is present on the Orin Nano.
However, with modprobe nvidia-modeset it gives the Error could not insert 'nvidia-modeset': No such device
and with insmod /lib/modules/5.10.104-tegra/extra/opensrc-disp/nvidia-modeset.ko it gives the Error could not insert module /lib/modules/5.10.104-tegra/extra/opensrc-disp/nvidia-modeset.ko: Unknown symbol in module
Does the module need to be loaded in a special way?
Thanks and best regards,
Olivier

Hi, then can you check what symbols you are missing in the system?

Hello, I have the same problem. I am using SSD for storage. I followed the below link steps. I get a message about successful installation from SDK Manager, but the screen does not appear. I am able to SSH connection and I think the device is working. How can I solve it?

Hi Dave,
My missing simbols are ‘nvidia_register_module’, ‘nvidia_get_rm_ops’ and ‘nvidia_unregister_module’. They are not present in /proc/kallsyms as well.

Hi @furkan.sakaci1,

we are not sure if your issue is the same with described in this topic, so please file a new topic for your issue.

Hi @olivier.lutz,

are nvidia.ko and nvidia-drm.ko also present in your system?
Can you show the dmesg log when the module fails to be loaded?

Hi Dave,
Yes, they are both present.
sudo insmod /lib/modules/5.10.104-tegra/extra/opensrc-disp/nvidia.ko results in the following dmesg entry:
[ 1839.095671] platform 13800000.display:nvdisplay-niso: Adding to iommu group 55
[ 1839.095885] nv_platform 13800000.display: Unbalanced pm_runtime_enable!
[ 1839.103430] MRQ_EMC_DVFS_LATENCY table has no entries
[ 1839.103435] NVRM: failed to get bpmp data
[ 1839.103730] platform 13800000.display:nvdisplay-niso: Removing from iommu group 55
[ 1839.103830] nv_platform: probe of 13800000.display failed with error -1
[ 1839.111162] NVRM: The NVIDIA probe routine was not called for 1 device(s).
[ 1839.111164] NVRM: This can occur when a driver such as: NVRM: nouveau, rivafb, nvidiafb or rivatv NVRM: was loaded and obtained ownership of the NVIDIA device(s).
[ 1839.111165] NVRM: Try unloading the conflicting kernel module (and/or NVRM: reconfigure your kernel without the conflicting NVRM: driver(s)), then try loading the NVIDIA kernel module NVRM: again.
[ 1839.111167] NVRM: No NVIDIA devices probed.
(none of the mentioned drivers are loaded in lsmod)

and sudo insmod /lib/modules/5.10.104-tegra/extra/opensrc-disp/nvidia-drm.ko results in nvidia_drm: Unknown symbol nvKmsKapiGetFunctionsTable (err -2).
nvKmsKapiGetFunctionsTable is not present in /proc/kallsyms as well

Hi,

do you have an DevKit board for validation?
See if you can get display to work with the same BSP package on a DevKit.

Hello Dave,
I am pleased to inform you that we have successfully identified the cause of the issue we were experiencing.
The problem arose because we were using the same .config file to flash our Orin Nano 8GB, Nano 4GB, and NX 16GB devices. E.g. the BPFDTB_FILE (tegra234-bpmp...dtb) was hardcoded for the NX, which resulted in the Nano 8GB having no display output and the 4GB Nano being unflashable, as mentioned here.
To resolve these issues, we have made the necessary changes to our .config file. We started with the jetson-orin-nano-devkit.conf and replaced the .dtb and pinmux file with our custom board-specific ones. This solution has successfully resolved both the display issues of the 8GB Nano and the flashing issues of the 4GB Nano.
We appreciate your time and assistance throughout this process.
Thank you once again for your support.
Best regards,
Olivier

1 Like

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