Hi Nvidia Team,
We are facing an issue where a blank white screen appears when we shut down the board using the sudo shutdown or /sbin/shutdown -h now command.
We tested with different HDMI cables and monitors, but we are still seeing this issue on some boards.
I have attached the logs below for your reference.
[ 271.423146] tegra-ivc-bus bc00000.rtcpu:ivc-bus:echo0: ivc channel driver missing
[ 271.423150] tegra-ivc-bus bc00000.rtcpu:ivc-bus:dbg1: ivc channel driver missing
[ 271.423152] tegra-ivc-bus bc00000.rtcpu:ivc-bus:dbg2: ivc channel driver missing
[ 271.423156] tegra-ivc-bus bc00000.rtcpu:ivc-bus:diag5: ivc channel driver missing
[ 271.792028] nvgpu: 17000000.gpu pmu_wait_message_cond:664 [ERR] PMU wait timeout expired.
[ 271.792032] nvgpu: 17000000.gpu pmu_wait_message_cond:664 [ERR] PMU wait timeout expired.
[ 271.792637] nvgpu: 17000000.gpu ga10b_intr_log_pending_intrs:306 [ERR] Pending TOP[0]: 0x00000004, LEAF[4]: 0x11000000
[ 271.793643] arm-smmu 12000000.iommu: disabling translation
[ 271.793741] arm-smmu 10000000.iommu: disabling translation
[ 271.793775] arm-smmu 8000000.iommu: disabling translation
[ 271.793776] tegra-mc 2c00000.memory-controller: nvdisplayr: read 0x0000007fff5dab80: EMEM address decode error (EMEM decode error)
[ 271.793788] tegra-mc 2c00000.memory-controller: nvdisplayr: read 0x0000007fff5dba00: EMEM address decode error (EMEM decode error)
[ 271.793795] tegra-mc 2c00000.memory-controller: nvdisplayr: read 0x0000007fff5dd600: EMEM address decode error (EMEM decode error)
[ 271.793801] tegra-mc 2c00000.memory-controller: nvdisplayr: read 0x0000007fff5de200: EMEM address decode error (EMEM decode error)
[ 271.793807] tegra-mc 2c00000.memory-controller: nvdisplayr: read 0x0000007fff5df200: EMEM address decode error (EMEM decode error)
[ 271.793813] tegra-mc 2c00000.memory-controller: nvdisplayr1: read 0x0000007fff5e0000: EMEM address decode error (EMEM decode error)
[ 271.793820] tegra-mc 2c00000.memory-controller: nvdisplayr1: read 0x0000007fff5dae00: EMEM address decode error (EMEM decode error)
[ 271.793826] tegra-mc 2c00000.memory-controller: nvdisplayr1: read ����Shutdown state requested 0
Shutting down system …
��0x0000007fff5e1000: EMEM address decode error (EMEM decode error)
[ 271.793832] tegra-mc 2c00000.memory-controller: nvdisplayr1: read 0x0000007fff5e1200: EMEM address decode error (EMEM decode error)
[ 271.793839] tegra-mc 2c00000.memory-controller: nvdisplayr: read 0x0000007fff5e2200: EMEM address decode error (EMEM decode error)
[ 271.797580] reboot: Power down
We are getting the same logs for the working boards as well. When I execute this command, the board freezes with a white screen. Sometimes the board powers off after a few minutes, and sometimes it boots again.
I suspect that the watchdog might be getting triggered. Kindly let me know if you need any further clarification.
Hi akash.kumar2,
Are you using the devkit or custom board for Orin NX?
Could you also verify with the latest Jetpack 6.2.2(r36.5)?
Please share the full log as file instead of part of them.
Hi @KevinFFF, We are using a custom board based on the NVIDIA Orin NX platform with JetPack 6.1 (36.4.0).
This issue is not reproducible across all boards. Out of 10 boards tested, the issue was observed on only 1 board.
For reference, we have attached the relevant log file. We are currently investigating the root cause and would appreciate your input.
Please let us know if you need any additional information or clarification.
display_issue.log (91.9 KB)
Do you mean the issue is specific to one board?
Is the issue reproducible 100% on this failed board?
Could you try running to following command to check if there’s the similar blank white screen issue?
$ sudo systemctl isolate multi-user.target
$ sudo poweroff
For the display related issue, you can also check /var/log/Xorg.0.log.
Hi @KevinFFF
Yes, the issue appears to be specific to one board.
Out of the 10 boards tested, the issue is observed only on one board. On this particular board, the issue is 100% reproducible every time.
I also tried the commands that you shared above, but the issue is still reproducing on that board.
For reference, I have attached the Xorg.0.log file along with the relevant logs. Please let me know if you need any additional information or clarification.
Xorg.0.log (7.1 KB)
What is the input voltage on your custom carrier board?
Please refer to Errata-2 in Jetson Orin NX Series Modules - Hardware Errata. It sounds like the failing module may be related. The Errata has a link for a serial number lookup.
Hi @ChrisB_NV , The input voltage used for the custom carrier board is 12V ± 0.5V. We checked the serial number against the errata link, and it appears to be not impacted. I have attached the screenshot for your reference.
Please let me know if any further clarification is required.
Thank you for checking your serial number. The errata would only apply if it’s impacted and was ever supplied with >12V.
Does it ever remain on and keep displaying the white screen? Or does it only power off after minutes or reboot itself?
Does it ever remain on and keep displaying the white screen? Or does it only power off after minutes or reboot itself?
Yes @ChrisB_NV , It keeps showing a white screen, sometimes it turns off after a few minutes, and sometimes it restarts itself.
From the log, we don’t see the watchdog triggering the reset.
Could you also verify with the latest Jetpack 6.2.2(r36.5) to check if there’s the similar issue?
Hi @KevinFFF , Thanks for the response. Currently, we have a time constraint for porting to JP 6.2.2. This issue is observed only on one board, which is our main concern, while other boards are working fine.
Apart from porting, is there any other way we can verify or debug this issue? As per the errata, we are already using the latest pinmux configuration, and it is also mentioned there that this serial number is not impacted.
Please let us know your suggestions.
Since the issue is only occurring on a specific board, it may not be related to the software configuration.
Could you help clarify whether the problem is specific to that individual module or the board itself?
@KevinFFF , This issue appears to be specific to an individual SOM. We tested by swapping the SOM with other boards, and today we observed the same issue on another SOM as well.
We checked the serial numbers against the errata list, and they are marked as not impacted. Hence, we are unsure why this issue is occurring.
Please let us know if any additional information is required from our side.
How many failed modules in your case? and how about the failed rate?
@KevinFFF , Currently, we have observed this issue on 2 boards, with a failure rate of 10/10.
Based on your previous result, the issue is specific to the module rather than the board.
How many modules do you have with 2 failed modules?
@KevinFFF , We have more than 150 modules in total. So far, we have tested 15 modules, and issues were identified in 2 of them.
Thanks for your information.
Do you have the Orin Nano devkit?
I would like to know if there’s the similar issue if you put the failed module on the devkit and flash the latest JP6.2.2(r36.5).
@KevinFFF , We tested one failed module on the Orin dev kit using versions R36.5 and R36.4.0, and the issue could not be reproduced. we observed a white screen lasting around 2 seconds.
hello
When I execute sudo shutdown or sudo poweroff, there will also be a white screen on HDMI display, and the error log is the same as the one you submitted. My SDK version is R36.5
