Jetson Orin Nano Display Problem

Hello,

On our own carrier card, we get the following display error on the Jetson Orin Nano 4GB module.

The screen works in the “jetson-orin-nano-devkit” configuration file, there is no problem.

But it gives the following error in our own configuration file.

[   12.274861] nv_platform 13800000.display: Adding to iommu group 41
[   12.278740] platform 13800000.display:nvdisplay-niso: Adding to iommu group 42
[   12.282512] platform 13800000.display:nvdisplay-niso: Removing from iommu group 42
[   12.294736] nv_platform: probe of 13800000.display failed with error -1
[   21.397362] platform 13800000.display:nvdisplay-niso: Adding to iommu group 43
[   21.397608] nv_platform 13800000.display: Unbalanced pm_runtime_enable!
[   21.407426] platform 13800000.display:nvdisplay-niso: Removing from iommu group 43
[   21.407559] nv_platform: probe of 13800000.display failed with error -1
[   21.615218] platform 13800000.display:nvdisplay-niso: Adding to iommu group 43
[   21.615497] nv_platform 13800000.display: Unbalanced pm_runtime_enable!
[   21.623690] platform 13800000.display:nvdisplay-niso: Removing from iommu group 43
[   21.623803] nv_platform: probe of 13800000.display failed with error -1

The content of our own .conf file is as follows.

jetson Orin Nano 4GB-DRAM (P3767-0004) PROD MODULE

source “${LDK_DIR}/p3767.conf.common”;

DEFAULT_EMC_FUSE=“0”;
PINMUX_CONFIG=“custom-pinmux.dtsi”;
PMC_CONFIG=“custom-padvoltage-default.dtsi”;
BPFDTB_FILE=“tegra234-bpmp-3767-0000-a02-as-3767-0004-3509-a02.dtb”;
DTB_FILE=“tegra234-p3767-0004-p3509-a02.dtb”;
TBCDTB_FILE=“${DTB_FILE}”;
EMMC_CFG=“flash_t234_qspi_sd.xml”;

We used Jetson_Orin_NX_Series_+HDMI A03 as the Pinmux table and did not change anything in the HDMI part of the pins.

There is no eeprom on the carrier card, we bypassed it.

Also, the naming part seems very complicated. We use A03 on Pinmux, there is no A03 in the dtb section, we use A02 instead.

I know there is a 4 GB nano with 0004, I am trying to detect the correct file from there.

Is there a guide for this?

How can I quickly fix the display problem in its my configuration?

Thanks.

First, if you are using HDMI, then “jetson-orin-nano-devkit” board config does not really support that. jetson-orin-nano-devkit is for Orin nano devkit carrier board which does not have HDMI port on it at all.

The reason your HDMI seems working with that configuration is just because DP++ is supported on display driver. But if you try 4k HDMI monitor on it, it will fail.
Run xrandr on jetson and it shall still give you “DP-0” instead of “HDMI-0” for this case.

Second, I would suggest you can directly flash the p3509+ p3767 board config, instead of directly modifying the whole board config by yourself.

Your own conf file looks not make sense…

For example, what is “tegra234-bpmp-3767-0000-a02-as-3767-0004-3509-a02.dtb”? There is totally no such thing exist in BSP at all.

My card only has HDMI connection. Display port is not available.

It is connected as in the diagram below.

hdmi

In order to make changes to Pinmux, I have to make changes to the configuration.

It also worked when I set it as below in the config file.

BPFDTB_FILE=“tegra234-bpmp-3767-0004-3509-a02.dtb”;
DTB_FILE=“tegra234-p3767-0004-p3768-0000-a0.dtb”;

But I really don’t know the difference between these, what do the a0 and a2 at the end mean, why do we use a03 in pinmux?

Also, you are talking about whether HDMI connection is not supported, it only works because it is DP++.
In this case, you say that it will not work when you connect a 4K monitor to the HDMI connection.

Is this problem hardware or configuration issue?

Can I fix the 4K monitor issue by fixing the configuration file?

Thanks.

In this case, you say that it will not work when you connect a 4K monitor to the HDMI connection.
Is this problem hardware or configuration issue?

Just software configuration issue.

The point is just if you use “jetson-orin-nano-devkit.conf” to flash, you will only have DP mode.

But if you use p3509+ p3767 board config to flash, it will have HDMI mode.

You are not the first one asking such question. So I know what you are doing here.

But I really don’t know the difference between these, what do the a0 and a2 at the end mean, why do we use a03 in pinmux?

The naming really does not matter here. It is just a sku name. If it can work, then it is a good configuration.

If you use this DTB, you will only run in DP/DP++ mode.

DTB_FILE=“tegra234-p3767-0004-p3768-0000-a0.dtb”;

p3768 is Orin nano devkit carrier board which only supports DP.
P3509 is Xavier NX devkit carrier board which supports HDMI.

p3767 is Orin Nano/NX module.

You cannot and don’t need compare the board sku between p3768 (a0) and p3509(a02) because they are totally not related.

Also, please be aware that you are the one who gave out “tegra234-bpmp-3767-0000-a02-as-3767-0004-3509-a02.dtb” as below screenshot. I am not sure if this is just a typo or you really created something wrong here.

1 Like

This “tegra234-bpmp-3767-0000-a02-as-3767-0004-3509-a02.dtb” file actually exists, I don’t remember whether it was created while testing.

I normally use this file. “tegra234-bpmp-3767-0004-3509-a02.dtb”.

But I understood what you said.

I’m trying it this way and I will let you know the results.

BPFDTB_FILE=“tegra234-bpmp-3767-0004-3509-a02.dtb”;
DTB_FILE=“tegra234-p3767-0004-p3509-a02.dtb”;

Thanks.

In the last case, the device gives an error and turns on again.

Used configuration:

BPFDTB_FILE=“tegra234-bpmp-3767-0000-a02-as-3767-0004-3509-a02.dtb”;
DTB_FILE=“tegra234-p3767-0004-p3509-a02.dtb”;

I am uploading the records in attachment.

all.txt (86.2 KB)

[ 16.858079] CPU:0, Error: dce-fabric@0xde00000, irq=28
[ 16.863378] **************************************
[ 16.868307] CPU:0, Error:dce-fabric, Errmon:2
[ 16.872793] Error Code : SLAVE_ERR
[ 16.876648] Overflow : Multiple SLAVE_ERR
[ 16.881134]
[ 16.882667] Error Code : SLAVE_ERR
[ 16.886523] MASTER_ID : DCE
[ 16.889742] Address : 0x2212000
[ 16.893331] Cache : 0x1 – Bufferable
[ 16.897629] Protection : 0x3 – Privileged, Non-Secure, Data Access
[ 16.904435] Access_Type : Read
[ 16.907927] Access_ID : 0x0
[ 16.907928] Fabric : dce-fabric
[ 16.914742] Slave_Id : 0x3b
[ 16.917967] Burst_length : 0x0
[ 16.921461] Burst_type : 0x1
[ 16.924779] Beat_size : 0x2
[ 16.928004] VQC : 0x0
[ 16.930785] GRPSEC : 0x3f
[ 16.933836] FALCONSEC : 0x0

There was no such problem in the configuration below.

BPFDTB_FILE=“tegra234-bpmp-3767-0004-3509-a02.dtb”;
DTB_FILE=“tegra234-p3767-0004-p3768-0000-a0.dtb”;

Am I doing something wrong?

Thanks.

No, you didn’t do anything wrong. We are in the right direction.

This is a known issue. Please read release note and apply patch to bypass the error.

Please forget about what you did in p3768 dtb. You didn’t hit error just because it was DP mode. It didn’t mean correctness.

1 Like

Yes,

As you said, I made the patch in the document and it is now working with the configuration below.

BPFDTB_FILE=“tegra234-bpmp-3767-0004-3509-a02.dtb”;
DTB_FILE=“tegra234-p3767-0004-p3509-a02.dtb”;

Thanks.

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