"pcie failed to prase device tree" using default UPHY1 and pcie config

Please provide the following info (tick the boxes after creating this topic):
Software Version
DRIVE OS 6.0.8.1
[*] DRIVE OS 6.0.6
DRIVE OS 6.0.5
DRIVE OS 6.0.4 (rev. 1)
DRIVE OS 6.0.4 SDK
other

Target Operating System
[*] Linux
QNX
other

Hardware Platform
DRIVE AGX Orin Developer Kit (940-63710-0010-300)
[*] DRIVE AGX Orin Developer Kit (940-63710-0010-200)
DRIVE AGX Orin Developer Kit (940-63710-0010-100)
DRIVE AGX Orin Developer Kit (940-63710-0010-D00)
DRIVE AGX Orin Developer Kit (940-63710-0010-C00)
DRIVE AGX Orin Developer Kit (not sure its number)
other

SDK Manager Version
1.9.3.10904
[*] other

Host Machine Version
[*] native Ubuntu Linux 20.04 Host installed with SDK Manager
native Ubuntu Linux 20.04 Host installed with DRIVE OS Docker Containers
native Ubuntu Linux 18.04 Host installed with DRIVE OS Docker Containers
other

I find the default UPHY1 and pcie config as below:

1、using the default config, have error log :


tegra194-pcie 141a0000.pcie: Failed to parse device tree: -517

tegra194-pcie 141c0000.pcie_ep: Failed to parse device tree: -517

2、In document “Jetson_AGX_Orin_Series_Design_Guide_DG-10653-001_v1.6.pdf” and " Jetson AGX Orin Platform Adaptation and Bring-Up — Jetson Linux Developer Guide documentation (nvidia.com) " there is no 4 lane config option for UPHY1:

image

3、So, is the default 4 lane config for UPHY1 to pcie correct ? In this develop kit (DRIVE AGX Orin Developer Kit (940-63710-0010-200)) and SDK (DRIVE OS 6.0.6 & Linux) can config UHPY1 to 2 pcie 4 lane,if yes or not, how to modify?

Dear @huangxx8,
May I know when you notice this issue like flashing or compilation?. Did you make any changes or noticed issue with default files in DRIVE OS?

1、We want to use “nvscistream_event_sample” to test nvscic2c_pcie function, I follow " Chip to Chip Communication | NVIDIA Docs " when excute “sudo modprobe nvscic2c-pcie-epc” or “sudo modprobe nvscic2c-pcie-epf” there is no extra pcie log out

2、form the error log “tegra194-pcie 141a0000.pcie: Failed to parse device tree: -517” I think the driver is not ready, so i want to know is the default config correct?

3、There is’t any changes and I can’t find the default UPHY1 config in document “Jetson_AGX_Orin_Series_Design_Guide_DG-10653-001_v1.6.pdf” and " Jetson AGX Orin Platform Adaptation and Bring-Up — Jetson Linux Developer Guide documentation (nvidia.com) "

Dear @huangxx8,
Just to confirm is this error noticed in bootlog or running modprobe or nvscistream event sample?

In bootlog, running modprobe or nvscistream event sample have no extra pcie log out

I don’t see any such log using DRIVE OS 6.0.8.1

nvidia@tegra-ubuntu:~$ sudo modprobe nvscic2c-pcie-epf
nvidia@tegra-ubuntu:~$ sudo modprobe nvscic2c-pcie-epc
nvidia@tegra-ubuntu:~$

The above error is not notice in 6.0.8.1 bootlog.

Is it possible to test on 6.0.8.1?

I want to know is there the same error in SDK 6.0.6 on your side, Do you have test nvscic2c-pcie function in SDK 6.0.6? If same error and nvscic2c-pcie can not running OK, I will try to test on 6.0.8.1

Dear @huangxx8,
Currently, I have 6.0.8.1 flashed on target and unfortunately I don’t have c2c use case test setup for quick checking.

Dear @huangxx8 ,
May I know the used command for nvscistream event sample?

I use the sample in drive-linux/samples/ncsci/nvscistream/event, in README.txt, the command:

Hi SivaRamaKrishnaNV,
Any update for this?

Dear @huangxx8,
As I don’t have 2 P3710, I had used only one p3710 as PCIE master and slave. I connected the miniSAS cable port-A to miniSAS port-B in 1 P3710s. Followed the steps from Chip to Chip Communication | NVIDIA Docs

I could run nvscistream event sample

#Run on Devkit 1 Terminal
nvidia@tegra-ubuntu:~$ sudo modprobe nvscic2c-pcie-epc
[sudo] password for nvidia:

#Run below steps on  Devkit2 terminal if you have connected two devkits

nvidia@tegra-ubuntu:~$ sudo modprobe nvscic2c-pcie-epf  
nvidia@tegra-ubuntu:~$ sudo -s
root@tegra-ubuntu:/home/nvidia# cd /sys/kernel/config/pci_ep/
root@tegra-ubuntu:/sys/kernel/config/pci_ep# mkdir functions/nvscic2c_epf_22CC/func
root@tegra-ubuntu:/sys/kernel/config/pci_ep# echo 0x10DE > functions/nvscic2c_epf_22CC/func/vendorid
root@tegra-ubuntu:/sys/kernel/config/pci_ep# echo 0x22CC > functions/nvscic2c_epf_22CC/func/deviceid
root@tegra-ubuntu:/sys/kernel/config/pci_ep# ln -s functions/nvscic2c_epf_22CC/func controllers/141c0000.pcie_ep
root@tegra-ubuntu:/sys/kernel/config/pci_ep# echo 0 > controllers/141c0000.pcie_ep/start
root@tegra-ubuntu:/sys/kernel/config/pci_ep# echo 1 > controllers/141c0000.pcie_ep/start
root@tegra-ubuntu:/sys/kernel/config/pci_ep# lspci
0005:00:00.0 PCI bridge: NVIDIA Corporation Device 229a (rev a1)
0005:01:00.0 Serial controller: NVIDIA Corporation Device 22cc
0007:00:00.0 PCI bridge: NVIDIA Corporation Device 229a (rev a1)
0007:01:00.0 Ethernet controller: Microchip Technology / SMSC Device 7431 (rev 11)
root@tegra-ubuntu:/sys/kernel/config/pci_ep# ls /dev/nvsci*
/dev/nvscic2c_pcie_s0_c5_1   /dev/nvscic2c_pcie_s0_c5_5   /dev/nvscic2c_pcie_s0_c6_11  /dev/nvscic2c_pcie_s0_c6_7
/dev/nvscic2c_pcie_s0_c5_10  /dev/nvscic2c_pcie_s0_c5_6   /dev/nvscic2c_pcie_s0_c6_12  /dev/nvscic2c_pcie_s0_c6_8
/dev/nvscic2c_pcie_s0_c5_11  /dev/nvscic2c_pcie_s0_c5_7   /dev/nvscic2c_pcie_s0_c6_2   /dev/nvscic2c_pcie_s0_c6_9
/dev/nvscic2c_pcie_s0_c5_12  /dev/nvscic2c_pcie_s0_c5_8   /dev/nvscic2c_pcie_s0_c6_3   /dev/nvsciipc
/dev/nvscic2c_pcie_s0_c5_2   /dev/nvscic2c_pcie_s0_c5_9   /dev/nvscic2c_pcie_s0_c6_4
/dev/nvscic2c_pcie_s0_c5_3   /dev/nvscic2c_pcie_s0_c6_1   /dev/nvscic2c_pcie_s0_c6_5
/dev/nvscic2c_pcie_s0_c5_4   /dev/nvscic2c_pcie_s0_c6_10  /dev/nvscic2c_pcie_s0_c6_6
root@tegra-ubuntu:/sys/kernel/config/pci_ep# exit
exit

#Devkit 1 terminal
nvidia@tegra-ubuntu:~$ ./nvscistream_event_sample -P 0 nvscic2c_pcie_s0_c5_1 -Q 0 f
Producer setup completed
Producer finished sending 32 payloads


#Devkit 2 terminal
nvidia@tegra-ubuntu:~$ ./nvscistream_event_sample -C 0 nvscic2c_pcie_s0_c6_1 -F 0 3
Consumer setup completed
Consumer disconnected after receiving 32 payloads

Note: I had used only 1 P3710 in my test. So I ran all commands on same p3710.

图像

may be we use the wrong Mini-SAS cable,we will use the supported cables late and update.

1 Like

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