Problems and Solutions: Micro Constantly Waking by PCI to PCI Bridge (PCI VEN_1022&DEV_1483&SUBSYS_87C01043&REV_003&11583659&0&19)

This problem had been occurring with my PC after the update with the ASUS ROG Crosshair VIII Hero Wi-Fi AMD X570 Motherboard and AMD Ryzen processor 9 3900X. This seems to have started after some updates to the Windows, including optional drivers while I was trying to resolve a blue screen crash issue.

Issue: The computer started waking up shortly after sleeping (suspension). He goes on hold and soon (after about 20 ~ 30 minutes) is awake.

To investigate the problem I opened Windows Powershell as an administrator and used the following command:

powercfg - lastwake

I got this answer:

Activation History Count - 1
Activation History [0]
Activation Source Count - 1
Activation Source [0]
Type: Dispositivo
Caminho da Instância: PCI  VEN_1022&DEV_1483&SUBSYS_87C01043&REV_003&11583659&0&19
Friendly Name:
Description: Ponte PCI para PCI
Fabricante: (Standard system devices)
Using powercfg -lastwake to find out who woke the PC.

Using powercfg -lastwake to find out who woke the PC.

Attention to string instance path. Using HWiNFO64 I found this same string tapping with the following device AMD Starship/Matisse/Vermeer – PCIe GPP Bridge[7:0].

Investigating the device tree in HWiNFO.

Investigating the device tree in HWiNFO.

I noticed that all the items in the tree of this device are related to the video card.

Then, I went back to Powershell and used the following command:

powercfg -devicequery wake_armed

I then got that list:

HID-compatible mouse (001)
Dispositivo de teclado HID
Intel(R) I211 Gigabit Network Connection
Dispositivo de teclado HID (002)
Realtek PCIe 2.5GbE Family Controller
NVIDIA USB Type-C Port Policy Controller
Mouse compatível com HID (008)
HID keyboard device (008)
Intel(R) Wi-Fi 6 AX200 160MHz
Investigating which devices can wake the PC with powercfg -wakearmed.

Investigating which devices can wake the PC with powercfg -wakearmed.

Here the device draws attention NVIDIA USB Type-C Port Policy Controller which is the only one related to the video card. It is also listed there on HWiNFO64 as NVIDIA TU106 – USB Type-C Port Policy Controller.

I found this device in the device manager, I opened its properties and noticed that in the tab Energy management, he was allowed to wake the computer. I deselected the checkbox Allow this device to wake the computer.

Disabling permission to wake the PC in Device Manager.

Disabling permission to wake the PC in Device Manager.

This temporarily solves the problem. But, when restarting the computer this option is always checked again, and then the problem returns. To get around it I created a file batch with the following command:

powercfg -devicedisablewake "NVIDIA USB Type-C Port Policy Controller"

Then, I put this batch to be executed at each system startup, using Task Scheduler from the Windows, and remembering to check to run with administrator privileges and whether the user is logged in or not.

But a new problem arose, now another PCI to PCI bridge started to wake the PC:

PS C:\Windows\system32> powercfg -lastwake
Contagem de Histórico de Ativação - 1
Activation History [0]
Activation Source Count - 1
Activation Source [0]
Type: Dispositivo
Caminho da Instância: PCI  VEN_1022&DEV_1483&SUBSYS_87C01043&REV_003&11583659&0&0A
Nome Amigável:
Description: Ponte PCI para PCI
Fabricante: (Standard system devices)

The device that corresponds to that string instance path has many more elements in its branches than the other.

I then focused on the list of devices allowed to wake the PC:

PS C:\Windows\system32> powercfg /devicequery wake_armed
Mouse compatível com HID (001)
Dispositivo de teclado HID
GRUPO : Ethernet Team - Intel(R) I211 Gigabit Network Connection
Dispositivo de teclado HID (002)
GROUP : Ethernet Team - Realtek PCIe 2.5GbE Family Controller
Mouse compatível com HID (008)
HID keyboard device (008)
Intel(R) Wi-Fi 6 AX200 160MHz

Here I noticed something strange: are 3 keyboards and 2 mice. Watching the device manager I found 6 keyboards and 5 mice. Among the keyboards, there was a virtual keyboard “em Corsair composite virtual input device” and another “em Logitech G HUB Virtual Keyboard”. I also found a mouse “em Corsair composite virtual input device”. These are the strings that appear as the device's location. But, these virtual devices, probably created by Logitech G Hub and Corsair iCUE, is not allowed to wake the PC. Furthermore, keyboard and mouse do not even appear in the device list under that PCI to PCI bridge that is waking up the PC.

My attempt then was to get permission to wake the PC from the WiFi adapter, which is a “Intel (R) Wi-Fi 6 AX200 160 MHz”. I'm using Ethernet connections and WiFi is idle, so I decided to do a test. Anyway, the WiFi adapter was configured to wake the PC only when receiving a magic packet, as well as Ethernet adapters.

It… Problem solved! Now the computer can sleep in peace until it is woken by the keyboard, by mouse, for “magic pack” Wake-on-LAN on adapters Ethernet, or for any event scheduled at Task Scheduler, but no longer waking up randomly.

Because the “NVIDIA USB Type-C Port Policy Controller” kept waking up the PC? This remains an unknown for me, but at least the problem has been properly worked around. I've seen the problem reported in some messages on the forum NVidia, but without solution. It is likely that NVidia need to resolve this in your drivers and software.

Why the WiFi adapter kept waking up the PC even though it was configured to do this only when it received some magic packet? I also didn't know. Nesse caso é possível que alguma atualização de driver resolva o problema. Estou usando a versão 21.10.2.2 from 13/05/2019 e sei que existem versões mais novas, mas não instalei ainda porque estou investigando alguns problemas de tela azul que parecem estar relacionados com driver.

Share this article with your friends if you liked 😉 . The Skooter Blog needs your help in spreading to continue existing.

0 0 vote
Article Rating

Permanent link to this article: https://www.skooterblog.com/2020/12/26/problemas-e-solucoes-micro-acordando-constantemente-por-ponte-pci-para-pci-pciven_1022dev_1483subsys_87c01043rev_00311583659019/

Sign up
Notify about
guest
0 Comments
Inline Feedbacks
See all reviews
0
We would like to know what you think, Leave your commentx
()
x
Enable Notifications    OK No thanks