Uncorrectable PCI Express Error

[GordonThree]

I have a dead server that fails to boot with a Windows STOP error and this entry logged in ILO

Uncorrectable PCI Express Error (Embedded device, Bus 0, Device 28, Function 7, Error status 0x00100000)

Server:

Proliant DL380p Gen8, dual 2620 xeon, 256GB ram
Embedded HP Ethernet 1Gb 4-port 331FLR Adapter
Embedded Smart Array P420i Controller
PCI-E Slot 1 HP Ethernet 10Gb 2-port 560SFP+ Adapter
PCI-E Slot 2 HP Ethernet 10Gb 2-port 560SFP+ Adapter

Steps:

1. Installed Server 2016 Datacenter with Desktop Experience via virtual cd, no issues
2. Windows Update, reboot, no issues
3. Ran the latest (July 2017) SPP on the machine, reboot unsuccessful - pci express error

I notice HP specifies BIOS P71 for Server 2016, however Intelligent Provisioning and SUM is not seeing P71 as being available, showing the 2015 P70 as latest. However Support Downloads does show a P71 available for this system.

Any ideas?

[jvplopper]

I encountered this situation almost exactly.  Installed Server 2016, reboot it was fine, updated with lates SPP and I've got BSOD with the same error you mentioned.  

[JamesMoondew]

Last known good configuration saved us from a continuous blue screen cycle with NMI errors.

I found it was the matrox graphic driver on my ML350 G8 that caused this.  Ran the SPP with everything but the updated graphics driver and have had no issues so far.

Regards

James

[cbka]

Same error here

installed Server 2016 Standard and after updates and so on finaly matrox driver and bang - BSOD with same Stop code

any news on this ?

Cheers,

Chris

[GreeTz]

Same problem here, I receive the BSOD at the moment i connect to the machine via Integrated Remote Console to this machine. 

Server 2016, System ROM P71 01/22/2018 ILO firmware version 2.55 Aug 16 2017

PCI Bus 04/04/2018 20:24 04/04/2018 20:17 2 Uncorrectable PCI Express Error (Embedded device, Bus 0, Device 28, Function 7, Error status 0x00100000)

System Error 04/04/2018 20:24 04/04/2018 20:17 2 Unrecoverable System Error (NMI) has occurred. System Firmware will log additional details in a separate IML entry if possible

[SeanPaul]

We've just had exactly the same issue - ML350p G8 upgraded to Server 2016 and after running the SPP the performance was terrible / ILO RC causing blue screen instantly. At one point the IML advised to replace the processor!

So far removing the Matrox graphics driver has massively boosted performance and stopped the blue screens.

All working perfectly now with the basic graphics driver that came with 2016.

[Timau]

Same problem as well... What a bloody waste of time!

Clean install - ran SPP 2018.03 BANG BSOD

Had to reinstall OS from scratch as there wasnt a roll back :(

Anyone tried later versions - doesnt seem to be any other than a (B)

https://support.hpe.com/hpsc/swd/public/detail?swItemId=MTX-52a46bcb04a84bb3bb20b4c710#tab-history

[JS7]

Solution:

- Boot into Safe Mode

- Open Device Manager > Display Adapters > Matrox G200eh (HP) WDDM 2.0 > Properties > Driver > Roll Back Driver

- It should revert to previous version: 4.3.1.5 (Date: 7/12/2016)

- Reboot into Default Mode

IMPORTANT: 

HPSUM (7.6.0 + SPP_2017.04.0) is the last production SPP to contain components for the G7 and Gen8 server platforms.

HPSUM (8.0.0 + SPP_2017.07.0) is the new production SPP to contain components for the Gen9 and Gen10 server platforms.

[Zombie64]

You can use the CMD in the Recovery Mode and edit the Registry to Disable the servive as instructed at:
https://support.microsoft.com/en-us/help/927525/after-you-install-a-device-or-update-a-driver-for-a-device-windows-vis

There you will see the service correspondent to Driver for Matrox \ MxG2hDO64.sys 
Then set the 'Start' key to 4, which will disable it. Reboot computer and all good

----

1. In the System Recovery Options dialog box, click Command Prompt.
2. Type the following commands. Press ENTER after you type each command.
   cd \Windows\INF
   notepad setupapi.app.log
3. Note the date at the start of each new device or driver installation section. Use these dates to determine the last driver that was installed.
4. After you determine which driver was installed last, determine whether the driver is required to start the computer. To do this, read the information in the section of the Setupapi.app.log file that describes this driver. If the driver is related to the disk controller or to the chipset, or if the driver is provided by the operating system, search for the driver name and for the symptom of the problem on the following Microsoft Web site:
   http://www.microsoft.com
   Determine whether the driver can be disabled before you continue. If the last driver that was installed is not required to start the computer, go to step e.
5. At the command prompt, type regedit, and then click OK.
6. Click HKEY_LOCAL_MACHINE, and then click Load Hive on the File menu.
7. Locate and then click the C:\Windows\System32\Config\System file, and then click Open.
8. In the Load Hive dialog box, type Offline, and then click OK.
9. Expand System under 'Offline' Hive , and then click Select.
10. In the right-pane, locate Current, and then note the value in the Data column.
11. Expand ControlSet00x, and then expand Services. x is the value from the
    Data column that you noted in step j.
12. Locate the subkey that corresponds to the last driver that was installed. If you cannot locate a match, click Services, click Find on the Edit menu, type the name of the driver in the Find what box, and then click Find Next.
13. Click the subkey that has the driver name.
14. In the right-pane, right-click Start, and then click Modify.
15. In the Value data box, type 4, and then click OK. This step stops the driver from starting.
16. Locate and then click the following registry subkey:
    HKEY_LOCAL_MACHINE\Offline
17. On the File menu, click Unload Hive, and then click Yes in the Confirm Unload Hive dialog box.
18. Exit Registry Editor.
19. Restart the computer.

[mwiora]

finally resolved with 224:

https://support.hpe.com/hpsc/swd/public/detail?swItemId=MTX_de2c0f0e28124d849c6fa5c0ac#tab4