Xeon Owners Club

[crazydrve`]

Hello,

Does anyone know how to troubleshoot "General Protection Fault UEFI startup errors?

I cant seem to get rid of it fully and it keeps coming back every so often.

View attachment 1694479161259.webp

 

[lexluthermiester]

Hello,

Does anyone know how to troubleshoot "General Protection Fault UEFI startup errors?

I cant seem to get rid of it fully and it keeps coming back every so often.

View attachment 313208

This is one the Dell T620 in your specs?

 

[crazydrve`]

This is one the Dell T620 in your specs?

Yes, its a Dell T620 and Ive taken it down to just the 2 cpu and ram. ex e2967 and 256 ram

Is there a certain amount of time needed to drain the power? Or is there issues with SSD's? Ive used them ever since I got this but now lately its erroring on the uefi. Doing a bios mbr test now.

 

[lexluthermiester]

Yes, its a Dell T620 and Ive taken it down to just the 2 cpu and ram. ex e2967 and 256 ram

Is there a certain amount of time needed to drain the power? Or is there issues with SSD's? Ive used them ever since I got this but now lately its erroring on the uefi. Doing a bios mbr test now.

What happens when you remove the SSD(unplug it) and let the system boot without it?

 

[crazydrve`]

It still gives the error, the one thing I am not sure of is how to get into the actual error message using the serial cable, like it suggests.

I did get it working earlier today but only bios mode and with no ssd, and that was after I let it sit overnight.

It would also stall at initializing firmware interfaces

The Idrac only gives generic...


EFI0011: CPU Exception Type 0x0D: General Protection (Software).
2023-09-11T17:45:46-0500
Log Sequence Number: 678
Detailed Description:
A CPU general protection (type 0x0D) exception occurred in the pre-boot UEFI environment.
Recommended Action:
Recently installed hardware or updated software may cause this exception. Update or remove the hardware or software.

Edit:

I have it running now in bios mode, using a 3.5" hdd, and Server 2019 installed. So now I can at least do some updates and tests.

Correction, it just gave me the error again, no ssd connected no usb devices and only vga and network cords

 

[lexluthermiester]

It still gives the error, the one thing I am not sure of is how to get into the actual error message using the serial cable, like it suggests.

I did get it working earlier today but only bios mode and with no ssd, and that was after I let it sit overnight.

It would also stall at initializing firmware interfaces

The Idrac only gives generic...


EFI0011: CPU Exception Type 0x0D: General Protection (Software).
2023-09-11T17:45:46-0500
Log Sequence Number: 678
Detailed Description:
A CPU general protection (type 0x0D) exception occurred in the pre-boot UEFI environment.
Recommended Action:
Recently installed hardware or updated software may cause this exception. Update or remove the hardware or software.

Edit:

I have it running now in bios mode, using a 3.5" hdd, and Server 2019 installed. So now I can at least do some updates and tests.

Correction, it just gave me the error again, no ssd connected no usb devices and only vga and network cords

Honestly, this one has me stumped.

 

[crazydrve`]

Honestly, this one has me stumped.

Me too, its so inconsistent.

 

[lexluthermiester]

Me too, its so inconsistent.

Have you tried clearing the CMOS and resetting all of the BIOS settings?

 

[crazydrve`]

Have you tried clearing the CMOS and resetting all of the BIOS settings

Yep and the nvram.

 

[lexluthermiester]

Yep and the nvram.

Have you tried swapping video cards? Cards that are not UEFI compatible can cause issues on some UEFI boards, but I've never seen that error before. The Intel Management Engine might be misconfigured, but clearing the CMOS would have solved that issue short term. The only thing left that I can think of is the motherboard might be going bad and spitting out errors. The CPU would not be doing this and nothing else plugged in would give UEFI warnings like this either.

 

[crazydrve`]

Have you tried swapping video cards? Cards that are not UEFI compatible can cause issues on some UEFI boards, but I've never seen that error before. The Intel Management Engine might be misconfigured, but clearing the CMOS would have solved that issue short term. The only thing left that I can think of is the motherboard might be going bad and spitting out errors. The CPU would not be doing this and nothing else plugged in would give UEFI warnings like this either.

I only have the perc h710p installed and using onboard video.

Do you know how much time is needed to let it sit and have not power, drain the power from it?

 

[lexluthermiester]

Do you know how much time is needed to let it sit and have not power, drain the power from it?

10seconds as long as you pull the cord, press & hold the power button.

 

[crazydrve`]

10seconds as long as you pull the cord, press & hold the power button.

ok, ive been doing 20 secs

 

[lexluthermiester]

ok, ive been doing 20 secs

My hunch is a bad motherboard.

 

[crazydrve`]

My hunch is a bad motherboard.

God I hope not

 

[lexluthermiester]

God I hope not

The good news is that for such a high-end board, they're only $250ish on ebay, stateside.

 

[crazydrve`]

The good news is that for such a high-end board, they're only $250ish on ebay, stateside.

Well Im still gonna see if its just something weird, right now I got into the Lifecycle controller, so I can run some tests.

If you got a used system off ebay, etc. what would be the steps to see how stable and working a system it is?

Obviously physical check, no bad caps, burn marks, etc. clear of dust.

I am currently able to have usb mouse and keyboard with idrac network and 1 network port being used.

No pcie cards installed and no hdd or blueray connected.

Im trying to think of what else?

 

[Edwired]

Seems like some sort of bios corruption through byte rot can happen on older hardware try and reflashing a clean bios to see if that clears up or seems like the bios chip going bad

 

[crazydrve`]

Seems like some sort of bios corruption through byte rot can happen on older hardware try and reflashing a clean bios to see if that clears up or seems like the bios chip going bad

I'm working on that right now, I just finished the through Lifecycle controller hardware test.

Now to see if I can boot a windows OS...

Update, I have windows server 2019 installed, so now for updating and testing.

 

[Edwired]

I'm working on that right now, I just finished the through Lifecycle controller hardware test.

Now to see if I can boot a windows OS...

Update, I have windows server 2019 installed, so now for updating and testing.

Any progression since?

 

[crazydrve`]

Any progression since?

I currently have server 2019 running on a hdd in bios/mbr mode using onboard sata, ran all updates, now installing omsa, etc before I go any further.

Making sure I go real slow and carefull.

 

[Edwired]

The whole thing about uefi compatibility it shouldn't really affect the graphic card per say as those are usually backward compatible regardless it seems something along the microcode for the CPU to fit to the board unless you got a different revision or something on the.board is misjumpered to a different configuration is what I am thinking

 

[crazydrve`]

The whole thing about uefi compatibility it shouldn't really affect the graphic card per say as those are usually backward compatible regardless it seems something along the microcode for the CPU to fit to the board unless you got a different revision or something on the.board is misjumpered to a different configuration is what I am thinking

This was all working great last week, uefi mode, ssd drive for os. 20+ TB in raid using perc h710p. So not really sure what happened yet

 

[Edwired]

This was all working great last week, uefi mode, ssd drive for os. 20+ TB in raid using perc h710p. So not really sure what happened yet

There have to be some sort of log explaining why
Found this about the expections error
It stated causing the error by trying to set the onboard nics to default in the bios. If I don't do that setting than all is good.

Came across this on searching through google

Another possible cause is the socket or dimm have bent pins or corrosive somewhere that is interferening with something on the board. Unless the one of the cpu is a dud or memory stick gone bad

 

[lexluthermiester]

Im trying to think of what else?

Umm..

Seems like some sort of bios corruption through byte rot can happen on older hardware try and reflashing a clean bios to see if that clears up or seems like the bios chip going bad

This might be a thing worth trying.

Have you tried reflashing your firmware?