Help with reverting/repairing CSME of Dell XPS 15 9560

Special Topics Intel Management Engine
5

Unfortunately, no solution for the slow boot, however undervolting can still be done (which is a must for this laptop).

The solution to allow undervolt (the easy way) is to flash 1.18 bios, immediately enter the bios and factory reset the bios. Then boot again into the bios and disable capsule updates. Do whatever else you want and boot up. Within Windows, I use throttlestop and undervolt works (can be confirmed with CPUz)

More info
https://www.reddit.com/r/Dell/comments/hgzpdi/dell_xps_9560_bios_1192_disabled_undervolting/

Another method that might work on the newer bios, for me it did nothing on 1.18.
https://www.reddit.com/r/Dell/comments/hgzpdi/comment/fw7cyzd/?utm_source=share&utm_medium=web2x&context=3

6

Long boottime may be a sign of a corrupt ME region. That can’t be solved by an update since updates aren’t configured and don’t overwrite settings/data but only code.

That might help:

1 Like
7

One step further (see this).

I found the area for ME Reflash, but I still have to read through (again) the cleaning thread.

By the way, I was able to update ME to 11.8.92.4249_COR_H_DA_PRD_RGN.bin in the meantime.

8

I found this section in the PE32 Image:

0x3981F 				One Of: ME Unconfig on RTC Clear, VarStoreInfo (VarOffset/VarName): 0x13F6, VarStore: 0x1, QuestionId: 0x2B5, Size: 1, Min: 0x0, Max 0x1, Step: 0x0 {05 91 27 10 28 10 B5 02 01 00 F6 13 10 10 00 01 00}
0x39830 					One Of Option: Disabled, Value (8 bit): 0x0 {09 07 04 00 00 00 00}
0x39837 					One Of Option: Enabled, Value (8 bit): 0x1 (default) {09 07 03 00 30 00 01}
...
0x3D187 	Form: OEM Flags Settings, FormId: 0x27B1 {01 86 B1 27 25 14}
0x3D18D 		Checkbox: MEBx hotkey Pressed, VarStoreInfo (VarOffset/VarName): 0x6A0, VarStore: 0x1, QuestionId: 0x433 {06 8E 27 14 28 14 33 04 01 00 A0 06 10 00}
0x3D19B 		End {29 02}
0x3D19D 		Checkbox: MEBx Selection Screen, VarStoreInfo (VarOffset/VarName): 0x6A1, VarStore: 0x1, QuestionId: 0x434 {06 8E 29 14 2A 14 34 04 01 00 A1 06 10 00}
0x3D1AB 		End {29 02}
0x3D1AD 		Checkbox: Hide Unconfigure ME Confirmation Prompt, VarStoreInfo (VarOffset/VarName): 0x6A2, VarStore: 0x1, QuestionId: 0x435 {06 8E 2B 14 2C 14 35 04 01 00 A2 06 10 00}
0x3D1BB 		End {29 02}
0x3D1BD 		Checkbox: MEBx OEM Debug Menu Enable, VarStoreInfo (VarOffset/VarName): 0x6A3, VarStore: 0x1, QuestionId: 0x436 {06 8E 2D 14 2E 14 36 04 01 00 A3 06 00 00}
0x3D1CB 		End {29 02}
0x3D1CD 		Checkbox: Unconfigure ME, VarStoreInfo (VarOffset/VarName): 0x6A4, VarStore: 0x1, QuestionId: 0x437 {06 8E 2F 14 30 14 37 04 01 00 A4 06 10 00}
0x3D1DB 		End {29 02}
0x3D1DD 	End Form {29 02}

Do you think this would fix the issue?

Edit:
Since ME Unconfig on RTC Clear is enabled by default, removing the RTC (and obviously the laptop battery) should unconfigure it. However, I’ve already tried this years ago with no success.

I noticed that writing some variables take some time (same as changing stuff in the bios itself).

9

ME config is stored in ME region itself and can’t be undone completely without ‘re-initializing’ the ME region and flashing it back. It’s (like NVRAM) non- volatile and you can remove whatever battery as long as you like, but ME initialization won’t be impressed…

Follow the linked procedure.

10

I still did not take the chance to do this procedure.

I just came across this post and this image from the 9550.

image
image838×516 87.2 KB

Maybe this is an easier way to clear the ME region? Although I have no idea how to proceed with this.

11

I also came across the ME Cleaner tool

12

I have an XPS 15 9570. The motherboard/MB platform is identical to the
Precision 5530. Dell is still supporting the Precision 5530, and I can still apply Precision 5530 bios updates to my XPS 9570. I’m assuming you can do the same - apply the Precision 5520 bios update to your 9560.

It’s worth a shot, and if that bios won’t apply, the Dell bios program should warn and then auto crap out. You can also extract the latest xps_15_9560_1.24.0.exe bios and the Precision_5520_1.24.0.exe bios and hex editor compare them, but I’m almost positive that shouldn’t be necessary.

If I’m right, that may take care of your issue.

13

I didn’t know that. I successfully flashed v1.40 bios, but nothing changed about the slow booting.

14

The applicable drivers for the Precision 5520 will work on your XPS 9560. I would apply updated drivers, and specifically the 2435.6.35.0, A20 Intel Management Engine Components Installer.

It could be a number of issues, including, perhaps, bios settings. Good luck!

15

I updated the drivers, no luck (I wouldn’t assume that they’d fix the slow booting).

I did reset the bios multiple times in the past, that didn’t work either.

16

As lfb6 previous mentioned, this is most probably a corruption on the ME FW partition, NO reset, no bios update, etc will fix a corrupted ME FW image on a working system.
It must be reinitialized and flashed back or full SPI IC(s) programm with a fixed/initialized image.

17

No, they wouldn’t.

I’m having a tough time seeing how a dell bios update would corrupt your ME, but I guess it’s possible. If that’s it, you’ll have to follow the guide lfb6 pointed you to.

The issue with ME is that I’m not sure there’s a way to test it for corruption. I think you just have to assume it, and then clean and physically flash. I’m pretty sure it will require a physical flasher.

What I didn’t know (and still don’t know) is whether a new dell bios update would get around the “need to clean” issue, since dell bios updates will overwrite current ME data versions. I’ve had my issues, but thankfully never a corrupted ME.

18

IMG_20250118_205016
IMG_20250118_2050161920×2560 691 KB

Mine has a Winbond 25Q256FVEQ.

I have a CH341A and a programmer clip that seems to fit the size. I’ll see if I can pull a copy from it.

19

Carefull, this is a WSON IC pachage, the standart SOIC8 clips wont attach correctly.
You need a a needle type one or working on it desoldered from the pcb.

20

Yep. Unfortunately, mine is for a SOP8. While it does have the right pitch, the plastic sides don’t allow the pins to reach the soldering blobs. So I’ll have to order a specific clip.

To be continued…

21

@lfb6 (or anyone who wants to take a shot at this), Dell usually posts BIOS Recovery Image Files, and that’s the case here. The latest Precision 5520 has a BIOS Recovery Image File/BIOS_IMG.rcv, and Dell has instructions on how to apply the recovery. Would there be any benefit in trying to apply the BIOS Recovery Image File process?

1 Like
23

user lfb6 has a different “EGO” of mine

Well we found something we agree on.

1 Like
24

Regarding the recovery file for the 5520 you could’ve found out yourself. The file unpacks with Dell_PFS_Extract and its content has the same structure as a normal update. You can simply check the ME binary with MEA and found out that it’s ‘unconfigured’ (and it has ‘update’ in it’s name). I haven’t seen this changing for newer ME versions where update and region have a different structure.
.

Corruption:

If you had checked MEInfo output once you’d have found a line that says “ME File System Corrupted Yes/No”. if it says ‘yes’ you can be quite sure that there’s corruption, but if it’s no you can’t be completely sure that there’s no corruption.

If you had checked ME tools set more thoroughly you’d have found a tool called MEManuf which checks the ME and - run with ‘-verbose’ switch - gives some information about the ME state, se output for ME 14

MEManuf64 -verbose (ME 14)

LPC Device Id: 687.
Platform: Cometlake Platform
General FW Information
FW Status Register1: 0x94000245
FW Status Register2: 0x38850106
FW Status Register3: 0x00000030
FW Status Register4: 0x00004000
FW Status Register5: 0x00000000
FW Status Register6: 0xC0400000

CurrentState:                                Normal
ManufacturingMode:                           Disabled
FlashPartition:                              Valid
OperationalState:                            CM0 with UMA
InitComplete:                                Complete
BUPLoadState:                                Success
ErrorCode:                                   No Error
ModeOfOperation:                             Normal
SPI Flash Log:                               Not Present
Phase:                                       BringUp
PhaseStatus:                                 CM0_MKHI_HANDLER_STOP
ME File System Corrupted:                    No
FPF and ME Config Status:                    Committed
RPMC status:                                 OK

FW Capabilities value is 0x7DF6D147
Feature enablement is 0x5DF6D147
Platform type is 0x42001492
No Intel Wireless device was found
Feature enablement is 0x5DF6D147
ME initialization state valid
ME operation mode valid
Current operation state valid
ME error state valid
MFS is not corrupted
PCH SKU Emulation is correct

Request Intel(R) ME BIST status command… done

Get Intel(R) ME test data command… done

Get Intel(R) ME test data command… done

Get Intel(R) ME test data command… done
Total of 19 Intel(R) ME test result retrieved

Policy Kernel - Power Package : Live Heap Test - Passed

Common Services - LAN : Connectivity to NIC in M3 - Passed

Policy Kernel - Boot Guard : Self Test - Passed

VDM - General : VDM engine - Passed

GFX - General : Sampling engine - Passed

USBr - General : Storage - Passed

USBr - General : KVM - Passed

Common Services - LAN : Connectivity to NIC in M0 - Passed

AMT - KVM : Compression engine - Passed

AMT - KVM : Compare engine - Passed

PAVP - General : Verify Edp and Lspcon Configurations - Passed

PAVP - General : Set Lspcon Port - Passed

PAVP - General : Set Edp Port - Passed

IP Loading - TCSS IPL Tests : TCSS IPL Health Test - IOM - Passed

Policy Kernel - ME Password : Validate MEBx password - Passed

Common Services - EHBC State : EHBC and Privacy Level states compatibility - Passed

Common Services - EHBC State : Valid Embedded Host Based Configuration (EHBC) state - Passed

Common Services - Privacy Level : Valid Privacy Level settings - Passed

AMT - Power : Valid LAN power well - Passed

Clear Intel(R) ME test data command… done

MEManuf Operation Passed

If ME updates by FwUpdLcl don’t run you can be quite sure that there’s corruption. So if you have a machine where the firmware updates contains a ME update binary (like many Dell) and the bios region get’s updated but ME doesn’t you can be quite sure, too.

All other cases it’s on a ‘might be worth a try’- base

plutomaniac earlier asked users to run MEInfo and MEManuf, but there’s a lot of cases where you can’t come to a conclusion. So if in doubt about firmware, reset your firmware (NVRAM and re-initializing ME) to come as close to a stock ‘new’ firmware. If this is done properly and you still have problems with your machie you can rule out firmware issues for good.

2 Likes
25

Thanks once again, my friend. For the BIOS Recovery Image File issue, many have said that you can’t overwrite a corrupted ME with a new ME firmware to fix/correct the corruption, so I was wondering if implementing the Dell BIOS Recovery Image File process would overcome the “need to clean”. In other words, I was wondering if the Dell process was so thorough, that it did the cleaning at re-boot. Come to think of it, if you were not in boot dire straits, and could use the regular exe upgrade, I guess the same question would apply.

I still have my doubts about a Dell update corrupting the ME region, but one thing is for sure; your post will get to the answer/solution one way or another! Thanks again.