Zitat von plutomaniac im Beitrag #15

---

I don't know. There are three models at Dell: "Latitude E6420", "Latitude E6420 ATG" and "Latitude E6420 XFR". I chose the bare "Latitude E6420".

---

Zitat von plutomaniac im Beitrag #17

---

At their website of course, Support > Drivers & Downloads.

---

Of course, I've splited it and flashed both. Reverse of making dump procedure. But how did you made programmer.bin from hdr?
P.S. Tested with ES cpu. No changes (((
So I think ES CPU blacklisted in bios

@rustam2v - I have a few questions/thoughts, are we sure it's 206A2 CPUID? Maybe BIOS doesn't know for sure, or assigns that incorrectly? Can you please link to listing on ebay where you got these. Did you ask the seller what the s-spec is or what CPUID should be used?
Older 206A2 microcode may work better than latest/newer, so that should be tested as well.

Additional thing that came to mind is virtualization and Vt-d etc, disable this with other CPU installed and save, then put back in this CPU. This may help if this CPU does not have those functions enabled.

You said >> Also this theory supported by mabx_main section in bios region, the only CPU ID present there is 206A2 << Here, first instance = half the 206a2 is the end part of an integer extraction command in assembly (20), rest 6A 20 = push 20h, similarly exact same for next two instances in the MEBX_Main
(So this unrelated, just coincidence it's using these values)

Here, I made many single byte changes on various modules (CPUDXE First), where a check on 206a2/a0 06 02 is performed (change from setup byte if=1 to if= 0) Probably no help, but first thing I see to try, this is built on dump you gave at post #10 (FD unlocked also)
Additional edits on 4-5 modules included too in a second BIOS, which adds to the above one module change, similar check at SMBIOSBoard / Powermanagement / MemoryInit / CPUPEIBeforeMem
*** Never mind! I can't replace any modules, UEFITool 25 and 26 ail on rebuild, even if working on BIOS region only. PhoenixTool fails on insertion of edited module with module count mis-match. MMTool can't open properly, H20EZE fails to open with any version that allows module replacements properly
So, while I may be able to fix this issue, I can't modify the BIOS!! I've never seen such a thing!

**** OK, I can edit the stock BIOS region extracted from the HDR's (tested A08 and A20) - not sure where this leaves us though because you don't want to use those BIOS regions.


plutomaniac - What if you disable ME? Or, disable/enable AMT/MEBx, since he mentioned it was looking for that possibly.
Can you tell why I can't replace a module from his dump and rebuild with UEFITool? If you want to try and see what I mean, the first one I was trying to replace was CPUDXE PE32 body
Weird I can't rebuild his dump, or region extracted and rebuild region either, but I can edit the stock region from HDR extracts. This bums me out, spent all that time finding what might possibly fix this, and then can't apply it

@plutomaniac

Zitat

---

BIOS A01 and ME v7.0.4?

---

yes, versions as you specified
@Lost_N_BIOS

Zitat

---

are we sure it's 206A2 CPUID?

---

I'm not sure, but this only CPU ID displayed in bios setup without dependence on bios version. And A25 started to bios setup after 206A2 microcode injection. Look this http://www.streamtech.net/Downloads/bios.jpg

Zitat

---

virtualization and Vt-d etc,

---

I have tried to disable all possible dependence points - no changes. As I described before, the problem not in downgrade ME, but that bios initialized and it "well know" that ME will failed before it starts - missed MEBx choice in boot manager! This line missed if CPU replaced with ES.
http://www.streamtech.net/Downloads/F12.JPG
Then any boot choice (inc. diagnostic) ends as described before.
"Failed to get ME Kernel Host Interface version ..."
"Error sending End of Post message to ME"
"(A7) Me FW Downgrade - Request MeSpiLick Failed"

Zitat

---

CPUDXE PE32 body ...Weird I can't rebuild his dump, or region

---

I think it because 0.26 have not read CPUDXE PE32 body splited to different location, but N55 can, but still alpha editing

Needs to ask @CodeRush add "defrag" function to NE55 release )))

P.S. I'm free for experiments with bios, because I can reflash with A07 at any moment, and for flashing i need only remove keyboard.

@plutomaniac

Zitat

---

My best guess is that this mistake happened during the process to add the 206A2 microcode manually.

---

206A2 microcore present in dell bios. I have injected it only to A25 bios.
I have to return the BIOS to version A25, because the earlier ones (inc. A19) do not see 16GB ram. And because the problem is not in ME, but in the initialization of ME. If necessary, it's possible replace the ME region from A01 by @plutomaniac
But apparently the problem is CPU blacklist in the bios region.
So i provide you A25 bios dump shortly

Ready A25 pure Dell dump

May be THIS will helpfull
FPT full dump, it's lite different with programmer dump one, some changes in bios setup.

@rustam2v - FPT Dump should match programmer dump, except for what FD locks you out of dumping (ME, possibly PDR, EC, etc) BIOS setup should be same for either programmer dump or FPT, you do mean "Setup" module from within BIOS region correct?
I'm don't quite get what you are trying to explain about CPUDXE module, doesn't make sense to me the CPUDXE module is not and should not be split/fragmented to somewhere else, hopefully that is all only caused by what plutomaniac found instead << Yes, this was only issue, what he found, confirmed before posting!

Yes, I saw your images of the BIOS from the links previously, my thoughts still apply, especially since we can't see s-spec on IHS-less CPU. Is there s-spec printed in the other black squares?
Can you scan and upload that QR code to somewhere and see what it says. Or provide better image of that so I can upload to somewhere, less blur, more crisp, macro, up close.

For the ME issue, have you tried what's outlined here exactly, maybe simple fix!
https://www.dell.com/community/Desktops-...led/m-p/4728275
Similar mentioned here too - https://forums.tomshardware.com/threads/...failed.2636825/

If those do not help, please put in other CPU and get into windows and run MEInfoWin -verbose + MeInfoWin -fwsts (From V7 ME System Tools Package)
Then plutomaniac can check the result images and see if the ME looks healthy or not

plutomaniac Thanks for noticing this issue with the BIOS provided previously! Hopefully that was the cause of my attempts at a rebuild failing. (It was, thanks! )

I am not sure the changes I made will help, but they do look promising to me, however there may be others aside from ones I mentioned above.
I've not done this type of inspection and bypassing CPU checks before with much luck, but I've only tried a few times and never really found things as easily as I did here so there is hope.
Maybe after this, if successful I can go back and possibly figure something out for the "SAGE" guys, that would be nice

rustam2v - since 16GB or 8GB etc is not part of the current issue/concerns, and because I would have to make a bunch of new edits on your A25 dump, for now I will use the file plutomaniac fixed above to try and apply the changes I meant to the other day, since they use this versions BIOS files.
If this works, then yes I will redo on A25 for you. Here is two BIOS, one with CPUDXE edit only, other with all rest I mentioned + CPUDXE edit too. Please test both, programmed in, unless you have unlocked FD current, then you can FPT entire BIOS file in there (FD is unlooked in these too)
http://s000.tinyupload.com/index.php?fil...121857309486328

Note to all, please check all padding files in last volume (maybe other volumes too) before/after edit with UEFITool 26, replacing some modules remove 2x padding here (Sometimes relocates one too), other edits even within same volume do not, so it's random, always check or use 25 instead.
Some modules can only be extracted/replaced with MMTool, otherwise similar issues as mentioned above, all due to editing PEI modules/known issues.

Grrrrrrrr!!! Still stuck, two modules in PEI volume I cannot replace without breaking BIOS (padding files issue in UEFITool, Invalid FFS in MMTool no matter how extracted from MMTool and edited, checksums corrected etc)
Got it, see above for links, what a wrestling match that was

Zitat von Lost_N_BIOS im Beitrag #29

---

FPT Dump should match programmer dump, except for what FD locks you out of dumping (ME, possibly PDR, EC, etc) BIOS setup should be same for either programmer dump or FPT, you do mean "Setup" module from within BIOS region correct?

---

Zitat von Lost_N_BIOS im Beitrag #29

---

to me the CPUDXE module is not and should not be split/fragmented to somewhere else, hopefully that is all only caused by what plutomaniac found

---

Zitat von Lost_N_BIOS im Beitrag #29

---

I saw your images of the BIOS from the links previously, my thoughts still apply, especially since we can't see s-spec on IHS-less CPU. Is there s-spec printed in the other black squares? [/qoute] I've did these all, with macro QR photo. 1st post of thread. These were as pics url, now inserted as pics for normal view.

[quote=Lost_N_BIOS|p73862]For the ME issue, have you tried what's outlined here exactly, maybe simple fix!
https://www.dell.com/community/Desktops-...led/m-p/4728275
https://forums.tomshardware.com/threads/...failed.2636825/

---

Zitat von Lost_N_BIOS im Beitrag #29

---

If those do not help, please put in other CPU and get into windows and run MEInfoWin

---