Well, reconfiguring the ME was the first step, just to be sure that it’s OK. Having used this many posts for flashing a clean configured ME makes it look a littele like a heroic action, but it normally would’ve been download of stock bios and ME tools and one command line.
As stated in the first part of the thread, the error has to be in the bios region. There’s very little board specific data in bios, one padding and the NVRAM store for variables, see picture. Both areas could contain corrupted information that can cause your problems, both areas won’t be overwritten when updating bios the normal way. The padding contains blocks related to several bios versions, looks like storage for configuration/overclocking(?) data. The static bios parts in your bios are OK (100% identical to stock 3.70 bios).
If you don’t have done so already create an USB stick for booting into DOS, copy the ME tools to the stick 1.) Run MEINfo, check if this gives an error 2.) Run MEINfo -verbose > MEinfo.txt, post the complete content of MEInfo.txt (preferably put it into a spoiler to save some space)
3.) Check on which bios you are (Multi bios A or B?) 4.) Switch to the other bios, don’t boot into OS 5.) Check if this bios shows ME 6.) Make a dump of this bios, regardless of whether working properly or not (fpt -d spi_2.bin) 7.) Zip this dump, post file and info if this was bios A or bios B
Please post the relevant information into the thread. Maybe you misunderstood ‘saving space’, this is not storage space, it’s display space. It’s difficult to follow a thread if one post has half a page with pictures, for example, so putting them into a spoiler (eye-symbol, fourth from right) makes a single post taking less (view-) space.
The Windows XP command window is not DOS.
Relevant infor from MEInfo: "CurrentState: Normal ManufacturingMode: Enabled FlashPartition: Valid OperationalState: Bring Up InitComplete: Initializing BUPLoadState: Success ErrorCode: No Error ModeOfOperation: Unsecured mode by H/W jumper Phase: BringUp ICC: Valid OEM data, ICC programmed ME File System Corrupted: No PhaseStatus: UNKNOWN
HECI device is found to be disabled.“
This is all conclusive with either a (non documented?) service- jumper set or a bios option for firmware or ME servicing/ downgrade enabled.
Check the board for a undocumented jumper and check the bios for an option like “ME servicing enable”, “ME downgrade enable”, “Firmware unlock”, “Firmware downgrade”, Firmware servicing” … On MSI pictures there’s at least one undocumented switch beside the 2-Digit Debug Code LED, is there a label?
Hi, i followed your latest instructions, and the files generated are in the Zip file attached [HERE] (saved to Dropbox as total file size is 33mb) many thanks, hopefully you can spot something from the files …
thanks , I will check in the Bios for that possibility.
i will also try again from a proper Dos boot, i thought using a Ramdisk (Hirens boot CD) was just the same as using Dos, since its all stored in “ram” …? perhaps not…
On the board, there is a 4-pin ‘sys-fan’ header right next to the Digital display panel, but there is also a 2 pin header just above the 19-pin USB3 header. But thats just a Chassis Intrusion header. see photo attached… (spoiler)
How is it though, I am still able to flash to the ME section successfully, with this line in operation ModeOfOperation: Unsecured mode by H/W jumper ?
OK, saw that switch on picture on MSI page for this board, but it’s no longer there:
This 'hardware pinmod' is an option for allowing flashing of the ME region, used on many server boards, or dell bussiness machines for example. ME is not working with jumper 'on' but can be flashed, or is working with jumper 'off' but can't be flashed (if ME is no longer in manufacturing mode = FD set to standard values which don't allow for flashing the ME region).
Unfortunately I remember also one case when this was non- solvable, possibly a hardware- fault in the mainboard. But since your bios has a MePolicyInit dxe and MeFwDowngrade dxe modules there might be an option in bios setup.... (but that's guessing...)
Thanks, but you don’t need to do this again, the information needed from this we already got! The important thing is now if there’s a undocumented switch/ jumper on the mainboard, or if there’s a bios setting.
Did you use the built-in M-flash utility when you updated your bios?
The settings related to ME reflashing are in a possibly hidden Intel- submenu:
One Of: Me FW Image Re-Flash, VarStoreInfo (VarOffset/VarName): 0x1313, VarStore: 0x1, QuestionId: 0xD0B, Size: 1, Min: 0x0, Max 0x2, Step: 0x0 {05 91 3F 14 40 14 0B 0D 01 00 13 13 10 10 00 02 00} 0x3A4FA One Of Option: Auto, Value (8 bit): 0x0 {09 07 76 03 00 00 00} 0x3A501 One Of Option: Disabled, Value (8 bit): 0x1 {09 07 F6 13 00 00 01} 0x3A508 One Of Option: Enabled, Value (8 bit): 0x2 (default) {09 07 F5 13 30 00 02} 0x3A50F End One Of {29 02}
One Of: ME State, VarStoreInfo (VarOffset/VarName): 0x16FB, VarStore: 0x1, QuestionId: 0xD0C, Size: 1, Min: 0x0, Max 0x1, Step: 0x0 {05 91 5B 14 5C 14 0C 0D 01 00 FB 16 10 10 00 01 00} 0x3A60B One Of Option: Disabled, Value (8 bit): 0x0 {09 07 F6 13 00 00 00} 0x3A612 One Of Option: Enabled, Value (8 bit): 0x1 (default) {09 07 F5 13 30 00 01} 0x3A619 End One Of {29 02}
One Of: Me FW Image Re-Flash, VarStoreInfo (VarOffset/VarName): 0x1316, VarStore: 0x1, QuestionId: 0xD0F, Size: 1, Min: 0x0, Max 0x1, Step: 0x0 {05 91 3F 14 40 14 0F 0D 01 00 16 13 10 10 00 01 00} 0x3A681 One Of Option: Disabled, Value (8 bit): 0x0 (default) {09 07 F6 13 30 00 00} 0x3A688 One Of Option: Enabled, Value (8 bit): 0x1 {09 07 F5 13 00 00 01} 0x3A68F End One Of {29 02}
0x3A69D One Of: Local FW Update, VarStoreInfo (VarOffset/VarName): 0x1317, VarStore: 0x1, QuestionId: 0xD10, Size: 1, Min: 0x0, Max 0x1, Step: 0x0 {05 91 41 14 42 14 10 0D 01 00 17 13 00 10 00 01 00} 0x3A6AE One Of Option: Disabled, Value (8 bit): 0x0 (default) {09 07 F6 13 30 00 00} 0x3A6B5 One Of Option: Enabled, Value (8 bit): 0x1 {09 07 F5 13 00 00 01} 0x3A6BC End One Of {29 02}
They can be found in the NVRAM as IntelSetup. Problem is that in your dumps there are several of them, and it’s unclear which one is used. Probably the M-Flash utility has to change some of these settings to get access to the Me region for flashing, but doesn’t delete the entry afterwards.
1.) Flash a stock bios 370 with fpt: fpt(w) -f E7882IMS.370 (Link to bios 370)
If flash is not successfull make picture of error message and post, stop here
If flash is successfull continue with 2.)
2.) Clear CMOS
3.) Boot into bios (don’t boot into windows!) Check ME info displayed in bios
4.) Load default values for bios (‘restore defaults’ on ‘save and exit’)
5.) Reboot into bios
6.) Check again ME info displayed in bios
7.) Dump the complete bios with ftp(w) -d 370_1.bin Attach/ post file
8.) Stop here, don’t use this bios for booting into windows, switch to the other bios if this is necessary (Your DMI information is missing at the moment, so you might lose Windows activation)
Thanks! This is not good. I hoped to see a relatively clean NVRAM now and to be able to see changes done by the upgrade process, but there are already 11 IntelSetup entries after once loading the defaults and one reboot…
Just to be sure- please run once again MEInfo -verbose on the freshly installed stock bios 370 and post the result.
ok thanks, i will do this later today. There is a weird symptom on this board and I have read it somewhere online, about it rebooting itself once after you have made changes in the Bios. And Im wondering is this throwing off the results you are expecting…
I.E. Lets say I am in the bios and I save some settings (not sure what one triggers this but) then hit save and reboot. It will shutdown and reboot and then a few seconds later, it repeats that before it even reaches the Bios. I know there is an article somewhere on it online and the MSI techies did give a "valid" reason, but I cant rem what it was. If i find it i will post it.
will redo the 370 flash and verbose test again shortly…
Thanks, that’s as expected. I checked all the NVRAM entries for the chipset, not a single one with settings which would cause this behaviour from bios side.
Please check this spot on your board, see the 2 contacts JME1?
A connection there would be putting the ME in the operation mode "Unsecured mode by H/W jumper" for servicing the ME/ Bios
Any chance that something is shortening these contacts on the topside or bottomside og the board? Picture?
Unfortunately though, after checking thoroughly (on both sides), I can see nothing causing a short on the two points.
I see on your image that the dual-bios chips are easily accessible, but on my board there is a “cosmetic” cover over them obstructing access. And I’m not sure exactly how to remove it. (cant see how its attached)
i do have a USB CH341a Programmer , if this all went pear-shaped …lol!!
I’m sorry, but there’s nothing more we could do here.
ME firmware is OK, tried both cleaned, and from stock bios
Bios is OK, Intel ME servicing settings are hidden, but have in all NVRAM entries the correct values.
The board still behaves as if this JME1 jumper was ‘on’…
One could try an older bios (flashed via fpt), at least first bios (3.0) having a different chipset initialisation, but that would be a wild goose chase.
One could read the ME variables on the PCI bus and try to find out if the information found in NVRAM is congruent, but that again would be a wild goose chase and would need reading a lot of documentation, if even possible.
One could search for schematics and check if there were other connections where the lines to the JME1 jumper could be shortened…
At the end you might flash back your own bios to the bios where you flashed the 370 stock bios with fpt since there’s some DMI Information. But as you showed already- it’s not much infor, not even a serial number.
Please write if you find out something! Good luck!
Did you buy this board new? Did you ever see a working ME on this board?
If the ME was set into servicing mode while you owned the board (You’ve seen it once working), then it either should be something you did actively yourself like setting a jumper, changing a switch position or it did just happen, then it’s a hardware fault.
yes the board was brand new , but bought a few years ago when first released. From day 1, it always had that issue that everyone has had with it where, sometimes if left powered off (at the mains), it would completely lose the killer wifi and bluetooth function…like, simply vanish (internet is full of others posting same issues, with this board…including this forum).
But yes, ME was definitely there before.
I have plugged in and removed that many things on the board and flashed it that many times over the last few years, i just dont know when it happened, that ME vanished…?
There was a strange issue a few weeks ago come to think of it… where an Intel NIC card I had in failed. (actually caused the board to power off when in its slot…?) So it was binned and i went back to using the onboard Killer Lan adapter.
At that time, i took the opportunity to remove the Killer wifi/BT module from the board, as it was never used. Im wondering now did the faulty Lan card blow something on the motherboard ? If it did, its weird that only Onboard Sound and the Via VL805 USB controller was affected ? (Yes, the Via VL805 USB controller chip is gone also, not detected in windows at all? as is Onboard sound.)
but to me , this is all ME related. I dont understand how I can easily flash the board with Full or partial (ME) Bios without any issue, but that after rebooting, Bios reports ME as being , N/A…
