I thought so too, but then I compared an output in hex and found lots of differences
Here, compare this file output IFR in text in hex compare, lots of removed, added, and modified + 1kb difference without even looking in hex

Though, I'm still not sure which is better, or newer? https://www.sendspace.com/file/pysvul

@davidm71

Today I installed 8700k CPU and tried to find variables in RU.efi. But I can´t find 1FAB and following, setup module ends with 10FC. Did I take a look into the wrong module or did I miss something?

Edit: Next I will try is to take screenshot of every page in setup module, change core 1-4 in bios to different value, and take a look if some bits are changing.

@oldirdey ,

Can you please dump your tables to a file and post the txt file in a zip? I think F2 dumps them to a file after setting random values in your cpu ratio tables. I did it last week but can't remember which F-Key to be certain.

Thanks

Also what bios version are you using?? And have you tried hitting PG-Down?

Edit:

Debug with RU:
Save registers to a file including:
Single page F2, memory binary SHIFT-F2 and all PCI registers

@oldirdey

I can answer why you may not find that variable in your setup file. Thats because the variable exists in anothers Varstore. Varstore 34 to be exact which corresponds to:

 
QuestionId: 0x290D equals value in list (0x0) {14 08 0D 29 01 00 00 00}
0x4D178 			String: 1-Core Ratio Limit, VarStoreInfo (VarOffset/VarName): 0x1FAC, VarStore: 0x34 
 

 
0x4B2BC 	VarStore: VarStoreId: 0x34 [A3A3B874-7636-4182-BA1E-F52C584494E3], Size: 0x245D, Name: VARSTORE_OCMR_SETTINGS_NAME 
 
{24 32 74 B8 A3 A3 36 76 82 41 BA 1E F5 2C 58 44 94 E3 34 00 5D 24 56 41 52 53 54 4F 52 45 5F 4F 43 4D 52 5F 53 45 54 54 49 4E 47 53 5F 4E 41 4D 45 00}
 

 
 

On my X99 system items and variables related to the cpu originated in the IntelSetup module. I couldn't find that one in your bios so it could be any module but I think its findable with trial and error.

Zitat von davidm71 im Beitrag #28

---

I just wanted share the latest version of Universal IFR Extractor ver LS 0.3 by LongSoft which is a fork of Donavon6000's IFR Extractor which only goes to version 0.5 I believe. I compiled it myself for 64 bit Windows as LongSoft (aka CodeRush) only published the source code and I got to say it does a nicer job in that its more descriptive translating the byte code.

Here it is for others to play with: http://www.mediafire.com/file/777arjm4cb...tractor.7z/file

---

@davidm71 I will give you more information later from home when I have access to the system. Will try to find the other module and if still needed, uploading the dump files. As for now thank you very much for helping me out.

@mtothaj and @oldirdey

Glad the new version worked. Someone posted ver 0.7 a few posts back you should also try.

As far as finding that module at this point its going take some rolling up our sleeves to find it.

If we can't find it I would suggest asking Coderush.

@davidm71 The worst case scenario at this point is to take a look into each module. Time consuming but doable. I am ready to do my homework :)

@davidm71 @oldirdey - I noticed this a while back, but compared to e.g. the MSI X99A Godlike Gaming bios you mentioned, the ASUS bioses have been prepared in a lazy / sloppy way in that the text descriptors for the bios guid's are missing i.e. text column detailing module names in UEFI tool is not populated. This makes it harder to find the respective module as one has only the guid numbers to go on. However, with both bioses being AMI from a similar time frame, they mostly match up, so it is possible to use the MSI bios as a refrence table for the guid names to narrow down the search.


Asus bios:






MSI bios:

@mtothaj

Try using Uefitool NE Alpha 51

@davidm71 @mtothaj I found the module, it´s named VARSTORE_OCMR_SETTINGS_NAME like mentioned in post #34

First I changed value 204F to 24 and after additional 2078 to 23. But nothing happend.



As far as I know, under full load and if all cores used, the core clock does not exceed the value set for the last core. So it must be 35 (23 in hex), but windows shows 37 on all cores, like set on core 4 in bios from me.

@oldirdey

Are you sure thats the correct module? For example if you set values in bios for 1..4 core ratios do you note the change in values in the OCMR module?

@davidm71 I am very sure about the module. In ru.efi the module shows the value according to what I set in bios. Here are my bios settings for core ratio. Multiplier 45-42 is choosen. (take a look at the bottom of picture, bios gives a hint to core 5 "4-core ratio value must be higher or equal to 5-core ratio value...")



Here is what I have in RU.efi. Values matching bios values.



Then I changed values for 5-core and 6-core ratio in ru.efi to 41 and 40 (hex 29/28)

Nothing happens in Windows, but Intel XTU does notice that there is some changes in bios/nvram. The little X in circle behind core 5 and 6 has the Info "undo changes" And I am wondering why Intel XTU shows core 5 and 6 in yellow and not in blue font. Maybe cause it´s not adjusted via bios. Under full load system goes up to core 42, that´s the value set to 4-core ratio in bios.



And if I then use XTU to change values for 5-core or 6-core ratio, font changes to blue color and changings are accepted by windows. Maximum core clock under full load is now 40 instead of 42, before I used XTU to change value for 5-core and 6-core ratio.

@oldirdey,

My only guess is maybe there is an undocumented variable that sets the number of cores to control via bios setup.

For example when adjusting the fit table I noticed there was a value corresponding to the number of entries in the microcode list.
Such that changing a 4 to a 5 added another microcode to the list of microcodes that could be recognized by the bios.

Perhaps there is a similar variable or value. In anycase changing the values corresponding to the 5th and 6th cores creates a recognizable change in XTU?

Thats a positive step forward. Dump your variables to a file and I'll help you look.

Edit: Maybe this:

One Of: Active Processor Cores, VarStoreInfo (VarOffset/VarName): 0x4D0, VarStore: 0x1, QuestionId: 0x1DF, Size: 1, Min: 0x0, Max 0x3, Step: 0x0 {05 91 44 01 45 01 DF 01 01 00 D0 04 10 10 00 03 00}

Change Max 0x3 to 0x5

Yes, XTU recognizes the changes in nvram via ru.efi. If I don´t write anything to 204B and 2078 XTU has no X in circle behind 5-active cores and 6-active cores. If I hover over the circled X it gives the info "undo changes". This only happens if altering the values before over ru.efi.

In case of fit table it would make sense that bios is reading this and just adding the info. If I find the time maybe I try to use another bios without patched acpi tables and only 8 core support and take a look if this information is still there.

€dit: Good finding, I will try to change the value from 0x3 to 0x5 at adress 4D0 and will report what happened.