@kgbab
@Fernando

I have said many times that there are some difficulties with the addition of AMD. I have no motherboard is for testing. A motherboard is the risk strangers I will not.
I see, of course, once again, what can be done, but I will not promise anything. So please forgive me in advance.

Added:
At this point, I have an idea. Make 2 packages UBU. The first package for Intel, for example UBU_I and one for AMD, for example UBU_A.
And as I said, I have no motherboard is for testing. So I'm not sure compatibility OROM AMD, even within the same chipset with different revisions.
If someone is willing to risk his motherboard is, then send me the file BIOS that point I had started.

Zitat von Fernando im Beitrag #696

---

Unfortunately I don't have an AMD chipset mainboard myself. So I cannot do any tests regarding the AMD OROM/EFI modules.

---

@lordkag
How do you determine the version in files MISC_SIG? If you are not hard, please explain on an example of version 3.2.1540. Thank you.

Hello Sonix

Here is the latest bios from my motherboard.
Probably I will test the UBU Tool for the AMD and I will tell you the results.

Good luck to me and you...

@kgbab
Thank you for your feedback. As soon as I have prepared an option, I shall write to you immediately. :)

Hi soniX

I have the asrock 990FX Professional
The last original bios P2.00
The my mod bios for the comparation mdclab

good job

Ok. I will try this weekend to release the first beta of UBU supporting AMD SATA AHCI/RAID. :)

@SoniX

It is not possible to do that, since this is just a helper module. The only way is to pair it with the main AMD RAID OROM. But this is where the tricky part starts:

- usually BIOS has 1002-4391 (AHCI) + 1002-4392 (RAID) + 1002-4393 (RAID) | OR | 1022-7801 (AHCI) + 1022-7802 (RAID) + 1022-7803 (RAID). I have not seen mixed 43xx and 78xx so far.
- the rule is to have lower ID (4392 or 7802) in 9BD5C81D-...FC and higher ID (4393 or 7803) in 9BD5C81D-...FD.
- but some boards have them backwards, so extra check is needed.
- older 1002-4392 from 3.2.xxxx.xx series are uncompressed and have all this in hex: 81FA0210, 81430210, 91430210, 92430210, 93430210.
- all others are usually compressed so you need 7z.exe in UBU. They don't have the ID in hex.
- run "7z.exe e temp\miscFC.ffs -otemp misc.bin -r >nul && ren temp\misc.bin miscFC.bin" and "7z.exe e temp\miscFD.ffs -otemp misc.bin -r >nul && ren temp\misc.bin miscFD.bin"
- run "hexfind 64000057568B5E048B473A8946FC662BC0668946C4B910008D7EC81607F3AB66 temp\miscFD.bin >nul". If true, it is 4393 or 7803, otherwise check with miscFC.bin, otherwise dead end.
- I have only a few misc.bin from each ID, so this is only an assumption. Maybe kgbab can share other modules (even if not AMD), since he seems to have some high access.

So do something like this:

set miscFClow=0
set miscFDlow=0
hexfind 92430210 temp\miscFC.ffs >nul && hexfind 93430210 temp\miscFC.ffs >nul && set miscFClow=1
hexfind 92430210 temp\miscFD.ffs >nul && hexfind 93430210 temp\miscFD.ffs >nul && set miscFDlow=1
if %miscFClow%==1 if %miscFDlow%==0 ------> then FC is 4392/7802 + FD is 4393/7803 and go to ...
if %miscFClow%==0 if %miscFDlow%==1 ------> then FD is 4392/7802 + FC is 4393/7803 and go to ...
rem Both are either compressed or uncompressed
7z.exe e temp\miscFC.ffs -otemp misc.bin -r >nul && ren temp\misc.bin miscFC.bin
7z.exe e temp\miscFD.ffs -otemp misc.bin -r >nul && ren temp\misc.bin miscFD.bin
rem Search this string. It is only found in 4393/7803.
if exist temp\miscFC.bin hexfind 64000057568B5E048B473A8946FC662BC0668946C4B910008D7EC81607F3AB66 temp\miscFC.bin >nul && set miscFDlow=1
if exist temp\miscFD.bin hexfind 64000057568B5E048B473A8946FC662BC0668946C4B910008D7EC81607F3AB66 temp\miscFD.bin >nul && set miscFClow=1
if %miscFClow%==1 if %miscFDlow%==0 ------> then FC is 4392/7802 + FD is 4393/7803 and go to ...
if %miscFClow%==0 if %miscFDlow%==1 ------> then FD is 4392/7802 + FC is 4393/7803 and go to ... else Dead end.
 

It also needs to account for one or none being compressed, so more polishing is needed. But this AMD is going to be a beast, since it has 4-5 (!!!) AMD RAID EFI files, then there are 2 more for FM2+, then there are 2-4 OROM+MISC, then there are VBIOS files.

@lordkag
It turns out that there is a direct dependence on versions? That is, if for example 4002-4392 is version 3.0.0.000, then MISC (for 4392) will be the same version 3.0.0.0000?

Added:
I see that there is a compression LH5 (LHARC), I wonder whether all will understand the motherboard is compressed MISC? Or have to do even unpacked?

Yes, they are linked with the main module. The problem is to determine with which one. The rule 4392/7802 in ...FC and 4393/7803 in ...FD is not always followed. Some of them are bugged with reversed order.

MISC.BIN for 1002-4392 up until 3.2.xxxx.xx is uncompressed, only after 3.3.xxxx.xx is compressed. You should leave them as they are found, since they have a checksum check in byte 0x0A.

Zitat von lordkag im Beitrag #705

---

YSome of them are bugged with reversed order.

---

The first beta version with support AMD.
While only display versions for OROM 4391-4393 and EFI AMD RAID / Utility.
https://www.sendspace.com/file/ebf16p

I do not have a BIOS for 7801-7803. :(

Added:
@asrlab
@kgbab
@All
If there are problems in displaying versions then send screenshots.