[890FX Dlx3] HOI-HOI-HOI! Mmmmery christmas from the "mad scientist"!

**Ket** · 01-20-2011, 09:54 AM

Originally Posted by felix_w

If you have time, check this out....Raid Rom has already been updated after my communication with GB Tech....when i have time, i'll check again for modifying the values of CPU-NB Voltage

Thats award BIOS, the BIOSes on these Asrock boards are AMI, which is why we are having so much trouble replacing AGESA.CPU and the AGESA code itself because AMI build it into their main BIOS module instead of making the modules independent like Award do.

**Formula350** · 01-20-2011, 09:56 AM

Originally Posted by felix_w

Formula350, u can also check the link i attached to my previous link....my objective was quite similar to yours as far as i can see now...maybe we can exchange our information...

Well I don't have the 890GX board in use right now (was actually /trying/ to sell it), since I got this 890FX heh

All the info I have can be found in this single thread

Started out with modding it before I had the board even, but then I updated it once I managed to find it.
http://techpowerup.com/forums/showthread.php?t=116661

I've been through Rebel's Haven a number of times though, just not much recently is all. Couldn't find anything of help. The forums don't have new registration opened, so I emailed Polygon (I think that's his name), asking if he knew anything or could do a CRC check. He just said no and only Gigabyte could. So I left it at that >_>

I was successful in modding Award/Award-Phoenix BIOSes with MODBIN up to S939 though :\ If I manage to get this CPU I'm hoping to, I'll have enough spare parts to build the 890GX again, so I might try with MODBIN 2.04.03 that I came across recently (it's on RH actually I think).

**NEOAethyr** · 01-20-2011, 10:14 AM

Some newer award bios'es with ageis have staticly linked code.
Like my dfi board, which has at least 4 (can't remember off hand how many) static jumps or calls, or just mem ref's.
Whatever they are.
If you move a rom module just a byte, it'll offset these static addresses and ruin the whole bios.

I can mod my dfi, and I got a 2nd chip just for it.
But since I lost my x4 I don't really care about it so much anymore.
That and the implementation was dfi's fault, not all newer award bios'es are this hard to mod.

Plus award bios is about dead anyways.
So someday I gotta redo my compiler for ami bios'es.
But I won't mess around until I get a spare chip for my board.

When I was doing research on it I found some of the ami tools and was able to peek inside my bios.
Found some extra spread spectrum help msg's that I was kind of interested in (mentioned that with certain ones enabled together it would cause corruption).
But otherwise there was nothing that interesting in it to be unlocked.
That's when I read online somewhere that one of the the ami tools would corrupt the bios when saving it.

Award bios'es it isn't hard to add new settings and adding cmos parsing code into the bootblock is easy too, heck even adding debug msg's on the post screen is easy.
Finding a cmos location that isn't used as scratch ram during post is not so easy though lol.

I don't know squat about modding ami bios'es.

But anyways I just wanted to post so I could mention that award bios'es sometimes have static address that you need to patch during compile.

Corrupted bios menu's come from a bad ptr to MLSTRING (en_code) and _item.
Or sometimes just an offset ver of the above files (byte 0 offset ='s byte 1 type of thing).

Ageis is modable but everything must be written out correctly.
If _item or _en_code is messed up, it's ok because you can still post and re-flash.
Much easier then swapping bios chips by hand...

Has anyone seen any old ami src codes around the net?
I've got 2 rev's of award src but no ami src to work with.
Having the original src (even if it's 10 years old) helps a ton with mapping and decompiling post functions into asm src.

Oh and one more small ?.
Anyone know the compression format that ami uses?

**felix_w** · 01-20-2011, 10:24 AM

Originally Posted by NEOAethyr

Corrupted bios menu's come from a bad ptr to MLSTRING (en_code) and _item.
Or sometimes just an offset ver of the above files (byte 0 offset ='s byte 1 type of thing).

Ageis is modable but everything must be written out correctly.
If _item or _en_code is messed up, it's ok because you can still post and re-flash.
Much easier then swapping bios chips by hand...

Could that mean that if i swap _en_code from another sister board model there's probably no harm done , since i can re-flash with the unmodded bios ?

**Formula350** · 01-20-2011, 10:27 AM

Originally Posted by Ket

Thats award BIOS, the BIOSes on these Asrock boards are AMI, which is why we are having so much trouble replacing AGESA.CPU and the AGESA code itself because AMI build it into their main BIOS module instead of making the modules independent like Award do.

lol He knows that Ket

We were just [hijacking], talking about the Gigabyte 890GX, sorry

Originally Posted by NEOAethyr

Some newer award bios'es with ageis have staticly linked code.
Like my dfi board, which has at least 4 (can't remember off hand how many) static jumps or calls, or just mem ref's.
Whatever they are.
If you move a rom module just a byte, it'll offset these static addresses and ruin the whole bios.

<snip>
llllllllllllllooots of info
</snip>

Oh and one more small ?.
Anyone know the compression format that ami uses?

I found Borg's (Number One, the original) pack a few years back, and then nabbed Number Two's (different guy, just trying to carry on the 'legacy') with some updated stuff. I did peek inside a few AMI BIOSes, this 890FX included, but I've never been able to get AMIBCP to save my "changed" options

So, I just gave up. I bugged Ket enough back then (not that I still don't on other topics

) and figured it was basically pointless since I don't know HEX :P

**NEOAethyr** · 01-20-2011, 10:39 AM

As long as the checksum is ok and nothing else is wrong with the bios it will post.
But you won't have a working bios menu or bootup menu.

If you need to mod the menu's, you might have to learn how to doit by hand, and that does take a while to write the thing out into asm.
I was gonna try to make a decompiler just for this, because it's the most time consuming part about modding an award bios but I never got toit.

Award encode goes something like this:

Code:

;Item #: Starts with 0, not 1.
;Group #: 0xFF = 1, 0x00 = 2, 0x01 = 3, 0x02 = 4, 0x03 = 5, 0x04 = 6, 0x05 = 7, 0x06 = 8 (guessing, close enough)

db "_EN_CODE.BIN", 0x00, 0x00, 0x00, 0x00

Menu_Group_List:
dw (((Menu_Group_List_End - Menu_Group_List) / 2) - 1)
dw Menu_Group_0
dw Menu_Group_1
dw Menu_Group_2
dw Menu_Group_3
dw Menu_Group_4
dw Menu_Group_5
dw Menu_Group_6
dw Menu_Group_7
dw Menu_Group_8
dw Menu_Group_9
if Cleanup = 1
	dw Menu_Group_10
end if
Menu_Group_List_End:

........

if Cleanup = 1
	Menu_Group_10_Item_0:	;0x0A00
	db "[tCLK]: ", 0x00

	Menu_Group_10_Item_1:
	db "1X    ", 0x00
	db "1.125X", 0x00
	db "1.25X ", 0x00
	db "1.375X", 0x00
	db "1.5X  ", 0x00
	db "1.625X", 0x00
	db "1.75X ", 0x00
	db "1.875X", 0x00
	db "2X    ", 0x00
	db "2.5X  ", 0x00
	db "3X    ", 0x00
	db "3.5X  ", 0x00
	db "4X    ", 0x00

	Menu_Group_10_Item_2:
	db "BUI/SIP Strap: ", 0x00

	Menu_Group_10_Item_3:
	db "Auto             ", 0x00		;0x00
	db "100              ", 0x00		;0x01
	db "113              ", 0x00		;0x02
	db "125              ", 0x00		;0x03
	db "138              ", 0x00		;0x04
	db "150              ", 0x00		;0x05
	db "163              ", 0x00		;0x06
	db "175              ", 0x00		;0x07
	db "188              ", 0x00		;0x08
	db "200              ", 0x00		;0x09
	db "213              ", 0x00		;0x0A
	db "225              ", 0x00		;0x0B
	db "238              ", 0x00		;0x0C
	db "250              ", 0x00		;0x0D
	db "256              ", 0x00		;0x0E
	db "263              ", 0x00		;0x0F
	db "269              ", 0x00		;0x10
	db "275              ", 0x00		;0x11
	db "Hellfire 3EG-REV2", 0x00		;0x12
	db "tictac XT        ", 0x00		;0x13
	db "Test Strap       ", 0x00		;0x14

	Menu_Group_10_Item_4:
	db "Debug ROM: ", 0x00

	Menu_Group_10_Item_5:
	db "Disabled        ", 0x00
	db "Skip DMI Pooling", 0x00
	db "Debug           ", 0x00

	Menu_Group_10:
	dw (((Menu_Group_10_End - Menu_Group_10) / 2) - 1)
	dw Menu_Group_10_Item_0
	dw Menu_Group_10_Item_1
	dw Menu_Group_10_Item_2
	dw Menu_Group_10_Item_3
	dw Menu_Group_10_Item_4
	dw Menu_Group_10_Item_5
	Menu_Group_10_End:
end if

This is not the greatest example.
It's all about the offsets of each setting and there values.
And the length's of these tables and so on.
The very begining of the file lists each group offset.

If the bios gets this wrong then everything gets offset and everything appears corrupted in the bios menu and boot menu (sometimes f8 or hammering on the esc key during post).

If the menu and item file was made correctly, then it's possible a value in the uncompressed section of the bios is incorrect, or maybe it's just in the wrong location (modbin will compile it and it'll be wrong type of thing).

Here's an example of _item.bin

Code:

;;; Begining Of 2nd Segment

;0x09	should be the last group
;0x0A	offset or group (group if I defined it, this might be a prob...)
;0x00	group2, used as the dmi pool text

;0x0D	direct offset
;0x0E	direct offset
;0x0F	Direct offset
;0x10	direct offset
;0x11	direct offset
;0x12	direct offset
;0x13	direct offset

;db Item_2, Menu_Group_4	;ItemName

Item_Null	equ	0x0000

Menu_Group_0	equ	0x00
Menu_Group_1	equ	0x01
Menu_Group_2	equ	0x02
Menu_Group_3	equ	0x03
Menu_Group_4	equ	0x04
Menu_Group_5	equ	0x05
Menu_Group_6	equ	0x06
Menu_Group_7	equ	0x07
Menu_Group_8	equ	0x08
Menu_Group_9	equ	0x09
Menu_Group_10	equ	0x0A
Menu_Group_11	equ	0x0B
Menu_Group_12	equ	0x0C
Menu_Group_13	equ	0x0D
Menu_Group_14	equ	0x0E
Menu_Group_15	equ	0x0F
Menu_Group_16	equ	0x10
Menu_Group_17	equ	0x11
Menu_Group_18	equ	0x12
Menu_Group_19	equ	0x13
Menu_Group_20	equ	0x14
Menu_Group_21	equ	0x15
Menu_Group_22	equ	0x16
Menu_Group_23	equ	0x17
Menu_Group_24	equ	0x18
Menu_Group_25	equ	0x19
Menu_Group_26	equ	0x1A
Menu_Group_27	equ	0x1B
Menu_Group_28	equ	0x1C
Menu_Group_29	equ	0x1D
Menu_Group_30	equ	0x1E
Menu_Group_31	equ	0x1F

Item_0		equ	0x00
Item_1		equ	0x01
Item_2		equ	0x02
Item_3		equ	0x03
Item_4		equ	0x04
Item_5		equ	0x05
Item_6		equ	0x06
Item_7		equ	0x07
Item_8		equ	0x08
Item_9		equ	0x09
Item_10		equ	0x0A
Item_11		equ	0x0B
Item_12		equ	0x0C
Item_13		equ	0x0D
Item_14		equ	0x0E
Item_15		equ	0x0F
Item_16		equ	0x10
Item_17		equ	0x11
Item_18		equ	0x12
Item_19		equ	0x13
Item_20		equ	0x14
Item_21		equ	0x15
Item_22		equ	0x16
Item_23		equ	0x17
Item_24		equ	0x18
Item_25		equ	0x19
Item_26		equ	0x1A
Item_27		equ	0x1B
Item_28		equ	0x1C
Item_29		equ	0x1D
Item_30		equ	0x1E
Item_31		equ	0x1F
Item_32		equ	0x20
Item_33		equ	0x21
Item_34		equ	0x22
Item_35		equ	0x23
Item_36		equ	0x24
Item_37		equ	0x25
Item_38		equ	0x26
Item_39		equ	0x27
Item_40		equ	0x28
Item_41		equ	0x29
Item_42		equ	0x2A
Item_43		equ	0x2B
Item_44		equ	0x2C
Item_45		equ	0x2D
Item_46		equ	0x2E
Item_47		equ	0x2F
Item_48		equ	0x30
Item_49		equ	0x31
Item_50		equ	0x32
Item_51		equ	0x33
Item_52		equ	0x34
Item_53		equ	0x35
Item_54		equ	0x36
Item_55		equ	0x37
Item_56		equ	0x38
Item_57		equ	0x39
Item_58		equ	0x3A
Item_59		equ	0x3B
Item_60		equ	0x3C
Item_61		equ	0x3D
Item_62		equ	0x3E
Item_63		equ	0x3F
Item_64		equ	0x40
Item_65		equ	0x41
Item_66		equ	0x42
Item_67		equ	0x43
Item_68		equ	0x44
Item_69		equ	0x45
Item_70		equ	0x46
Item_71		equ	0x47
Item_72		equ	0x48
Item_73		equ	0x49
Item_74		equ	0x4A
Item_75		equ	0x4B
Item_76		equ	0x4C
Item_77		equ	0x4D
Item_78		equ	0x4E
Item_79		equ	0x4F
Item_80		equ	0x50
Item_81		equ	0x51
Item_82		equ	0x52
Item_83		equ	0x53
Item_84		equ	0x54
Item_85		equ	0x55
Item_86		equ	0x56
Item_87		equ	0x57
Item_88		equ	0x58
Item_89		equ	0x59
Item_90		equ	0x5A
Item_91		equ	0x5B
Item_92		equ	0x5C
Item_93		equ	0x5D
Item_94		equ	0x5E
Item_95		equ	0x5F
Item_96		equ	0x60
Item_97		equ	0x61
Item_98		equ	0x62
Item_99		equ	0x63
Item_100	equ	0x64
Item_101	equ	0x65
Item_102	equ	0x66
Item_103	equ	0x67
Item_104	equ	0x68
Item_105	equ	0x69
Item_106	equ	0x6A
Item_107	equ	0x6B
Item_108	equ	0x6C
Item_109	equ	0x6D
Item_110	equ	0x6E
Item_111	equ	0x6F
Item_112	equ	0x70
Item_113	equ	0x71
Item_114	equ	0x72
Item_115	equ	0x73
Item_116	equ	0x74
Item_117	equ	0x75
Item_118	equ	0x76
Item_119	equ	0x77
Item_120	equ	0x78
Item_121	equ	0x79
Item_122	equ	0x7A
Item_123	equ	0x7B
Item_124	equ	0x7C
Item_125	equ	0x7D
Item_126	equ	0x7E
Item_127	equ	0x7F

org 0x0000

include "Makefile"

........

db "$ml$"
dw 0x0000	;Kernel_Do_NO
dw 0x0000	;Kernal_Show_NO
db "$SpeCR"	;Kernal Sepcial Code or something

......

dw 0x0000				;ItemStat	;"Advanced BIOS Features"
db Item_0, Menu_Group_1			;ItemName
dw 0xF800				;CtReg
dw 0x0001				;CtRegMask
db 0xFD					;CmosLoc
dw 0x0001				;CmosMask
dw Item_Null				;ItemValName
dw 0x0000				;ItemMin
dw 0x0000				;ItemMax
db 0x48					;Item_Xaxis
db 0x18					;Item_Yaxis
dw 0x0000				;RomDefault
dw 0x0000				;SetupDef
db Item_20, Menu_Group_1		;HelpOffset

dw 0x0000				;ItemStat	;"Advanced Chipset Features"
db Item_1, Menu_Group_1			;ItemName
dw 0xF800				;CtReg
dw 0x0002				;CtRegMask
db 0xFD					;CmosLoc
dw 0x0001				;CmosMask
dw Item_Null				;ItemValName
dw 0x0000				;ItemMin
dw 0x0000				;ItemMax
db 0x48					;Item_Xaxis
db 0x20					;Item_Yaxis
dw 0x0000				;RomDefault
dw 0x0000				;SetupDef
db Item_21, Menu_Group_1		;HelpOffset

dw 0x0000				;ItemStat	;"Power Management Setup "
db Item_3, Menu_Group_1			;ItemName
dw 0xF800				;CtReg
dw 0x0004				;CtRegMask
db 0xFD					;CmosLoc
dw 0x0001				;CmosMask
dw Item_Null				;ItemValName
dw 0x0000				;ItemMin
dw 0x0000				;ItemMax
db 0x88					;Item_Xaxis
db 0x28					;Item_Yaxis
dw 0x0000				;RomDefault
dw 0x0000				;SetupDef
db Item_22, Menu_Group_1		;HelpOffset

dw 0x0000				;ItemStat	;"PnP/PCI Configurations"
db Item_2, Menu_Group_1			;ItemName
dw 0xF800				;CtReg
dw 0x0008				;CtRegMask
db 0xFD					;CmosLoc
dw 0x0001				;CmosMask
dw Item_Null				;ItemValName
dw 0x0000				;ItemMin
dw 0x0000				;ItemMax
db 0x88					;Item_Xaxis
db 0x30					;Item_Yaxis
dw 0x0000				;RomDefault
dw 0x0000				;SetupDef
db Item_23, Menu_Group_1		;HelpOffset

dw 0x0000				;ItemStat	;"Integrated Peripherals"
db Item_4, Menu_Group_1			;ItemName
dw 0xF800				;CtReg
dw 0x0010				;CtRegMask
db 0xFD					;CmosLoc
dw 0x0001				;CmosMask
dw Item_Null				;ItemValName
dw 0x0000				;ItemMin
dw 0x0000				;ItemMax
db 0x48					;Item_Xaxis
db 0x38					;Item_Yaxis
dw 0x0000				;RomDefault
dw 0x0000				;SetupDef
db Item_24, Menu_Group_1		;HelpOffset

dw 0x0000				;ItemStat	;"PC Health Status"
db Item_5, Menu_Group_1			;ItemName
dw 0xF801				;CtReg
dw 0x0080				;CtRegMask
db 0xFD					;CmosLoc
dw 0x0001				;CmosMask
dw Item_Null				;ItemValName
dw 0x0000				;ItemMin
dw 0x0000				;ItemMax
db 0x88					;Item_Xaxis
db 0x68					;Item_Yaxis
dw 0x0000				;RomDefault
dw 0x0000				;SetupDef
db Item_28, Menu_Group_1		;HelpOffset

dw 0x0000				;ItemStat	;"Genie BIOS Setting   "
db Item_6, Menu_Group_1			;ItemName
dw 0xF802				;CtReg
dw 0x0001				;CtRegMask
db 0xFD					;CmosLoc
dw 0x0001				;CmosMask
dw Item_Null				;ItemValName
dw 0x0000				;ItemMin
dw 0x0000				;ItemMax
db 0x88					;Item_Xaxis
db 0x08					;Item_Yaxis
dw 0x0000				;RomDefault
dw 0x0000				;SetupDef
db Item_29, Menu_Group_1		;HelpOffset


if Board_ID = "6A61BD49"
.....

That actually was the older ver of it before item existed.
But it was the most worked on ver I've done.
I've done real item.bin files, each setting thingy is the same layout.

The odd ptr's and stuff on top I don't even understand myself anymore since it's been so long.

Award bios is like, a dead art I guess you could say :|.

Edit:
Oh I remember, those were to tell each _item entry where the location of the setting or value was in the _en_code file...

**prznar1** · 01-20-2011, 10:52 AM

Originally Posted by Ket

^^ Only if the IMC is good enough. Theres probably also some PCB differences to help get those speeds as well such as better layout for the traces and additional phases for the DIMM slots.

maybe not. c3f has good memory overclocking abilities, you could gain lower timings on it compared to other mobos and better MHz overclocking overall but it still have that wierd wallblock. i think they could find the software problem there.

**Ket** · 01-20-2011, 10:58 AM

Originally Posted by Formula350

Well I don't have the 890GX board in use right now (was actually /trying/ to sell it), since I got this 890FX heh

All the info I have can be found in this single thread

Started out with modding it before I had the board even, but then I updated it once I managed to find it.
http://techpowerup.com/forums/showthread.php?t=116661

I've been through Rebel's Haven a number of times though, just not much recently is all. Couldn't find anything of help. The forums don't have new registration opened, so I emailed Polygon (I think that's his name), asking if he knew anything or could do a CRC check. He just said no and only Gigabyte could. So I left it at that >_>

I was successful in modding Award/Award-Phoenix BIOSes with MODBIN up to S939 though :\ If I manage to get this CPU I'm hoping to, I'll have enough spare parts to build the 890GX again, so I might try with MODBIN 2.04.03 that I came across recently (it's on RH actually I think).

I think polygon is trying to keep bios modding "exclusive". Years ago when RH was accepting new members I asked a question and polygon said it was impossible. I thought that was BS, and after a few days and teaching myself I accomplished what I was asking.

**Ket** · 01-20-2011, 11:08 AM

Originally Posted by prznar1

maybe not. c3f has good memory overclocking abilities, you could gain lower timings on it compared to other mobos and better MHz overclocking overall but it still have that wierd wallblock. i think they could find the software problem there.

A good memory table helps with IC compatibility and clocking certainly, I'm just saying its not totally reliant on the memory table. Such things as trace positioning on the PCB and amount of phases available to the DIMM slots can make quite a difference.

**prznar1** · 01-20-2011, 11:51 AM

but gtd pro got not more phases then c3f. traces sure, cos im not able to tell what is the diff in gtd vs c3f. im only suggesting that you should try to use those from gtd pro on deluxe

**Ket** · 01-20-2011, 12:02 PM

Thats the plan

The problem with it is all that info is crammed together in the 1B module, making it difficult.

**Formula350** · 01-20-2011, 07:47 PM

As a step towards checking the clock-wall, I busted out the Nanya sticks again (2x1GB 1066 C7 ECC) haha I decided to see what they'd handle for timings and started out trying them at what I was running my OCZ at (They're the old 1.9v Micro D9JLN 1600 C7s, but I was getting the same performance out of 1333 so I'm just running C6 @ 1.75v with tighter sub-timings). Well after a few restarts later, 1333 and all I had to do was change 6-6-6 to 7-7-7

Strangely they wouldn't work in the first 2 slots, but they did in the back 2 (channel B). THEN I figured the hell with it, and I crammed the 2 OCZ into the front dimms, with the Nanya still in haha Looks like C3 stepping did in fact allow for 4x1333 :P As a bonus, these ECC sticks came with on-die thermal sensors! Makes me want to get 4x2GB of Nanya ECC lol They run cooler than my OCZ, damn near as fast, and have temp readings. Suck on THAT Reaper HPC :\

Anyways, NEO, if that was intended for me then I'm sorry you wasted your time and I'd give you your 5minutes back if I could heh That was all greek to me

I'm the lab rat, not the scientist haha I only know my way around the maze from running it a few times and smelling the cheeze to guide me (which is akin to having a user friendly GUI to mod with [MODBIN]).

Ket (or anyone I guess), you have AMIBCP 3.46? I nabbed it earlier today and can send it. I mean, it's easy to find but I had to join a forum to download it, so I could save the time of having to do that.

**arroyo** · 01-21-2011, 02:11 AM

http://www.megaupload.com/?d=JTQHEAPQ
AMIBCP_V3.37.exe
MMTOOL_3.22_1B_21Fix-BKMOD.EXE

And this is AMIBCP 3.46:
http://www.megaupload.com/?d=6ZMYFC6L

Enjoy!

**Formula350** · 01-21-2011, 09:57 AM

Originally Posted by arroyo

http://www.megaupload.com/?d=JTQHEAPQ
AMIBCP_V3.37.exe
MMTOOL_3.22_1B_21Fix-BKMOD.EXE

And this is AMIBCP 3.46:
http://www.megaupload.com/?d=6ZMYFC6L

Enjoy!

hah :P

"The file you are trying to access is temporarily unavailable. Please try again later."

That was for the 3.37 and MMTool

You could just add them as attachments to the post... (I was only interested in that ver of MMTool, but I'm sure I can find it)

EDIT: Yea works here now too.

**Ket** · 01-21-2011, 11:24 AM

Works for me.

Thread: [890FX Dlx3] HOI-HOI-HOI! Mmmmery christmas from the "mad scientist"!

Thread Tools

Search Thread

Rate This Thread

Display

Bookmarks

Bookmarks

Posting Permissions