Bios Mod request : DFI Lanparty JR 790GX M2RS support for Phenom X6

[Gyxx]

Dear all,

I'm looking for to have Phenom X6 support for my DFI Lanparty Jr 790GX M2RS board (DDR2 AM2+ board). As you may know, DFI has stopped supporting these mobo, even if the power supply is well designed and the chipset supports new CPUs.

I'm actually running succesfully the latest BIOS available modified with SLIC 2.1 addition on a 940 BE CPU, of course made here at Biosmods .

I first searched for a ready to use bios already modfied for this board, but was not successful (apparently nobody made it, or it is no more on line)..

Then I thought to try with a modify by my own, following the suggestions and guide from this site, like I did already in the past to upgrade the RAID module of my old ABIT NF7s, since there's also the help of a brand new USB SPI capable programmer is on its way via express parcel service .....

But looking around i found this small 3d here from another user that apparently had problem doing it ....

http://www.bios-mods.com/forum/Thread-Ho...ESA-module

So I decided to ask help in order to understand what i was going to do and learn something .....

Here some download links (I higlighted the most important one):

Discussion here with the latest BIOS form my mobo improved with SLIC2.1 addiction.
http://www.bios-mods.com/forum/Thread-DF...lic2-1-req

Link to download same version of original bios (wituot SLIC ) from DFI :
http://img.dfi.com/Upload/BIOS/CM/R7JDA12.zip

Link to download a DONOR BETA BIOS from same board but with DDR3 instead of DDR2, still from DFI :
http://img.dfi.com/Upload/BIOS/CM/D7JD707.zip

Thanks to anybody that will give me help or prepare a ready to use BIOS for testing !

Happy epiphany to all

Gyxx

[Gyxx]

Dear all,

The USP Bios chip programmer is still on its way to arrive me, so I started to learn everything that I could on modding and/or upgrading CPU support of a mobo.

In the past I already upgraded the RAID or the bitmap logo of my older boards without problem, so I'm not completely novice to it, but I would like to know if what I'm trying to do has sense or not.

In a few words, I've understood that what I do basically need to upgrade the bios for Phenom 6 support is

1) the updated CPU microcode from AMD, either from a working BIOS wih full support to Ph x6 or just the one i can download from this discussion here :
http://www.bios-mods.com/forum/Thread-Ho...-Microcode

2) the AGESA.rom module (AMD Generic Encapsulated Software Architecture) from a donor bios, taken from a MOBO with Ph x6 support that has the same chipset, possibly also from the same producer. This is the module that suports all multicore functions and ( i do think) also memory controller intialization.

Taking these two things as true, i've made the following :

A) I've found both of these elements from another DFI MOBO identical to mine, except from the fact that it utilize DDR3 instead of DDR2. Since CPU microcode from the cited discussion was more recent and bigger (30kb Vs 28kb) than the one from the donor bios and the one form the base bios (24kb), I've chosen to include it in a test BIOS with the AGESA module taken from the DFI donor BIOS (rev. 3.7.0.1, 37 kb). AGESA version from base BIOS was 3.1.7.0. However, I'm afraid that the different DDR modules utilized by the two different boards (DDR£ Vs DDR2) can create problem for the AGESA part.

B) I've searched for at least another producer of Mobo that utilized the same AMD 790 GX chipset with DDR2 modules and that did incude Phenom X6 support to that mobo. I've found two different producers doing this, Jetaway and Sapphire, disassembled the modules and found with an Hex editor that they both do share the same AGESA module, about same size (35kb) of the one in the original BIOS, both revision 3.1.9.0. Moreover, one of these boards has also the debug port 80 on bord like mine, so It is really very similar. This AGESA version is lower than the one from A), but is higher than the one in my original BIOS and should be enought.
Also the CPU microcode was the same (26kb), bigger of the size of base bios (24kb only), so I prepared a modified BIOS for my board with both that AGESA and CPU microcode. This for my knowdlege is the best candiate.

C) I've found another board from Jetaway that shares Phenom II x6 support, 790GX chipset support, DDR2 support, but does utilize Laptop SO-DIMM DDR2 support instead of full lenght DDR2. The AGESA module of this board is slightly different from the others, revison is 3.3.2.3., but exactly same size. I had not made any BIOS mod utilizing this.

I don't know f these modifications were enought for a first try, or If I should replace also some other module in order to have a chance to succeed. I don't know either if i performed all the operation in a good way, even if what I did see from the hex editor confirms it. But i'm quite Noob to these things , any suggestion will be welcomed.

I've attached mod Bioses from both A) and B) option to this message, hoping I'll have from you all a confirmation that I had operated in a good way. Both of them has the working SLIC modified BIOS as base. Both of them were made utilizing CBROM V. 195 in a prompt command window under 7x64.

Here below also the link to download the other donor BIOS I've found

A)
- Original latest BIOS from DFI JR 790GX M2RS (DDR2) whitout SLIC mod http://img.dfi.com/Upload/BIOS/CM/R7JDA12.zip
- Same original bios DFI JR 790GX M2RS (DDR2) but with SLIC mod made here at Biosmods : Discussion : http://www.bios-mods.com/forum/Thread-DF...lic2-1-req
Direct SLIC BIOS download : http://dl.dropbox.com/u/747181/DFI%20JR%...34s282.zip
- AGESA Donor BIOS from DFI JR 790GX M3RS (DDR3) - latest BETA bios from DFI for these boards. http://img.dfi.com/Upload/BIOS/CM/D7JD707.zip
- Discussion where I took updated AMD microcode http://www.bios-mods.com/forum/Thread-Ho...-Microcode

B)
- Sapphire PC-AM2RS790G - PURE CrossFireX 790GX http://www.sapphiretech.com/presentation/product/?cid=2&gid=5&sgid=341&pid=240&psn=&lid=4&leg=0#
Direct BIOS wh. Ph. x6 support download : http://eu2.sapphiretech.com/archive/gm/d...540000.zip

- Jetaway HA07-GT/HA07-GTD http://www.jetway.com.tw/jw/motherboard_...ductid=576&proname=HA07-GT/HA07-GTD
Direct BIOS wh. Ph. x6 support download : http://www.jetway.org.uk/downloads/bios/HA7GAM05.zip

C)
- Jetaway HZ01 http://www.jetway.com.tw/jw/motherboard_...ductid=602&proname=HZ01
Direct BIOS wh. Ph. x6 support download : http://www.jetway.org.uk/downloads/bios/HZ01AM05.zip

Thanks to all in advance , if possible I'd like to learn how to do that, if not possible I'd like to learn why too !

Gyxx

[attachment=3324]
[attachment=3325]

[Gyxx]

Dear all.

While waiting for new Phenom x6 1100 BE used chip, I didn't resist to the temptation of Bios Modding, so first I've built mysel a simple serial to SPI programmer, then made some experiments, then the more performant USB programmer has arrived.

Having no fear of bricking the motherboards and some different environemt in my house ( I've two identical DFI with quite different settings , one as main PC with RAID disk and a 940 BE x4 phenom, another as mediacenter with Native IDE disks and only an athlon X3 435 2.9ghz CPU) I was able to discover some facts that were NOT reported in other thread I've seen around the web, and that maybe can help all of us in finding the solution to upgrade boards to newer CPU support / better understand this generation of BIOSes :

0) Upgrading NCPUCODE and adding different iSLIC at end of modules were only safe mods i could perform !
This is a fact, and CBROM 195 is the only one working under dos propmpt of 7x64.
Other did collapse even ordering them /d. However, using latest CBROM 198 in pure dos environment seems to produce no different results from 195 (and i could not find a changelog).

1) Order IS important !
Bricking was mathematecally sure if module order was not respected, meaning to substitute AGESA I've to extract & release first all the following modules, then sustitute AGESA, then rebuilt in the same order the BIOS, else (same modules, different orders). no chance to enter DOS mode. This did not happen on old Athlon motherboards, and I was quite surprised !

2) Better missing than misplaced !
If i put the AGESA as last module (no matter if followed or not by the three memory initialization commands MEMINIT, HT.DLL and HT32GATE.BIN) I'll brick the system, if i just release the AGESA module I'll have a booting system to DOS or Bart Pe environment, only problem was the lack of processor string recognition at boot and processor at fixed 3Ghz speed (By luck i do have a Phenom 940 BE chip, so maybe the setted informations iN BIOS were enought). Of course 7x64 stops booting after animated win flag composes from 4 lights, keeping pulse effect forever until reset.

3) You can rename modules, it doesen't matter (at least for video), but trying to subtitute video BIOS from another board of same chipset can prevent Windows boot or block Cool&Quiet, depending on SATA settings a thing that was quite unexpected from me.
In few words, after seeing that no memory seting was included in video BIOS with a RADEON bios editor, I tried to move upgraded BIOS from identical DDR3 DFI (same manufature) board to mine, changing module name in accord to the one present in the original BIOS or mantaining the same odf donor board.
In both cases I could enter DOS and Bart Pe, but no windows boot like in point 2) (Raid only board test)

Then tried to get video BIOS from Jetway/Sapphire (By Jetway) boards : they all (ddr2, DDr3) shared same module, it worked perfectly in RAID environment, even if completelyt blocking cool & quiet, but prevent proper booting of Windows in Native Ide environment of my mediacenter (but ok DOS and Bart Pe disk).

4) New AGESA allowed DOS and BArt Pe boot, but not windows boot (7x64)
Adding bigger new AGESA from a same producer DD3 board or similar size AGESA from a similar DDR2 board (both supporting Ph6) produced same effects of boot block on 7x64 as removing AGESA in point 2) (except for CPU string recognition) or putting wrong video BIOS bios, but no board brick.

5) Copying AGESA and HT.DLL from DDR2 board in my BIOS produced same poor results as point 4
No need to describe anymore .

6) Copying AGESA , MEMINIT, HT.DLL and HT32GATE.BIN from identical Jetway /sapphire DDR2 board bricked the MOBO Debug port 80 stopped with C1 code, meaning block at memory detection step in board boot sequence.

7) Changing or removing AHCI module prevented Windos boot and disks detection, not RAID disks detection , but i was still able to load USB boot Key equipped with awdflash and re-flash good bios.
This happened just changig the bootrom version by one step, including modules form both DFI DDR3 board and from JETWAY / Sapphire mobos (that producer utilized same modules for each bios of any mobo, I can presume hacking would be very easy in that case)

8) Removing SB750/RAID modules prevented loading of arrays but allowed windows boot in Native EIDE env. , as expected.

9) Changing SB750/RAID modules in a RAID environment blocked Raid routine after RAID screen display (no dots were produced , no array was searched for, complete block in this step.)
CMOS resetting and disabling module from BIOS however allowed me to usb Key but i was still able to load USB boot Key equipped with awdflash and re-flash good bios.

10) Changing AHCI and RAID modules blocked both Raid routine after RAID screen display (no dots were produced , no array was searched for, complete block in this step.) or any disk boot if RAID disabled.
CMOS resetting and disabling RAID module from BIOS however allowed me to usb Key but i was still able to load USB boot Key equipped with awdflash and re-flash good bios.

CONCLUSIONS

From what I've seen it seems that probably updated modules were not easily recognized (not really modular BIOS was written). And were REALLY interdependent. I've searched to understand what was the module header to try to rename it to what is expected to be, but at the moment I'nm not so skilled to do it.

But i can play as guinea pig for anybody or try to learn / understand from suggestions

Just to complete explanations, I did these experiments with the OLD cpu were supposed to work good on the boards, tomorrow new Ph6 1100 BE will arrive but i don' think it will work if older don't.

All these experiments were made flashing only main bios block (from DOS), than clearing cmos, while briking experiments induced a full SPI chip reprogramming via parallel port utilizing DOS environment from same USB key and SPIPGM2 with the undocumented command /s (my common SST 25V080B chip is not working with classical /p on SPIPGM2 or with classic SPIPGM program).

For all these reasons i did remove all the previous attachments, I don't want anybody can hurt its own system !

In case anybody would be interested in discovering how to setup a cheap and easy SPI programmer from a LPT port (not USB to LPT adaptor, at least a PCI card) i can post some links and explanations, all the explanations found around have at least on big error and were not easy to follow.

I attach to this mail for testing only working BIOS i produced, made simply upgrading CPU microcode to SLIC version, that works perfectly on both my test systems.

Of course DO USE IT AT YOUR OWN RISK and only if you do know what you are doing, I and nooney from this forum can be responsable for any damage it can cause to your system.

best wishes to everybody who reads, hope to get soon at least one message of interest by someone here.

Gyxx

P.S. here below structure of attached R7JDA14.BIN

Code:

cbrom V1.95 [12/14/07] (C)Phoenix Technologies 2001-2007

******** r7jda14.bin BIOS component ********

No. Item-Name Original-Size Compressed-Size Original-File-Name
================================================================================
0. System BIOS 20000h(128.00K) 13130h(76.30K) r7jda12.BIN
1. XGROUP CODE 0CB50h(50.83K) 08E78h(35.62K) awardext.rom
2. ACPI table 09506h(37.26K) 04266h(16.60K) ACPITBL.BIN
3. YGROUP ROM 0C6D0h(49.70K) 06364h(24.85K) awardeyt.rom
4. GROUP ROM[ 0] 07FC0h(31.94K) 035B5h(13.43K) _EN_CODE.BIN
5. SETUP0 032D0h(12.70K) 01368h(4.85K) _ITEM.BIN
6. BIOSF0 02490h(9.14K) 015BEh(5.44K) _DMI.BIN
7. FLSH CODE 02640h(9.56K) 00ED1h(3.70K) SPIFLASH.BIN
8. SMI32 002E0h(0.72K) 00246h(0.57K) SMI32COD.BIN
9. SMIAP 087E0h(33.97K) 00453h(1.08K) SMIAPCOD.BIN
10. VGA ROM[1] 0EA00h(58.50K) 0951Dh(37.28K) 780D.rom
11. PCI ROM[A] 0EC00h(59.00K) 093BEh(36.94K) SB750.BIN
12. PCI ROM[B] 06C00h(27.00K) 0665Fh(25.59K) AHCI.BIN
13. GV3 08978h(34.37K) 02DA3h(11.41K) AGESACPU.ROM
14. MINIT 0F08Fh(60.14K) 0F0BCh(60.18K) MEMINIT.BIN
15. HTINIT 05021h(20.03K) 05052h(20.08K) HT.DLL
16. 2 PE32 in MB 004D8h(1.21K) 00537h(1.30K) HT32GATE.BIN
17. LOGO BitMap 4B30Ch(300.76K) 1A477h(105.12K) 790JR.bmp
18. NoCompress ROM 00176h(0.37K) 001A7h(0.41K) iSLIC.BIN
(SP) NCPUCODE 07800h(30.00K) 07800h(30.00K) NCPUCODE.BIN
(SP) HOLE0 BC00h(47.00K) BC00h(47.00K) misc.BIN
(SP) HOLE1 7C08h(31.01K) 7C08h(31.01K) CIMRS780.B2
(SP) HOLE2 30B8h(12.18K) 30B8h(12.18K) CIMSB700.B2
(SP) HOLE3 FFFFh(64.00K) FFFFh(64.00K) ODMEC8.2.BIN

Total hole area space = 30000h(192.00K)
Total compress code space = A5000h(660.00K)
Total compressed code size = 7FB17h(510.77K)
Remain compress code space = 25509h(149.26K)

** Micro Code Information **
Bridge ID | Bridge ID
North South Patch ID CPUID | North South Patch ID CPUID
---------------------------------------+----------------------------------------
00000000 00000000 00000062 0414 | 00000000 00000000 00000068 040A
00000000 00000000 0000006D 041B | 00000000 00000000 00000083 0680
00000000 00000000 0000007A 0413 | 00000000 00000000 0000007B 0433
00000000 00000000 01000002 0000 | 00000000 00000000 01000020 1000
00000000 00000000 01000085 1040 | 00000000 00000000 01000086 1041
00000000 00000000 01000095 1022 | 00000000 00000000 01000096 1020
00000000 00000000 010000B6 1043 | 00000000 00000000 010000B7 1062
00000000 00000000 010000BF 10A0 |

[toettoetdaan]

Very interesting posts, I have also a DFI JR 790GX M2RS but my PII x4 940 died for an unknow reason
Now it runs on a Athlon x2 4800+ (@3Ghz oc ofc) but i have also a PIIx4 960T here. (cut down x6 1095T)

I no bios mod experiense, na a bit for GPU's and flashing a bios is also not new for me.
I was maken my own SPI flasher for this mobo some time ago, but then desided it was better to send is RMA.
(sinds i use te flash microcontrolers i am not so afraid for it)
Back to the point, if i am correct, you think that the header information from the module's isnt correct?

And i may have some other interesting point, the RMA bios i have is "(r7jda12) EVALUATION ROM - NOT FOR SALE"
It is the same nuber and date as the newest BIOS, but there might be a difference.

If you are interested to take a look: http://dl.dropbox.com/u/4896683/R7JDA12%28evu%29.BIN
(from the first look, it has 34 less chars and 17lines more)

EDIT:
I try to update the AGESA code in the rom, yust to see how it worked, interesting point is that if jou open the modded rom in a editor, it still make a couple of references to the old AGESA version number, i am acutaly very interseted in what happens if you edit those by hand to the new version number.

[toettoetdaan]

I did some playing whit CBROM, and did find some points i would like to disscuse.

1. The shapire and jetaway BIOS had both AGESA v3.1.x.x while the DFI BIOS is already 3.3.2.2 and x6 support is from 3.7.0.0.

2. AGESACPU.ROM and MEMINIT.BIN should always be added togetter, sindse MEMINIT.BIN refers to the AGESA version in AGESACPU.ROM.
I am not sure what to think about, HT.DLL and HT32GATE.BIN, HT.DLL seems universal and should be updatet whith the AGESACPU and MEMINIT.BIN, i think.
But HD32GATE.BIN seems to be board specific....

3. It is told that CBROM 1.95 meeses up MEMINIT.BIN, so use 1.98

4. I have found a better BIOS witch could help this project a bit more:
It is the Gigabite GA-MA790GP-UD4H 790GX + SB750 AM2 + DDR2
AGESA v3.7.1.0 and official X6 support. (on latest beta bios)
http://www.gigabyte.co.nl/products/produ...x?pid=3003#
BIOS F7C: http://www.gigabyte.co.nl/products/produ...=3003#bios

Some input would realy be appreciated

[dragoncrystals7]

[quote='Gyxx' pid='41125' dateline='1326921933']
Dear all.

in replying to your post I have a JR 790GX-M2RS and currently running a 965 quad core at 3.6 ghz , with 4 gb of 1066 corsair memory, and having trouble finding the right ram to increase it to 8gb, as I am new to posting bear with me if I don't post this right,
I have no knowledge of bios coding or anything like that but can build computers, I was told the 965's wouldn't work and have been running this for about 6 years or so, as to flashing a bios , I had 3 fails before I finally flashed a bios,
would your bios update by any chance have a ddr3 update too? was hoping to upgrade to fx 4-6-8 core cpu. any help there? thanks