MSI MS-6712 KT4V-L v1.0 - need BIOS 1.B minus PXE bootrom

Questions that don't belong in the other forums.
Post Reply
Virtual Larry
New visitors - please read the rules.
Posts: 8
Joined: Thu Sep 05, 2002 6:17 am

Greetings smart people of WimsBIOS, I need your help. Thank you.

I recently discovered an upgraded version of the BIOS for my board, and flashed it. But it causes every HD except for my first to not be recognized as a BIOS physical disk. Instead of BIOS disks 0x80, 0x81, and 0x82, I only get 0x80. This is true for both BIOS versions 1.A and 1.B that I tested. BIOS versions starting with 1.9 included a PXE boot-rom for the on-board Via Rhine II ethernet, and as best I can figure out, is causing the problem.

My three HDs are all connected to a Promise Ultra100 TX2 PCI IDE controller, and up until this recent BIOS flash, have never caused an issue. I have four CD-ROMs on the motherboard IDE channels.

I have flashed back to BIOS 1.7 for now, but would like to use the "core" from the newest BIOS 1.B version, minus the PXE boot-rom. (Because it contains stability fixes for my ATI AGP 8x Radeon card and USB issues.) BIOS version is AMI, not Award. Most of the guides for "modbin" hacking are for Award. :(

More details, my thread on Anandtech: http://forums.anandtech.com/messageview ... id=1418656
My thread on StorageReview: http://forums.storagereview.net/index.p ... opic=16596
MSI BIOS download page with version 1.B: http://www.msi.com.tw/program/support/d ... 362&kind=1
Ritchie
BIOS Guru
Posts: 761
Joined: Wed Oct 30, 2002 5:17 am

I'm not sure if it is possible that a BIOS upgrade could cause identically jumpered drives to be recognised differently, but believe that it is possible although would have thought that scenario to be more likely on the mobo IDE controllers than on your add-on card controller. Regardless, my first thoughts are since I suspect something like that to be the case, to try different jumper setups on the drives to see if the card will then recognise the drives. I also assume you have two IDE ports on the card - if this is the case then depending on how you have your drives connected then maybe for example the card is only recognising the first channel and not both channels.

I don't think your on-board LAN in itself would be causing a problem but you should be able to disable that in the mobo BIOS setup to see. Rather, the configuration with the new BIOS may be causing a device to be assigned a resource or IRQ that may take precedence over the resource assignments to the IDE ports on the controller card. This could be particularly true if the old BIOS had been customised and you had only used the defaults in the new BIOS rather than customising it to suit your configuration. For example, automatic IRQ assignment should probably be enabled and I observe this as set to manual when loading defaults for many BIOSs.

See how you go with that. I don't suspect your new BIOS versions would be incompatible with the IDE controller card - I would be quite surprised if this could not be resolved as merely a configuration obstacle. Good luck!
Virtual Larry
New visitors - please read the rules.
Posts: 8
Joined: Thu Sep 05, 2002 6:17 am

Ritchie, I'm sorry, but I have to respectfully disagree with everything that you have posted. Did you read my other two background-story links? I go into more detail there.

The problem has nothing to do with hardware resource assignments or configuration, that I can tell. If it did, then it doesn't make sense that the first and second HDs (which are master and slave, on the Promise's primary IDE channel), would be recognized differently. If they are on the same channel, they should both be recognized the same. The hardware resouces are also the same as they ever were, both in the BIOS boot-up screen and in W2K, and in DOS-based diagnostic boot floppies like WD's DataLifeguard and Maxtor's PowerMax tools. However, normal DOS that uses the BIOS disk APIs to access disks, don't see any of the HDs other than the first.

Also, have you never heard of an additional add-on module for onboard hardware being embedded into a motherboard BIOS, causing conflicts between that and add-on PCI card BIOSes? I certainly have.

(Btw, I am an ex-professional DOS/x86 assembly-language programmer, I could probably debug/patch the BIOS itself if I need to, to fix whatever is the issue, but I need the proper tools to compress / split / recompress / fixup the checksum of the BIOS to work.)

So does anyone here know how to remove a module from AMI BIOSes?
ajzchips
El cheapo dude
Posts: 3048
Joined: Sat Mar 23, 2002 12:41 pm
Location: Madrid, Spain

You can try using AMIBCP. By the way, it's ModuleID 20 you'll want to remove...
Copyright (C) 1997,1998,1999 Intel Corporation..VIA Rhine II Fast Ethernet Adapter.......................... v2.17 (2002/11/04)...Intel Corporation.Intel UNDI, PXE-2.0 (build 082)
Virtual Larry
New visitors - please read the rules.
Posts: 8
Joined: Thu Sep 05, 2002 6:17 am

Thanks ajzchips, that was helpful.

Ritchie, I apologize, I was wrong. It isn't strictly because of the presence of the PXE boot-rom, because some form of it has been present in the BIOS since version 1.0. There are three versions, first one was in 1.0-1.1 and is 38,912 bytes, second one was in 1.2-1.4 and is 57,344 bytes, and third one was in 1.5-1.B and is 53,248 bytes. However MSI's BIOS release notes document a change made to enable booting from LAN BIOS in 1.9, and both 1.A and 1.B show a new "LAN BIOS: enable/disable" option under "integrated peripherals", which wasn't there in 1.7. So they did make some changes to network boot support in later BIOSes, only that change wasn't made in the PXE boot-rom itself, but somewhere else. I'm going to keep looking for it.

Btw, there is a module "0A Notebook", size 65536. I have a desktop motherboard. Why would that module be in my BIOS?
ajzchips
El cheapo dude
Posts: 3048
Joined: Sat Mar 23, 2002 12:41 pm
Location: Madrid, Spain

I really wouldn't know. Long shot: mobile processor support???
Virtual Larry
New visitors - please read the rules.
Posts: 8
Joined: Thu Sep 05, 2002 6:17 am

No, apparently the BIOS module named "Notebook" is just the power-management BIOS kernel.

Btw, if I do end up removing some modules - how are they called from the main BIOS portion? Do I need to patch out the calls too, or are BIOS modules simply not called into if they aren't present in AMI BIOS?

I noticed that for the PXE boot-rom, it is listed as a PCI option rom module-type, and the "load address" is not an address, but the 16:16 PCI device-ID code.
ajzchips
El cheapo dude
Posts: 3048
Joined: Sat Mar 23, 2002 12:41 pm
Location: Madrid, Spain

I believe their non-presence is simply ignored, although this doesn't always apply when dealing with embedded VGA ROM. I've had some boards fail to boot when removing the onboard VGA ROM, so somehow a call from the BIOS kernel is performed.
But this never applied when dealing with LAN BOOTROM in my previous experiments.
Post Reply