OpenBIOS February 2000

openbios@openbios.org

43 participants
83 discussions

RE: [OpenBIOS] l440gx+ nvram writing ...
by Wallace I. Kroeker 08 Feb '00

08 Feb '00

Ron, Have a look at this ( I had to do a search of all the intel site) but it does give a description of how their Flash works (flow charts etc). It also give references to source code but I'm not sure if this is freely available or not. Professional I/O Application Developer's Kit User's Manual August 1998 Wallace - To unsubscribe: send mail to majordomo(a)freiburg.linux.de with 'unsubscribe openbios' in the body of the message

2 1

Floppy driver
by Petazzoni Thomas 08 Feb '00

08 Feb '00

Hi there, does somebody can help me to debug my beginning of floppy driver ? i don't want to write all my questions if nobody's can help me ! thx -- PETAZZONI Thomas Web : http://kos.enix.org thomas.petazzoni(a)kos.enix.org ICQ : 34937744 - To unsubscribe: send mail to majordomo(a)freiburg.linux.de with 'unsubscribe openbios' in the body of the message

2 2

RE: [OpenBIOS] l440gx+ nvram writing ...
by Wallace I. Kroeker 08 Feb '00

08 Feb '00

Hello Ron, Here is my take after reading the documentation (the documentation seems to be identical for the 440bx). Intel (or some other BIOS maker) write a BIOS. This BIOS has the SMI handler in it. The SMI handler has a signature that it expects from the Flash software (written by the said BIOS maker). The signature is something that they won't want to divulge to keep people from inadvertently (maliciously) flashing BIOS. So how can you get around this? The following steps seem logical: 1. Write your own BIOS that has an SMI handler with your own signature. 2. Use the Flash update utility for your Board to flash your BIOS into place 3. Write your own Flash update utility which supplies your signature to your BIOS Or the following sequence of steps: 1. Write your own BIOS that has an SMI handler with your own signature. 2. Use LoadBios to put your BIOS in place (NOTE: you will be redefining the SMI handler at this point to be your own) 3. Proceed to write your BIOS to FLASH Now that sounds simple enough but I suspect the tricky part is the SMI handler. I really don't think the signature is that magical but rather a hand-shaking/security feature. Wallace > -----Original Message----- > From: Ronald G. Minnich [SMTP:rminnich@lanl.gov] > Sent: Monday, February 07, 2000 4:32 PM > To: openbios(a)elvis.informatik.uni-freiburg.de > Subject: Re: [OpenBIOS] l440gx+ nvram writing ... > > On Mon, 7 Feb 2000, Edwin Rhodes wrote: > > > Well if thats the case, lets hope to god that freedos will do the trick! > :) > > not that simple. You'll still have to solve the problem for freedos of > unlocking the key. > > Go to intel and get doc # 245141-001 of 3/1999. Check page 19, para 2.3.2 > We need to see how that is handled. I've got requests in to Intel and > VALinux now. We'll see how it goes. > > ron > > - > To unsubscribe: send mail to majordomo(a)freiburg.linux.de > with 'unsubscribe openbios' in the body of the message - To unsubscribe: send mail to majordomo(a)freiburg.linux.de with 'unsubscribe openbios' in the body of the message

2 1

RE: [OpenBIOS] Floppy driver
by Wallace I. Kroeker 08 Feb '00

08 Feb '00

I have been looking into this and would be interested in helping out. So you should post your questions. > -----Original Message----- > From: Petazzoni Thomas [SMTP:peta@club-internet.fr] > Sent: Tuesday, February 08, 2000 1:04 AM > To: openbios(a)elvis.informatik.uni-freiburg.de > Subject: [OpenBIOS] Floppy driver > > Hi there, > > does somebody can help me to debug my beginning of floppy driver ? > i don't want to write all my questions if nobody's can help me ! > > thx > -- > PETAZZONI Thomas Web : http://kos.enix.org > thomas.petazzoni(a)kos.enix.org ICQ : 34937744 > - > To unsubscribe: send mail to majordomo(a)freiburg.linux.de > with 'unsubscribe openbios' in the body of the message - To unsubscribe: send mail to majordomo(a)freiburg.linux.de with 'unsubscribe openbios' in the body of the message

1 0

l440gx+ nvram writing ...
by Ronald G. Minnich 08 Feb '00

08 Feb '00

more. You can enable/disable smi_l in the bios. I'm going to check that today. If smi_l is disabled, your OS will never know that the BUD asserted an interruptt when you tried to write the BIOS> The remaining question, of course, is what you're supposed to do with that interrupt, but that's next. it's rather amazing, but intel is shipping a server motherboard that REQUIRES DOS to be upgraded. You have to wonder sometimes, what are people thinking? The only exisiting SMI_L interrupt handler is in the BIOS! ron - To unsubscribe: send mail to majordomo(a)freiburg.linux.de with 'unsubscribe openbios' in the body of the message

3 4

RE: [OpenBIOS] l440gx+ nvram writing ...
by Wallace I. Kroeker 07 Feb '00

07 Feb '00

I found this in the P6 Hardware Developer's Manual A.1.49 SMI# (I) The SMI# (System Management Interrupt) signal is asserted asynchronously by system logic. On accepting a System Management Interrupt, processors save the current state and enter System Management Mode (SMM). An SMI Acknowledge transaction is issued, and the processor begins program execution from the SMM handler. I am guessing SMI_L is a system management Interrupt and the SMI Acknowledge is what you will be trying to do. More digging required. Wallace > -----Original Message----- > From: Ronald G. Minnich [SMTP:rminnich@lanl.gov] > Sent: Monday, February 07, 2000 11:40 AM > To: openbios(a)elvis.informatik.uni-freiburg.de > Subject: [OpenBIOS] l440gx+ nvram writing ... > > more. > > You can enable/disable smi_l in the bios. I'm going to check that today. > If smi_l is disabled, your OS will never know that the BUD asserted an > interruptt when you tried to write the BIOS> > > > The remaining question, of course, is what you're supposed to do with that > interrupt, but that's next. > > it's rather amazing, but intel is shipping a server motherboard that > REQUIRES DOS to be upgraded. You have to wonder sometimes, what are people > thinking? The only exisiting SMI_L interrupt handler is in the BIOS! > > ron > > > > - > To unsubscribe: send mail to majordomo(a)freiburg.linux.de > with 'unsubscribe openbios' in the body of the message - To unsubscribe: send mail to majordomo(a)freiburg.linux.de with 'unsubscribe openbios' in the body of the message

3 2

VGA card and screen output
by Wallace I. Kroeker 07 Feb '00

07 Feb '00

Hello, I FINALLY managed to get the screen output working. (A side benefit is that I now know a lot more about the 440BX chipset as I assumed this was not being initialized properly) The card that I have in my system is an ATI Rage III PCI. I finally added some code in START.ASM to initialize the VGA BIOS (from the reading I've done the VGA cards have their own Video BIOS). This required the following: * Initialization into 07h mode (monochrome 80 x 25 ... MDA mode) * Loading of a Character set (the cards come with a few predefined) * Activate the Character set that was loaded * Now the VGA base has to be 0xB0000 (MDA location actually) in Firmware/init/vga.h OR configure the video to be MDA The actual initialization was done using INT 10h in start.asm. If anyone has suggestions on other ways to do this let me know. I am currently working on getting the floppy enabled. (The idea being to have a minimum BIOS that can actually be put into the BIOS flash but still allow boot from disk for debugging or re-flashing) Has anyone already started this? Glad to hear Ron managed to get around his NVRAM write issue. Wallace - To unsubscribe: send mail to majordomo(a)freiburg.linux.de with 'unsubscribe openbios' in the body of the message

2 1

l440gx write to NVRAM: ah ha!
by Ronald G. Minnich 07 Feb '00

07 Feb '00

such fun. There is a PAL (or PLD) on the motherboard called a BUD. The bud is a chunk of glue that takes care of stuff not in the various chipsets. The BUD has two interrupt lines it controls, one called SMI_L. IF you write to the NVRAM, BUD yanks SMI_L and the interrupt handler (in this case in the BIOS) is tasked with responding with a magic word or having put a magic word at some magic place in memory. This is the 'key'. No key, no write. Tough for you. This little detail is only documented in the l440gx+ baseboard tps, #245151-001, version 1.0, march 1999, and not in any of the other five manuals I have on this chipset! Security through obscurity, I guess. ron - To unsubscribe: send mail to majordomo(a)freiburg.linux.de with 'unsubscribe openbios' in the body of the message

1 0

Flashing the BIOS ...
by Wallace I. Kroeker 01 Feb '00

01 Feb '00

Ron, As I have been reading the documentation on my chipset there is mention of disabling certain features when the BIOS is to perform different functions. As I was and checking out the BIOS update for my board (an ABIT BM6) and was reading through the instructions and they mention disabling the following: 1. CPU Internal Cache 2. CPU External Cache 3. Video BIOS Cacheable 4. Video BIOS Shadow prior to running the Award flash program. So you might want to have a check in your documentation for issues relating to these features and the update of your BIOS. If I come across the documentation comments again (they didn't seem relevant when I read them the first time :-) ) I will pass them along. Wallace - To unsubscribe: send mail to majordomo(a)freiburg.linux.de with 'unsubscribe openbios' in the body of the message

1 0

RE: [OpenBIOS] linuxbios web page is there.
by Wallace I. Kroeker 01 Feb '00

01 Feb '00

Hello Ron, I'm not sure if you have looked into this but I took a look at the NVRAM (read FLASH ) from the perspective of building the board. If you look at the Design Specifications (I downloaded both the 440gx and 440bx and they are identical on this topic) and look at how they recommend implementing the memory they describe how you should implement BIOS memory. For the Flash BIOS it all eventually comes down to hardware control of the memory chips so you can actually write to them. The recommended schematic shows the PIIX4E BIOSCS# signal going out the Flash CE#. So what the heck is the BIOSCS# and how do I deal with it? I followed this in the 440BX to "PCI to ISA/EIO Bridge Register Description document" http://developer.intel.com/design/intarch/techinfo/440bx/bridgreg.htm The "XBCS-X-Bus Chip Select Register (Function 0)" seems to be the thing you are looking for. I couldn't find similar documentation for the 440gx There are a couple other flags in this register that may pertain to this. If you can get hold of the manufacturers schematics you should be able to tell if they have some special hardware config. (as mentioned by someone earlier). Wallace > -----Original Message----- > From: Ronald G. Minnich [SMTP:rminnich@lanl.gov] > Sent: Tuesday, February 01, 2000 8:55 AM > To: openbios(a)elvis.informatik.uni-freiburg.de > Subject: Re: [OpenBIOS] linuxbios web page is there. > > On Tue, 1 Feb 2000, Daniele Frijia wrote: > > > > I figure that we're taking a slightly different direction than > openbios so > > > I've set up the linuxbios web page, www.acl.lanl.gov/linuxbios. We'll > use > > > openbios and help out where we can. I'll be posting the 'linux boots > > > linux' code there this week as well as anything we've done to openbios > to > > > make it go. > > > > We shoudl we work against each other? > > No, I have no intention of working against anyone. It seems that the focus > of this list is replacing the BIOS with a plug-compatible equivalent, or > doing Open Firmware instead. My focus is much simpler: I want to get just > enough OpenBIOS working to boot linux, and we're almost there. Supporting > FreeDOS and other 16-bit OSes is quite neat, but is not a concern of mine. > I have no need for INT 12 support, for example. > > So we'll help all we can with OpenBIOS, but past a point our needs and > interests will diverge. I certainly will let you all know if I ever figure > out how to write NVRAM -- if I can't do that, then I'm dead in the water. > > Thanks > > ron > > > > - > To unsubscribe: send mail to majordomo(a)freiburg.linux.de > with 'unsubscribe openbios' in the body of the message - To unsubscribe: send mail to majordomo(a)freiburg.linux.de with 'unsubscribe openbios' in the body of the message

1 0

← Newer
1
2
3
4
5
6
7
8
9
Older →