[LinuxBIOS] Intel 440bx

[Shaddam Corrino IV]

Hello folks,

Ideally in few weeks I would like to use 440bx.
It is my impression that it is not working right now quite right.
Am I screwed?

What I would need is LB+VGA bios (so that I can use working vesafb in
linux).

# lspci
00:00.0 Host bridge: Intel Corporation 440BX/ZX/DX - 82443BX/ZX/DX Host
bridge (rev 02)
00:01.0 PCI bridge: Intel Corporation 440BX/ZX/DX - 82443BX/ZX/DX AGP bridge
(rev 02)
00:07.0 ISA bridge: Intel Corporation 82371AB/EB/MB PIIX4 ISA (rev 02)
00:07.1 IDE interface: Intel Corporation 82371AB/EB/MB PIIX4 IDE (rev 01)
00:07.2 USB Controller: Intel Corporation 82371AB/EB/MB PIIX4 USB (rev 01)
00:07.3 Bridge: Intel Corporation 82371AB/EB/MB PIIX4 ACPI (rev 02)
00:0a.0 Multimedia audio controller: Ensoniq ES1370 [AudioPCI] (rev 01)
00:0c.0 Ethernet controller: Lite-On Communications Inc LNE100TX (rev 20)
01:00.0 VGA compatible controller: S3 Inc. ViRGE/GX2 (rev 06)

Thanks,
Shaddam

[Corey Osgood]

Shaddam Corrino IV wrote:
> Hello folks,
>
> Ideally in few weeks I would like to use 440bx.
> It is my impression that it is not working right now quite right.
> Am I screwed?
>
> What I would need is LB+VGA bios (so that I can use working vesafb in
> linux).
>
> # lspci
> 00:00.0 Host bridge: Intel Corporation 440BX/ZX/DX - 82443BX/ZX/DX
> Host bridge (rev 02)
> 00:01.0 PCI bridge: Intel Corporation 440BX/ZX/DX - 82443BX/ZX/DX AGP
> bridge (rev 02)
> 00:07.0 ISA bridge: Intel Corporation 82371AB/EB/MB PIIX4 ISA (rev 02)
> 00:07.1 IDE interface: Intel Corporation 82371AB/EB/MB PIIX4 IDE (rev 01)
> 00:07.2 USB Controller: Intel Corporation 82371AB/EB/MB PIIX4 USB (rev 01)
> 00:07.3 Bridge: Intel Corporation 82371AB/EB/MB PIIX4 ACPI (rev 02)
> 00:0a.0 Multimedia audio controller: Ensoniq ES1370 [AudioPCI] (rev 01)
> 00:0c.0 Ethernet controller: Lite-On Communications Inc LNE100TX (rev 20)
> 01:00.0 VGA compatible controller: S3 Inc. ViRGE/GX2 (rev 06)
>
> Thanks,
> Shaddam

One of the developers is working on it now, we should have some code to
play with soon. In the meantime, what's the SuperI/O on your motherboard?

-Corey

[Shaddam Corrino IV]

On 11/12/06, Corey Osgood  wrote:
> Shaddam Corrino IV wrote:
> > 00:00.0 Host bridge: Intel Corporation 440BX/ZX/DX - 82443BX/ZX/DX
> > 00:07.0 ISA bridge: Intel Corporation 82371AB/EB/MB PIIX4 ISA (rev 02)
> > 01:00.0 VGA compatible controller: S3 Inc. ViRGE/GX2 (rev 06)

One of the developers is working on it now, we should have some code to
> play with soon. In the meantime, what's the SuperI/O on your motherboard?

Heh, thought it would be part of the NorthBridge. It is : Winbond
W83977TF-AW.

At the time I also do not have any extra EEPROMs so I'll have to pick
something and buy. Preferably cheap.

http://www.baber.com/baber/411/atrendatc6220.htm
http://www.baber.com/baber/gifs/411gifs/atc6220large.jpg
http://www.watch.impress.co.jp/Akiba/hotline/980418/image/atc1.jpg

[Peter Stuge]

On Sun, Nov 12, 2006 at 07:56:45PM +0100, Shaddam Corrino IV wrote:
> At the time I also do not have any extra EEPROMs so I'll have to pick
> something and buy. Preferably cheap.
>
> http://www.baber.com/baber/411/atrendatc6220.htm
> http://www.baber.com/baber/gifs/411gifs/atc6220large.jpg
> http://www.watch.impress.co.jp/Akiba/hotline/980418/image/atc1.jpg

That's a DIP-32. Could you peel of the shiny sticker and send the
part number, please?

Or find the datasheet yourself, of course. The important parameters
are voltage and size.

I'm not sure there are any non-5V DIP flashes but please check.
Size is almost certainly 128kb or 256kb on the board.

Older DIP flashes can have funky pin configuration depending on size
so you'll want to check the data sheet of your part. If you're not
sure how just send us the part number and we'll try to help.
Also, do you know any local electronic component stores that have
online inventories? Then we could try to find an equivalent for you.

//Peter

[Shaddam Corrino IV]

[please cc]

*Peter Stuge* wrote:
> On Sun, Nov 12, 2006 at 07:56:45PM +0100, Shaddam Corrino IV wrote:
> > At the time I also do not have any extra EEPROMs so I'll have to pick
>
> *>* something and buy. Preferably cheap.
> *>*
> *>* http://www.baber.com/baber/411/atrendatc6220.htm
> *>* http://www.baber.com/baber/gifs/411gifs/atc6220large.jpg
> *>* http://www.watch.impress.co.jp/Akiba/hotline/980418/image/atc1.jpg
>
> That's a DIP-32. Could you peel of the shiny sticker and send the
> part number, please?

Winbond W29C020-90

Or find the datasheet yourself, of course. The important parameters
> are voltage and size.

I found www.romstore.ru/pdf/*W29c020*.pdf
btw: does any kind soul has data sheet for s3 86C357 (Virge GX2)

I'm not sure there are any non-5V DIP flashes but please check.
> Size is almost certainly 128kb or 256kb on the board.

It seems it is DIP-32 5V 256KiB.
> Older DIP flashes can have funky pin configuration depending on size
> so you'll want to check the data sheet of your part. If you're not
> sure how just send us the part number and we'll try to help.

Is this one of those funky models?
> Also, do you know any local electronic component stores that have
> online inventories? Then we could try to find an equivalent for you.

My idea was to either take (a) the specs of the original chip or (b) list of
all equivalent (LB approved) models and visit all nearby electronics stores.
Not sure which idea is better.

[Peter Stuge]

On Mon, Nov 13, 2006 at 05:06:20AM +0100, Shaddam Corrino IV wrote:
> > That's a DIP-32. Could you peel of the shiny sticker and send the
> > part number, please?
>
> Winbond W29C020-90
> > I'm not sure there are any non-5V DIP flashes but please check.
> > Size is almost certainly 128kb or 256kb on the board.
>
> It seems it is DIP-32 5V 256KiB.

Yep, certainly is.
> > Older DIP flashes can have funky pin configuration depending on
> > size so you'll want to check the data sheet of your part. If
> > you're not ure how just send us the part number and we'll try to
> > help.
>
> Is this one of those funky models?

Nope, this has the standard pinout.
> > Also, do you know any local electronic component stores that have
> > online inventories? Then we could try to find an equivalent for you.
>
> My idea was to either take (a) the specs of the original chip or (b)
> list of all equivalent (LB approved) models and visit all nearby
> electronics stores.
> Not sure which idea is better.

Really any 5V 32-pin DIP 256 (probably also 512) kb flash with the
same pinout will be compatible. Lots of different chips.

The list of "LB approved" chips would be the list of supported chips
in the utility flashrom, but that just depends on what has been
available to developers since adding a new chip is quite simple with
few if any exceptions.

Here are the candidates: Am29F040B, At29C040A, Mx29f002,
SST29EE020A, SST28SF040A, SST39SF020A, SST39SF040, W29C020C,
M29F040B, M29F400BT from AMD, Atmel, Macronix, SST, Winbond and ST.

//Peter

[Shaddam Corrino IV]

On 11/13/06, Peter Stuge  wrote:
> > My idea was to either take [...] visit all nearby electronics stores.
>
> Really any 5V 32-pin DIP 256 (probably also 512) kb flash with the
> same pinout will be compatible. Lots of different chips.

I was looking around today and all that I could find was "certified
pre-owned" 256KiB flash ROMs for around $6 each.

Since as I understand 440BX is not ready for testing, I guess I can take
time looking around for better deal.
 I could check out two bazaars this weekend.

[Corey]

On Mon, 2006-11-13 at 17:45 +0100, Shaddam Corrino IV wrote:
> On 11/13/06, Peter Stuge  wrote:
>         > My idea was to either take [...] visit all nearby
>         electronics stores.
>     Really any 5V 32-pin DIP 256 (probably also 512) kb flash with
>     the
>     same pinout will be compatible. Lots of different chips.
>
> I was looking around today and all that I could find was "certified
> pre-owned" 256KiB flash ROMs for around $6 each.
>
> Since as I understand 440BX is not ready for testing, I guess I can
> take time looking around for better deal. 
>  I could check out two bazaars this weekend.

I'm in the same situation...I'm looking for them both for the P2-99 (NOT
P2B that it was supposed to be, dammit) that I'll be getting soon from
fleabay, and the Slot A board that I'm working on now, and possibly even
my Intel i815 board once I've messed around with (and messed up) the
Slot A. The only local store that might have such a thing pissed me off
last time I was in there, so I've turned to the net, but it seems most
suppliers are interested in orders larger than 4 chips (2 for each
board). Did manage to find out that FrozenCPU.com carries the Bios
Saviors, and I've never had a bad experience with them, I've used them
several times over the years, and when you figure in shipping, they're
quite a bit cheaper than direct from ioss.

-Corey

[Peter Stuge]

On Tue, Nov 14, 2006 at 01:43:13AM -0500, Corey wrote:
> On Mon, 2006-11-13 at 17:45 +0100, Shaddam Corrino IV wrote:
> > I was looking around today and all that I could find was "certified
> > pre-owned" 256KiB flash ROMs for around $6 each.

[..]
> board). Did manage to find out that FrozenCPU.com carries the Bios
> Saviors, and I've never had a bad experience with them, I've used
> them several times over the years, and when you figure in shipping,
> they're quite a bit cheaper than direct from ioss.

The BIOS savior $25 is definately worthwhile. Zero risk of shorting
the mainboard out while hotswapping flash chips then.

//Peter

[Shaddam Corrino IV]

Hello,

Said MBO has in their BIOS two options (a) support USB keyboard, and (b)
power on via hot key. USB keyboard works in bios, and power on via PS/2
keyboard works. But power on via usb keyboard does not seem to work. Is
there any chance I'll be able to power on this MBO via usb device using LB?

Also where can I find datasheets for Northbridge i82443BX (440BX), and
Southbridge i82371AB/EB/MB ?

[Uwe Hermann]

On Fri, Nov 17, 2006 at 01:28:16PM +0100, Shaddam Corrino IV wrote:
> Also where can I find datasheets for Northbridge i82443BX (440BX), and
> Southbridge i82371AB/EB/MB ?

www.intel.com, just search for the part names...

Uwe.

[Shaddam Corrino IV]

On 11/13/06, Peter Stuge  wrote:
> > Winbond W29C020-90
> > It seems it is DIP-32 5V 256KiB.
>
> Really any 5V 32-pin DIP 256 (probably also 512) kb flash with the
> same pinout will be compatible. Lots of different chips.

Is there some sort of memtest86 for eeproms?

I got an W29C020C-90B and an W29C020-12 for about total $10.

If I'm reading datascheets correctly the first one has about 1k cycles and
second one has about 10k cycles.

They have one week warrany on them so I would like to test them if they are
free of errors even if the 440bx is not ready for prime time. Is there some
easy way to do this?

[Peter Stuge]

On Mon, Nov 20, 2006 at 03:41:47PM +0100, Shaddam Corrino IV wrote:
> Is there some sort of memtest86 for eeproms?
>
> I got an W29C020C-90B and an W29C020-12 for about total $10.
>
> If I'm reading datascheets correctly the first one has about 1k
> cycles and second one has about 10k cycles.

Sounds about right, although 1k is a bit low.
> They have one week warrany on them so I would like to test them if
> they are free of errors even if the 440bx is not ready for prime
> time. Is there some easy way to do this?

If they're new I wouldn't worry, if they're used and haven't failed
yet there's no way to tell how many cycles are left unfortunately.

But you can test that they actually work right now.

It's a good idea to mark your chips somehow so they don't get mixed
up.

1. Boot your system with the normal BIOS.
2. Run: flashrom -r original.bin
3. Email original.bin to self.
4. Remove your original flash chip while the board is powered. Put
   one of the new chips in. Observe polarity! There's a notch on one
   end of the package.
5. Run: flashrom -wv original.bin
6. If the write+verify works you should be able to reboot with the
   new chip working as your BIOS.

Repeat for each chip.

//Peter

[Shaddam Corrino IV]

On 11/21/06, Peter Stuge  wrote:
> > They have one week warrany on them so I would like to test them if
> > they are free of errors even if the 440bx is not ready for prime
> > time. Is there some easy way to do this?
> But you can test that they actually work right now.
> Boot your system with the normal BIOS.
> Run: flashrom -r original.bin
> Email original.bin to self.
> Remove your original flash chip while the board is powered. Put
> one of the new chips in. Observe polarity! There's a notch on one
> end of the package.
> Run: flashrom -wv original.bin
> If the write+verify works you should be able to reboot with the
> new chip working as your BIOS.

Well both chips failed verify. Read and Erase seems to work fine, yet verify
seems to fail at the same place for both chips 0x2f. Doing less on written
binary it seems that there are areas that were left zeroed, and unwritten.
For both chips pattern seems to be the same.

Do I blame eeprom and return them or problem is elsewhere?

./flashrom -r ATC-6220.Award.1-1-03-WB.orig.bin
Calibrating delay loop... ok
No LinuxBIOS table found.
Enabling flash write on PIIX4/PIIX4E/PIIX4M...OK
W29C020C found at physical address: 0xfffc0000
Flash part is W29C020C (256 KB)
Reading Flash...done

[please cc]

[Shaddam Corrino IV]

Ok, I have attached a test of erasing (0xff) and then filling the eeproms
with null's. Steps taken are described in write.txt

-rw-r--r-- 1 root root 262144 Nov 21 16:50 W29C020-12.after.erase.bin
-rw-r--r-- 1 root root 262144 Nov 21 16:53 W29C020-12.after.zeroing.bin
-rw-r--r-- 1 root root 262144 Nov 21 16:59 W29C020C-90B.after.erase.bin
-rw-r--r-- 1 root root 262144 Nov 21 17:00 W29C020C-90B.after.zeroing.bin
-rw-r--r-- 1 root root   3896 Nov 21 17:03 write.txt

I would like opinion. Do I run to store and replace both eerpoms, or there
might be issue with flash writer. I need to decide by tomorrow.

[please cc]

[ron minnich]

yow. I would say timing problem but it is odd that it is just that one
line. It should repeat more I think.

can you afford to buy some parts? It would help figure this out.

ron

[Shaddam Corrino IV]

On 11/21/06, ron minnich  wrote:
> yow. I would say timing problem but it is odd that it is just that one
> line. It should repeat more I think.

Can I somehow slow-down the writes to see if that's a timing problem?

[ron minnich]

I don't think it's timing. I think the parts are crap.

ron