A bit OT: external equipment for writing...

[Vic Berdin]

Hi,

This maybe the wrong list to post this but since you guys are into
the field of creating your own BIOSware, I would like to ask if 
anyone knows of a device or equipment (connected via stadard PC port) 
that may be used in order to write data into a BIOS chip?

I am currently scouring the net for any material (software/hardware)
that can aid me in safely replacing the BIOSware on an SST49LF008A
FWH that socketted on an AMD GX2-based target board.
The said board and SST chip is not (yet) supported (or listed) by 
this project.

However, our goal is to replace the existing BIOSware with Linux.

TIA - Vic

[Antony Stone]

On Tuesday 30 March 2004 1:09 pm, Vic Berdin wrote:
> Hi,
>
> This maybe the wrong list to post this but since you guys are into
> the field of creating your own BIOSware, I would like to ask if
> anyone knows of a device or equipment (connected via stadard PC port)
> that may be used in order to write data into a BIOS chip?

BIOS chips are standard flash parts - a commercial flash programmer will do 
the job (but don't expect such a machine to be cheap).

Failing that, you could try using the DOS-based flash programming utility 
provided by the vendor of your motherboard?   That can write new BIOS images 
into the chip (in-place on the motherboard) - it's worth seeing if it'll 
write an arbitrary binary file for you.

Also, the MTD utilities may be able to help you out.

Regards,

Antony.

[Ian Smith]

At 13:25 30/03/2004, Antony Stone wrote:
> On Tuesday 30 March 2004 1:09 pm, Vic Berdin wrote:
> > Hi,
> >
> > This maybe the wrong list to post this but since you guys are into
> > the field of creating your own BIOSware, I would like to ask if
> > anyone knows of a device or equipment (connected via stadard PC port)
> > that may be used in order to write data into a BIOS chip?
>
> BIOS chips are standard flash parts - a commercial flash programmer will do
> the job (but don't expect such a machine to be cheap).

We use a Batronix USB programmer:
> http://progshop.com/shop/programmer/usb-chip-programmer/index.shtml

List price is 289 USD (plus you'll need a PLCC adapter) which is quite a 
bit cheaper than a lot of the others on the market, and it will program a 
huge variety of different chips. It's pretty fast too (much faster than any 
serial/parallel port programmer)

Cheers

Ian

[Richard Smith]

Ian Smith wrote:
> At 13:25 30/03/2004, Antony Stone wrote:
> > the job (but don't expect such a machine to be cheap).
>
> We use a Batronix USB programmer:
> > http://progshop.com/shop/programmer/usb-chip-programmer/index.shtml
>
> List price is 289 USD (plus you'll need a PLCC adapter) which is quite a 
> bit cheaper than a lot of the others on the market, and it will program 
> a huge variety of different chips. It's pretty fast too (much faster 
> than any serial/parallel port programmer)

Wow. our experience with this device has been totally opposite. In my 
opinion it has about the worst programmer user interface I've ever used. 
  It won't install and run on 2 of our machines (an XP box and a win 98 
box) It dosen't do offsets.  Many of the dialog boxes pop up in German. 
  It does a bad job of figuring out when it can skip 0xff's in blank 
areas.  It wants to always read a part first before programming it.  And 
our Advin Pilot U40 which is a parallel port programmer is faster.

I could go on but I'll stop before I get worked up.

The hardware itself seems sound but the software blows hard.  So hard 
that I've seriously considered snooping the data stream and writeing a 
short sweet python cli utility for it.

[Ian Smith]

At 17:40 30/03/2004, Richard Smith wrote:
> Ian Smith wrote:
> > At 13:25 30/03/2004, Antony Stone wrote:
> > > the job (but don't expect such a machine to be cheap).
> >
> > We use a Batronix USB programmer:
> > > http://progshop.com/shop/programmer/usb-chip-programmer/index.shtml
> >
> > List price is 289 USD (plus you'll need a PLCC adapter) which is quite a 
> > bit cheaper than a lot of the others on the market, and it will program a 
> > huge variety of different chips. It's pretty fast too (much faster than 
> > any serial/parallel port programmer)
>
> Wow. our experience with this device has been totally opposite. In my 
> opinion it has about the worst programmer user interface I've ever 
> used.  It won't install and run on 2 of our machines (an XP box and a win 
> 98 box) It dosen't do offsets.  Many of the dialog boxes pop up in 
> German.  It does a bad job of figuring out when it can skip 0xff's in 
> blank areas.  It wants to always read a part first before programming 
> it.  And our Advin Pilot U40 which is a parallel port programmer is faster.
>
> I could go on but I'll stop before I get worked up.
>
> The hardware itself seems sound but the software blows hard.  So hard that 
> I've seriously considered snooping the data stream and writeing a short 
> sweet python cli utility for it.

Interesting!

Maybe we've just been lucky as I've only installed it on a Win2K box, and 
have only really used it for blowing BIOS proms. I've not seen the German 
dialog boxes either.

As for performance, it's well ahead of the Stag & Data IO programmers I've 
used in the past, but granted that isn't saying much.

Cheers

Ian

[Steve Gehlbach]

Ian Smith wrote:
> > > We use a Batronix USB programmer:

The one I see mentioned the most, and the one I have, is from Needhams 
(www.needhams.com).  More money that what you are talking about, but I 
have never had a problem.  I heaved the DataIO into the dumpster, 
manuals, modules, everything.  The Needhams was less than a software 
upgrade on the DataIO.

[ron minnich]

flash_rom should work fine, depending on chipset etc.

ron

[Hendricks David W.]

The 49LF008A was added to flash_rom just recently. You should be able to 
use any mainboard with the appropriate socket and one of these 
southbridges: SiS630, SiS5595, Intel E7500, VIA VT8231, VIA VT8235, AMD 
SC1100, AMD CS5530, AMD 8111.

The AMDSC1100 was the southbridge on some GX1 based boards we used here 
at the Lab a while back. I'm uncertain whether flash_rom will work with 
whatever is on your GX2 based board, but it might be worth a try. If it 
doesn't work, you can probably grab a cheap mainboard off pricewatch.com 
and use flash_rom anyway. Beats paying hundreds for specialized hardware 
plus some rediculous amount for an MS Windows license :)

On Tue, 30 Mar 2004, Vic Berdin wrote:
> Hi,
>
> This maybe the wrong list to post this but since you guys are into
> the field of creating your own BIOSware, I would like to ask if 
> anyone knows of a device or equipment (connected via stadard PC port) 
> that may be used in order to write data into a BIOS chip?
>
> I am currently scouring the net for any material (software/hardware)
> that can aid me in safely replacing the BIOSware on an SST49LF008A
> FWH that socketted on an AMD GX2-based target board.
> The said board and SST chip is not (yet) supported (or listed) by 
> this project.
>
> However, our goal is to replace the existing BIOSware with Linux.
>
> TIA - Vic
>
> Linuxbios mailing list
> [e-mail]
> http://www.clustermatic.org/mailman/listinfo/linuxbios

[Vic Berdin]

Hi,
> -----Original Message-----
> From: Hendricks David W. [mailto:[e-mail]]
> Sent: Wednesday, March 31, 2004 5:09 AM
> To: Vic Berdin
> Cc: [e-mail]
> Subject: Re: A bit OT: external equipment for writing...
> The 49LF008A was added to flash_rom just recently. You should be able to 
> use any mainboard with the appropriate socket and one of these 
> southbridges: SiS630, SiS5595, Intel E7500, VIA VT8231, VIA VT8235, AMD 
> SC1100, AMD CS5530, AMD 8111.

This is good news! My companion southbridge chip however is an AMDCS5535. 
Will this bring about problems? Anyways, I'll checkout the recent updates
from CVS and try to give this a shot. Any other info are very much welcome.
> The AMDSC1100 was the southbridge on some GX1 based boards we used here 
> at the Lab a while back. I'm uncertain whether flash_rom will work with 
> whatever is on your GX2 based board, but it might be worth a try. If it 
> doesn't work, you can probably grab a cheap mainboard off pricewatch.com 
> and use flash_rom anyway. Beats paying hundreds for specialized hardware 
> plus some rediculous amount for an MS Windows license :)

Best regards,
Vic

[ron minnich]

On Wed, 31 Mar 2004, Vic Berdin wrote:
> This is good news! My companion southbridge chip however is an AMDCS5535.

try it, I think it will also work.

ron

[Vic Berdin]

Hi Ron, everyone,
> -----Original Message-----
> From: ron minnich [mailto:[e-mail]]
> Sent: Wednesday, March 31, 2004 11:16 AM
> To: Vic Berdin
> Cc: Hendricks David W.; [e-mail]
> Subject: RE: A bit OT: external equipment for writing...
> On Wed, 31 Mar 2004, Vic Berdin wrote:
> > This is good news! My companion southbridge chip however is an AMDCS5535.
> try it, I think it will also work.

I've also been going through the Sis630 online doc in order to
have an idea as to what I should do. Unfortunately, step 6
(replacing BIOS chip with a DOC), cannot be applied on my case.
My target board has no DOC support. Either (1) burn the kernel
image directly to the SST49LF008A chip that is socketted (not
soldered) on the target board, or (2) get an external device that
can write into an SST49LF008A.
In an attempt to do option (1), I downloaded driver codes for
49LF008 from SST:
http://www.sst.com/downloads/software_driver/SST49LF008A.txt,
and tried to make use of the "int Check_SST_49LF008A()" function.
Unfortunately, all I'm getting is a segfault. It seems that the
code is referring to an invalid address that should be specific
to my target SBC.
Anyways, if I may ask the list, has anyone tried the approach
to write directly into a BIOS chip using C when replacing the
BIOSware? If so, any hints/source codes/etc?
> ron

TIA - Vic

[ron minnich]

On Wed, 31 Mar 2004, Vic Berdin wrote:
> In an attempt to do option (1), I downloaded driver codes for
> 49LF008 from SST:
> http://www.sst.com/downloads/software_driver/SST49LF008A.txt,
> and tried to make use of the "int Check_SST_49LF008A()" function.
> Unfortunately, all I'm getting is a segfault. It seems that the
> code is referring to an invalid address that should be specific
> to my target SBC.

I would guess that will never work, as the stuff usually is designed to 
run under DOS with no protections.
> Anyways, if I may ask the list, has anyone tried the approach
> to write directly into a BIOS chip using C when replacing the
> BIOSware? If so, any hints/source codes/etc?

flash_rom

ron