EPIA-ME6000

[Leon Woestenberg]

      Hello,

I have been fiddling with etherboot 5.0.10 and freebios(v1) from CVS today
on a EPIA-ME6000 platform.

This is the unit with a low speed clock (500MHz?) Eden C3 processor
which copes with passive cooling. I think it is EPIA-M9000 alike otherwise.

Using a hardware flash programmer, I first copied the VIA BIOS from the
STT39SF040A (256Kbytes) into a STMicro 29F040B (512Kbytes) device.

That only worked after I copied the BIOS into the upper half of the larger
device (from 0x40000 to 0x7FFFF). So I recon the highest address line
is not connected on the board and so the FlashROM sees high-impedance
on its most significant address line (and addresses the upper half always)>

I followed the HOWTO/EPIA document to the letter (without the VGA
BIOS at first) booting linuxbios w/ via-rhine.elf payload.

When booting, I merely got garbled output on the serial port.

I have double checked that the config file reads 115200 and that my
terminal reads 115200 8N1. (The 115200 is assembly hardcoded in
the vt82c686 superio code -- does it cope with my different processor
speed? I do not have the datasheet (yet?).

After that, I have tried all speeds on the terminal without result.

Is the EPIA-ME6000 (aka M6000) already reported to work here,
or are you guys'n'gals refering to the M9000 and M10000 alone?

How can I help?

Regards,

Leon.

[SONE Takeshi]

      On Tue, Jul 29, 2003 at 12:09:11PM +0200, Leon Woestenberg wrote:
  That only worked after I copied the BIOS into the
upper half of the larger
 device (from 0x40000 to 0x7FFFF). So I recon the highest address line
 is not connected on the board and so the FlashROM sees high-impedance
 on its most significant address line (and addresses the upper half always)>
I think that it does not mean MSB is not connected.
Since BIOS is expected to live at highest address below 4GB,
you would have to write the 256KB image at highest portion of ROM anyway
if all address lines are connected.

If you had to write it in the lower half, you could say the highest
address line is grounded.

  I followed the HOWTO/EPIA document to the letter
(without the VGA
 BIOS at first) booting linuxbios w/ via-rhine.elf payload.
HOWTO/EPIA is written for original EPIA boards like EPIA-5000 or EPIA-800,
not for EPIA-M(E).
Did you use config file for EPIA-M?

--
Takeshi

[Brian G. Rhodes]

      That board has a VIA vt82c686b south on it?  I built linuxbios for a
similar VIA board (endat) with an 8601 north and vt82c686b south.  That
won't work with the epia or epia-m configs.  but you can go from the
via/vt5426 config.  You are perhaps getting garbage serial output because
you're setting up serial for the wrong superio.

Have you figured out how to enable flash memory write?  I have two 686
datasheets which seem to contradict one another on how it's done.  one
appears like the 8231 south, and another alike the sis630.

Brian G Rhodes
bgr(a)linespeed.net
brhodes(a)visualcircuits.com
[phone]


On Tue, 29 Jul 2003, Leon Woestenberg wrote:

  Hello,

 I have been fiddling with etherboot 5.0.10 and freebios(v1) from CVS today
 on a EPIA-ME6000 platform.

 This is the unit with a low speed clock (500MHz?) Eden C3 processor
 which copes with passive cooling. I think it is EPIA-M9000 alike otherwise.

 Using a hardware flash programmer, I first copied the VIA BIOS from the
 STT39SF040A (256Kbytes) into a STMicro 29F040B (512Kbytes) device.

 That only worked after I copied the BIOS into the upper half of the larger
 device (from 0x40000 to 0x7FFFF). So I recon the highest address line
 is not connected on the board and so the FlashROM sees high-impedance
 on its most significant address line (and addresses the upper half always)>

 I followed the HOWTO/EPIA document to the letter (without the VGA
 BIOS at first) booting linuxbios w/ via-rhine.elf payload.

 When booting, I merely got garbled output on the serial port.

 I have double checked that the config file reads 115200 and that my
 terminal reads 115200 8N1. (The 115200 is assembly hardcoded in
 the vt82c686 superio code -- does it cope with my different processor
 speed? I do not have the datasheet (yet?).

 After that, I have tried all speeds on the terminal without result.

 Is the EPIA-ME6000 (aka M6000) already reported to work here,
 or are you guys'n'gals refering to the M9000 and M10000 alone?

 How can I help?

 Regards,

 Leon.
 _______________________________________________
 Linuxbios mailing list
 Linuxbios(a)clustermatic.org
 http://www.clustermatic.org/mailman/listinfo/linuxbios

[Chung Ming]

      I Believe VIA ME6000 uses VT8235 SouthBridge.

Ming.

----- Original Message ----- 
From: "Brian G. Rhodes" <bgr(a)gw.linespeed.net>
To: "Leon Woestenberg" <leon.woestenberg(a)gmx.net>
Cc: <linuxbios(a)clustermatic.org>
Sent: Tuesday, July 29, 2003 10:05 PM
Subject: Re: EPIA-ME6000


  That board has a VIA vt82c686b south on it?  I built
linuxbios for a
 similar VIA board (endat) with an 8601 north and vt82c686b south.  That
 won't work with the epia or epia-m configs.  but you can go from the
 via/vt5426 config.  You are perhaps getting garbage serial output because
 you're setting up serial for the wrong superio.

 Have you figured out how to enable flash memory write?  I have two 686
 datasheets which seem to contradict one another on how it's done.  one
 appears like the 8231 south, and another alike the sis630.

 Brian G Rhodes
 bgr(a)linespeed.net
 brhodes(a)visualcircuits.com
 [phone]


 On Tue, 29 Jul 2003, Leon Woestenberg wrote:

 > Hello,
 >
 > I have been fiddling with etherboot 5.0.10 and freebios(v1) from CVS
today
  > on a EPIA-ME6000 platform.
 >
 > This is the unit with a low speed clock (500MHz?) Eden C3 processor
 > which copes with passive cooling. I think it is EPIA-M9000 alike
otherwise.
  >
 > Using a hardware flash programmer, I first copied the VIA BIOS from the
 > STT39SF040A (256Kbytes) into a STMicro 29F040B (512Kbytes) device.
 >
 > That only worked after I copied the BIOS into the upper half of the
larger
  > device (from 0x40000 to 0x7FFFF). So I recon the
highest address line
 > is not connected on the board and so the FlashROM sees high-impedance
 > on its most significant address line (and addresses the upper half
always)>
 
 I followed the HOWTO/EPIA document to the letter (without the VGA
 BIOS at first) booting linuxbios w/ via-rhine.elf payload.

 When booting, I merely got garbled output on the serial port.

 I have double checked that the config file reads 115200 and that my
 terminal reads 115200 8N1. (The 115200 is assembly hardcoded in
 the vt82c686 superio code -- does it cope with my different processor
 speed? I do not have the datasheet (yet?).

 After that, I have tried all speeds on the terminal without result.

 Is the EPIA-ME6000 (aka M6000) already reported to work here,
 or are you guys'n'gals refering to the M9000 and M10000 alone?

 How can I help?

 Regards,

 Leon.
 _______________________________________________
 Linuxbios mailing list
 Linuxbios(a)clustermatic.org
 http://www.clustermatic.org/mailman/listinfo/linuxbios
 _______________________________________________
 Linuxbios mailing list
 Linuxbios(a)clustermatic.org
 http://www.clustermatic.org/mailman/listinfo/linuxbios

[Andrew Ip]

      Hi Leon,

  I have been fiddling with etherboot 5.0.10 and
freebios(v1) from CVS today
 on a EPIA-ME6000 platform.
 This is the unit with a low speed clock (500MHz?) Eden C3 processor
 which copes with passive cooling. I think it is EPIA-M9000 alike otherwise.
 Using a hardware flash programmer, I first copied the VIA BIOS from the
 STT39SF040A (256Kbytes) into a STMicro 29F040B (512Kbytes) device.
 That only worked after I copied the BIOS into the upper half of the larger
 device (from 0x40000 to 0x7FFFF). So I recon the highest address line
 is not connected on the board and so the FlashROM sees high-impedance
 on its most significant address line (and addresses the upper half always)>
 I followed the HOWTO/EPIA document to the letter (without the VGA
 BIOS at first) booting linuxbios w/ via-rhine.elf payload.
 When booting, I merely got garbled output on the serial port.
 I have double checked that the config file reads 115200 and that my
 terminal reads 115200 8N1. (The 115200 is assembly hardcoded in
 the vt82c686 superio code -- does it cope with my different processor
 speed? I do not have the datasheet (yet?).
 After that, I have tried all speeds on the terminal without result.
 Is the EPIA-ME6000 (aka M6000) already reported to work here,
 or are you guys'n'gals refering to the M9000 and M10000 alone?
 How can I help?
Have you tried 57600?  EPIA-M6000 seems to have problem to use
115200
under LinuxBIOS unless it is boot from normal bios and then reset the
system.  That's why you are getting garbage.

-Andrew

-- 
Andrew Ip
Email:  aip(a)cwlinux.com
Tel:    [phone]
Fax:    [phone]
Mobile: [phone]

Cwlinux Limited
Unit 202B 2/F Lai Cheong Factory Building,
479-479A Castle Peak Road,
Lai Chi Kok, Kowloon,
Hong Kong.

For public pgp key, please obtain it from http://www.keyserver.net/en.

[Bari Ari]

      Andrew Ip wrote:

 
Have you tried 57600?  EPIA-M6000 seems to have problem to use 115200
under LinuxBIOS unless it is boot from normal bios and then reset the
system.  That's why you are getting garbage.

  
Any idea why? Broken hardware? Undocumented setups for the serial port?

-Bari

[Andrew Ip]

      Hi,

  Any idea why? Broken hardware? Undocumented setups for
the serial port?
Most likely undocumented setup, 'coz my code works on another
custom
board with the same chipset.

-Andrew

-- 
Andrew Ip
Email:  aip(a)cwlinux.com
Tel:    [phone]
Fax:    [phone]
Mobile: [phone]

Cwlinux Limited
Unit 202B 2/F Lai Cheong Factory Building,
479-479A Castle Peak Road,
Lai Chi Kok, Kowloon,
Hong Kong.

For public pgp key, please obtain it from http://www.keyserver.net/en.

[Leon Woestenberg]

      Hello Andrew,

  > I have been fiddling with etherboot 5.0.10 and
freebios(v1) from CVS
today
  on a
EPIA-ME6000 platform.
 ...
 When booting, I merely got garbled output on the serial port.
 I have double checked that the config file reads 115200 and that my
 terminal reads 115200 8N1. (The 115200 is assembly hardcoded in
 the vt82c686 superio code -- does it cope with my different processor
 Have you tried 57600?  EPIA-M6000 seems to have problem to use 115200
 under LinuxBIOS unless it is boot from normal bios and then reset the
 system.  That's why you are getting garbage.

 -Andrew
Just to make sure; where would I need to set the 57600 bitrate? The file I
have been
looking at is the assembler code file in the vt82c686 directory, which just
sets a
register with some values I cannot compare with a datasheet (as I haven't
found one yet).
Or am I looking at the wrong file here: ?

http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/freebios/freebios/src/superi…

Also, I blew up my board (probably ESD problem) during replacing the cooling
compounds. :-/
I am waiting for a new one... This _does_ give me an extra (and original)
FlashROM part. :-)

BTW, does anyone know a source for the STT39F020A Flashes? They're hard to
come by here.

Regards,

Leon.