[LinuxBIOS] W39V040BPZ or vt8237. Flashing problems.

[Adam Talbot]

Have three motherboards running cn700, VT8237 chips set.  All three
boards have W39V040BPZ bios chips.  I can not program them with
flashrom.  Flashrom see the chip as W39V040B.  Flashrom start on the
flash and just sits there.  When flashed with the -V option I can see
the memory address is not changing.  But, big problem is the fact that
the chip does get a bit flashed, as, I can no longer boot. So there is
something happening. 
Looking at the data sheet for the W39V040B, all the PZ gives me is the
PLCC package and lead free. So, is this a problem with the vt8237 south
bridge?  Where should I go to start debugging?
-Adam

[Corey Osgood]

Adam Talbot wrote:
> Have three motherboards running cn700, VT8237 chips set.  All three
> boards have W39V040BPZ bios chips.  I can not program them with
> flashrom.  Flashrom see the chip as W39V040B.  Flashrom start on the
> flash and just sits there.  When flashed with the -V option I can see
> the memory address is not changing.  But, big problem is the fact that
> the chip does get a bit flashed, as, I can no longer boot. So there is
> something happening. 
> Looking at the data sheet for the W39V040B, all the PZ gives me is the
> PLCC package and lead free. So, is this a problem with the vt8237 south
> bridge?  Where should I go to start debugging?
> -Adam

This is the jetway board? Hop into the factory BIOS, under
Miscellaneous, and disable "Flash Write Protect". I've been flashing
mine with the exact same chipset and flash part for months now, without
any problems.

-Corey

[Adam Talbot]

Yep, my jetway has that.  It is off.  But I have not tried flashrom on
that board. Can not seem to get the stock BIOS to boot from USB.  The
other two boards are PCChips V21G V1.0C
-Adam
Corey Osgood wrote:
> Adam Talbot wrote:
> > Have three motherboards running cn700, VT8237 chips set.  All three
> > boards have W39V040BPZ bios chips.  I can not program them with
> > flashrom.  Flashrom see the chip as W39V040B.  Flashrom start on the
> > flash and just sits there.  When flashed with the -V option I can see
> > the memory address is not changing.  But, big problem is the fact that
> > the chip does get a bit flashed, as, I can no longer boot. So there is
> > something happening. 
> > Looking at the data sheet for the W39V040B, all the PZ gives me is the
> > PLCC package and lead free. So, is this a problem with the vt8237 south
> > bridge?  Where should I go to start debugging?
> > -Adam
>
> This is the jetway board? Hop into the factory BIOS, under
> Miscellaneous, and disable "Flash Write Protect". I've been flashing
> mine with the exact same chipset and flash part for months now, without
> any problems.
>
> -Corey

[Corey Osgood]

Adam Talbot wrote:
> Yep, my jetway has that.  It is off.  But I have not tried flashrom on
> that board. Can not seem to get the stock BIOS to boot from USB.

It's a pain in the rear. Boot the board with the USB flash drive plugged
in. Depending on if you have a hard drive installed or not, shut it down
either at the grub prompt or "cannot find boot disk prompt". Then, boot
it BACK up, go into the BIOS menu, then go into the "Hard disk priority"
in the "Advanced" menu. USB boot option should be there now. I think
it's fixed in the latest BIOS.

-Corey
> The
> other two boards are PCChips V21G V1.0C
> -Adam

Dunno anything about those, probably some sort of GPIO protection in
place. The jetway has a similar effect if the write protection is
enabled, it takes forever attempting to program it and only one range
(60, I think) actually gets programmed.

-Corey

[Gregg C Levine]

Hello!
Would one of you good people please provide something so that I may examine
this "jetway" board? A link for example would be very helpful.
> -----Original Message-----
> From: linuxbios-bounces+hansolofalcon=[e-mail]

[mailto:linuxbios-
> bounces+hansolofalcon=[e-mail]] On Behalf Of Corey

Osgood
> Sent: Friday, September 28, 2007 5:28 PM
> To: Adam Talbot
> Cc: [e-mail]
> Subject: Re: [LinuxBIOS] W39V040BPZ or vt8237. Flashing problems.
>
> Adam Talbot wrote:
> > Have three motherboards running cn700, VT8237 chips set.  All three
> > boards have W39V040BPZ bios chips.  I can not program them with
> > flashrom.  Flashrom see the chip as W39V040B.  Flashrom start on the
> > flash and just sits there.  When flashed with the -V option I can see
> > the memory address is not changing.  But, big problem is the fact that
> > the chip does get a bit flashed, as, I can no longer boot. So there is
> > something happening.
> > Looking at the data sheet for the W39V040B, all the PZ gives me is the
> > PLCC package and lead free. So, is this a problem with the vt8237 south
> > bridge?  Where should I go to start debugging?
> > -Adam
>
> This is the jetway board? Hop into the factory BIOS, under
> Miscellaneous, and disable "Flash Write Protect". I've been flashing
> mine with the exact same chipset and flash part for months now, without
> any problems.
>
> -Corey

[Corey Osgood]

Gregg C Levine wrote:
> Hello!
> Would one of you good people please provide something so that I may examine
> this "jetway" board? A link for example would be very helpful.

Jetway J7F2WE-series, support is in the works and nearly complete ;) I
only have a J7F2WE1G, but the CPU speed is boot-strapped and those
registers aren't touched, so any of the other models should work fine,
when it's done.

http://www.jetway.com.tw/jetway/system/productshow.asp?id=383&cd=c3&...

-Corey

[ron minnich]

On 9/28/07, Adam Talbot  wrote:
> Have three motherboards running cn700, VT8237 chips set.  All three
> boards have W39V040BPZ bios chips.  I can not program them with
> flashrom.  Flashrom see the chip as W39V040B.  Flashrom start on the
> flash and just sits there.  When flashed with the -V option I can see
> the memory address is not changing.  But, big problem is the fact that
> the chip does get a bit flashed, as, I can no longer boot. So there is
> something happening.

I think this is a caching problem. Usually when it just sits there and
"part way flashes", it means that you were mostly programming the
cache, except for a few times when the cache flushed and incidentally
wrote the part.

Does this part have MTRR? IF so, can you do this
cat /proc/mtrr

ron

[Adam Talbot]

Thanks Ron, here is what is hiding in MTRR, explanation?
v21g ~ # cat /proc/mtrr
reg00: base=0x00000000 (   0MB), size= 512MB: write-back, count=1
reg01: base=0x1f000000 ( 496MB), size=  16MB: uncachable, count=1
reg02: base=0xe8000000 (3712MB), size= 128MB: write-combining, count=2
reg03: base=0x1ef00000 ( 495MB), size=   1MB: write-through, count=1

ron minnich wrote:
> On 9/28/07, Adam Talbot  wrote:
> > Have three motherboards running cn700, VT8237 chips set.  All three
> > boards have W39V040BPZ bios chips.  I can not program them with
> > flashrom.  Flashrom see the chip as W39V040B.  Flashrom start on the
> > flash and just sits there.  When flashed with the -V option I can see
> > the memory address is not changing.  But, big problem is the fact that
> > the chip does get a bit flashed, as, I can no longer boot. So there is
> > something happening.
>
> I think this is a caching problem. Usually when it just sits there and
> "part way flashes", it means that you were mostly programming the
> cache, except for a few times when the cache flushed and incidentally
> wrote the part.
>
> Does this part have MTRR? IF so, can you do this
> cat /proc/mtrr
>
> ron

[ron minnich]

On 9/28/07, Adam Talbot  wrote:
> Thanks Ron, here is what is hiding in MTRR, explanation?
> v21g ~ # cat /proc/mtrr
> reg00: base=0x00000000 (   0MB), size= 512MB: write-back, count=1
> reg01: base=0x1f000000 ( 496MB), size=  16MB: uncachable, count=1
> reg02: base=0xe8000000 (3712MB), size= 128MB: write-combining, count=2
> reg03: base=0x1ef00000 ( 495MB), size=   1MB: write-through, count=1

sadly, I don't think MTRRs are covering flash.
none of these cover flash space. This just got harder ...can you try a
different board, just to try to see if it is flashrom or something
else?

ron

[Adam Talbot]

-Ron
Just tried the Jetway board.  Flashed just fine.  Seems to be a problem
with the PCChips board.  What should I start checking?
-Adam

ron minnich wrote:
> On 9/28/07, Adam Talbot  wrote:
> > Thanks Ron, here is what is hiding in MTRR, explanation?
> > v21g ~ # cat /proc/mtrr
> > reg00: base=0x00000000 (   0MB), size= 512MB: write-back, count=1
> > reg01: base=0x1f000000 ( 496MB), size=  16MB: uncachable, count=1
> > reg02: base=0xe8000000 (3712MB), size= 128MB: write-combining, count=2
> > reg03: base=0x1ef00000 ( 495MB), size=   1MB: write-through, count=1
>
> sadly, I don't think MTRRs are covering flash.
> none of these cover flash space. This just got harder ...can you try a
> different board, just to try to see if it is flashrom or something
> else?
>
> ron

[ron minnich]

On 9/28/07, Adam Talbot  wrote:
> -Ron
> Just tried the Jetway board.  Flashed just fine.  Seems to be a problem
> with the PCChips board.  What should I start checking?

the partial flash is very confusing. That's usually a caching
interaction. But MTRRs are not set to cache that region. If there were
no flashing, then it would likely be a GPIO that disables flash
writes. If we can get linuxbios up on this PCChips board we can see if
flash works then. There may be SMM stuff going on here too.

ron

[Adam Talbot]

Some more information to follow.  If possible, I would like to try to
fix this.  So, if any one has a few good places to look, and some good
places in the code to add debug statements, that would be great.  I am
not all that familiar with flashrom, yet. What other BIOS chips are on
the CN700+VT8237 boards, I would like a different chip, just to prove it
is not the chip(s).
-Adam Talbot

v21g flashrom # ./flashrom -r backup.bin
Calibrating delay loop... ok
No LinuxBIOS table found.
Found chipset "VT8237": Enabling flash write... OK.
W39V040B found at physical address: 0xfff80000
Flash part is W39V040B (512 KB)
Reading Flash...done

v21g flashrom # time ./flashrom -wV backup.bin
Calibrating delay loop... 496M loops per second. ok
No LinuxBIOS table found.
Found chipset "VT8237": Enabling flash write... OK.
Probing for Am29F040B, 512 KB
probe_29f040b: id1 0x0, id2 0x0
Probing for Am29F016D, 2048 KB
probe_29f040b: id1 0xff, id2 0xff
Probing for AE49F2008, 256 KB
probe_jedec: id1 0xda, id2 0x54
Probing for At29C040A, 512 KB
probe_jedec: id1 0xda, id2 0x54
Probing for At29C020, 256 KB
probe_jedec: id1 0xda, id2 0x54
Probing for Mx29f002, 256 KB
probe_29f002: id1 0xda, id2 0x54
Probing for SST29EE020A, 256 KB
probe_jedec: id1 0xda, id2 0x54
Probing for SST28SF040A, 512 KB
probe_28sf040: id1 0x0, id2 0x0
Probing for SST39SF010A, 128 KB
probe_jedec: id1 0xda, id2 0x54
Probing for SST39SF020A, 256 KB
probe_jedec: id1 0xda, id2 0x54
Probing for SST39SF040, 512 KB
probe_jedec: id1 0xda, id2 0x54
Probing for SST39VF020, 256 KB
probe_jedec: id1 0xda, id2 0x54
Probing for SST49LF040B, 512 KB
probe_jedec: id1 0xda, id2 0x54
Probing for SST49LF040, 512 KB
probe_jedec: id1 0xda, id2 0x54
Probing for SST49LF020A, 256 KB
probe_jedec: id1 0xda, id2 0x54
Probing for SST49LF080A, 1024 KB
probe_jedec: id1 0xff, id2 0xff
Probing for SST49LF002A/B, 256 KB
probe_jedec: id1 0xda, id2 0x54
Probing for SST49LF003A/B, 384 KB
probe_jedec: id1 0xda, id2 0x54
Probing for SST49LF004A/B, 512 KB
probe_jedec: id1 0xda, id2 0x54
Probing for SST49LF008A, 1024 KB
probe_jedec: id1 0xff, id2 0xff
Probing for SST49LF004C, 512 KB
probe_49lfxxxc: id1 0x0, id2 0x0
Probing for SST49LF008C, 1024 KB
probe_49lfxxxc: id1 0xff, id2 0xff
Probing for SST49LF016C, 2048 KB
probe_49lfxxxc: id1 0xff, id2 0xff
Probing for SST49LF160C, 2048 KB
probe_49lfxxxc: id1 0xff, id2 0xff
Probing for Pm49FL002, 256 KB
probe_jedec: id1 0xda, id2 0x54
Probing for Pm49FL004, 512 KB
probe_jedec: id1 0xda, id2 0x54
Probing for W29C011, 128 KB
probe_jedec: id1 0xda, id2 0x54
Probing for W29C040P, 512 KB
probe_jedec: id1 0xda, id2 0x54
Probing for W29C020C, 256 KB
probe_jedec: id1 0xda, id2 0x54
Probing for W29EE011, 128 KB
probe_w29ee011: id1 0xff, id2 0xff
Probing for W49F002U, 256 KB
probe_jedec: id1 0xda, id2 0x54
Probing for W49V002A, 256 KB
probe_jedec: id1 0xda, id2 0x54
Probing for W49V002FA, 256 KB
probe_jedec: id1 0xda, id2 0x54
Probing for W39V040FA, 512 KB
probe_jedec: id1 0xda, id2 0x54
Probing for W39V040A, 512 KB
probe_jedec: id1 0xda, id2 0x54
Probing for W39V040B, 512 KB
probe_jedec: id1 0xda, id2 0x54
W39V040B found at physical address: 0xfff80000
Flash part is W39V040B (512 KB)
Flash image seems to be a legacy BIOS. Disabling checks.
Programming Page: 0000 at address: 0x00000000
### Ctrl-C ###
real    1134m28.975s
user    1134m49.425s
sys     0m0.277s

v21g flashrom # ./flashrom               
Calibrating delay loop... ok
No LinuxBIOS table found.
Found chipset "VT8237": Enabling flash write... OK.
No EEPROM/flash device found.

ron minnich wrote:
> On 9/28/07, Adam Talbot  wrote:
> > -Ron
> > Just tried the Jetway board.  Flashed just fine.  Seems to be a problem
> > with the PCChips board.  What should I start checking?
>
> the partial flash is very confusing. That's usually a caching
> interaction. But MTRRs are not set to cache that region. If there were
> no flashing, then it would likely be a GPIO that disables flash
> writes. If we can get linuxbios up on this PCChips board we can see if
> flash works then. There may be SMM stuff going on here too.
>
> ron

[ron minnich]

do you have URL for 8237 docs?

ron

[Adam Talbot]

no, sorry
-Adam

ron minnich wrote:
> do you have URL for 8237 docs?
>
> ron

[Corey Osgood]

Adam Talbot wrote:
> no, sorry
> -Adam
>
> ron minnich wrote:
> > do you have URL for 8237 docs?
> >
> > ron

http://www.via.com.tw/en/downloads/datasheets/chipsets/VT8237R_SouthBridge_R...

[Carl-Daniel Hailfinger]

Hi Adam,

did you figure out the reason of the bad interaction between vt8237 and 
W39V040BPZ?

On 29.09.2007 19:59, Adam Talbot wrote:
> Some more information to follow.  If possible, I would like to try to
> fix this.  So, if any one has a few good places to look, and some good
> places in the code to add debug statements, that would be great.  I am
> not all that familiar with flashrom, yet. What other BIOS chips are on
> the CN700+VT8237 boards, I would like a different chip, just to prove it
> is not the chip(s).
> -Adam Talbot
>
> v21g flashrom # ./flashrom -r backup.bin
> Calibrating delay loop... ok
> No LinuxBIOS table found.
> Found chipset "VT8237": Enabling flash write... OK.
> W39V040B found at physical address: 0xfff80000
> Flash part is W39V040B (512 KB)
> Reading Flash...done

Flash chip is found, reading out works fine.
> v21g flashrom # time ./flashrom -wV backup.bin
> Calibrating delay loop... 496M loops per second. ok
> No LinuxBIOS table found.
> Found chipset "VT8237": Enabling flash write... OK.
> [...]
> Probing for W39V040B, 512 KB
> probe_jedec: id1 0xda, id2 0x54
> W39V040B found at physical address: 0xfff80000
> Flash part is W39V040B (512 KB)
> Flash image seems to be a legacy BIOS. Disabling checks.

Chip found again, is being erased before the writing starts.
> Programming Page: 0000 at address: 0x00000000

It hangs here.
> ### Ctrl-C ###
> real    1134m28.975s
> user    1134m49.425s
> sys     0m0.277s
>
> v21g flashrom # ./flashrom               
> Calibrating delay loop... ok
> No LinuxBIOS table found.
> Found chipset "VT8237": Enabling flash write... OK.
> No EEPROM/flash device found.

I'm seeing exactly the same symptoms with a W39V040BPZ on my MCP51 
Asrock K8NF4G-SATA2 board. Strange fact: If I hot-unplug the flash chip 
and hot-replug it, it is recognized again. Reading it shows the erase 
was successful, but once I try to program even one byte, communication 
with the flash chip fails completely and I have to unplug and replug to 
fix that.

Regards,
Carl-Daniel

[Corey Osgood]

On Mon, 2008-03-03 at 22:50 +0100, Carl-Daniel Hailfinger wrote:
> Hi Adam,
>
> did you figure out the reason of the bad interaction between vt8237 and 
> W39V040BPZ?
>
> On 29.09.2007 19:59, Adam Talbot wrote:
> > Some more information to follow.  If possible, I would like to try to
> > fix this.  So, if any one has a few good places to look, and some good
> > places in the code to add debug statements, that would be great.  I am
> > not all that familiar with flashrom, yet. What other BIOS chips are on
> > the CN700+VT8237 boards, I would like a different chip, just to prove it
> > is not the chip(s).
> > -Adam Talbot
> >
> > v21g flashrom # ./flashrom -r backup.bin
> > Calibrating delay loop... ok
> > No LinuxBIOS table found.
> > Found chipset "VT8237": Enabling flash write... OK.
> > W39V040B found at physical address: 0xfff80000
> > Flash part is W39V040B (512 KB)
> > Reading Flash...done
>
> Flash chip is found, reading out works fine.
> > v21g flashrom # time ./flashrom -wV backup.bin
> > Calibrating delay loop... 496M loops per second. ok
> > No LinuxBIOS table found.
> > Found chipset "VT8237": Enabling flash write... OK.
> > [...]
> > Probing for W39V040B, 512 KB
> > probe_jedec: id1 0xda, id2 0x54
> > W39V040B found at physical address: 0xfff80000
> > Flash part is W39V040B (512 KB)
> > Flash image seems to be a legacy BIOS. Disabling checks.
>
> Chip found again, is being erased before the writing starts.
> > Programming Page: 0000 at address: 0x00000000
>
> It hangs here.
> > ### Ctrl-C ###
> > real    1134m28.975s
> > user    1134m49.425s
> > sys     0m0.277s
> >
> > v21g flashrom # ./flashrom               
> > Calibrating delay loop... ok
> > No LinuxBIOS table found.
> > Found chipset "VT8237": Enabling flash write... OK.
> > No EEPROM/flash device found.
>
> I'm seeing exactly the same symptoms with a W39V040BPZ on my MCP51 
> Asrock K8NF4G-SATA2 board. Strange fact: If I hot-unplug the flash chip 
> and hot-replug it, it is recognized again. Reading it shows the erase 
> was successful, but once I try to program even one byte, communication 
> with the flash chip fails completely and I have to unplug and replug to 
> fix that.

I'm seeing exactly the same thing with the cn700+vt8237r on Jetway
J7F2W, same flash chip, except that where it hangs seems random. I had
assumed the flash chips were write protected in some way that I didn't
care to work around, since they're the factory BIOS. Removing the chip
then reinserting it allows flashrom to detect it again, and writing
actually works and verifys, extremely slowly, perhaps 1 out of 100
times. PMC PM49FL004 works perfectly fine on the same board.

-Corey

[Peter Stuge]

On Mon, Mar 03, 2008 at 06:42:05PM -0500, Corey Osgood wrote:
> > If I hot-unplug the flash chip and hot-replug it, it is recognized
> > again. Reading it shows the erase was successful, but once I try
> > to program even one byte, communication with the flash chip fails
> > completely and I have to unplug and replug to fix that.
>
> I'm seeing exactly the same thing with the cn700+vt8237r on Jetway
> J7F2W, same flash chip, except that where it hangs seems random.

I take this to mean that 8237 and the W39V040BPZ aren't really
compatible.

I would investigate if the timing for the two is compatible on paper.
Also, maybe the 8237 has some speed setting like the ITE.

//Peter

[Carl-Daniel Hailfinger]

Hi Yinghai,

I am seeing problems with Winbond W39V040B and MCP51. It seems these 
problems are probably related to LPC timing. Did you check the MCP51 
datasheets for LPC timing and flashing when you were at AMD?

On 04.03.2008 01:15, Peter Stuge wrote:
> On Mon, Mar 03, 2008 at 06:42:05PM -0500, Corey Osgood wrote:
> > > If I hot-unplug the flash chip and hot-replug it, it is recognized
> > > again. Reading it shows the erase was successful, but once I try
> > > to program even one byte, communication with the flash chip fails
> > > completely and I have to unplug and replug to fix that.
> >
> > I'm seeing exactly the same thing with the cn700+vt8237r on Jetway
> > J7F2W, same flash chip, except that where it hangs seems random.
>
> I take this to mean that 8237 and the W39V040BPZ aren't really
> compatible.
>
> I would investigate if the timing for the two is compatible on paper.
> Also, maybe the 8237 has some speed setting like the ITE.

The interesting thing is that the chip is the factory installed chip and 
flashing it should (not yet tested) work fine with the vendor flash tool.

Regards,
Carl-Daniel

[Carl-Daniel Hailfinger]

On 04.03.2008 04:53, Carl-Daniel Hailfinger wrote:
> On 04.03.2008 01:15, Peter Stuge wrote:
> > On Mon, Mar 03, 2008 at 06:42:05PM -0500, Corey Osgood wrote:
> > > > If I hot-unplug the flash chip and hot-replug it, it is recognized
> > > > again. Reading it shows the erase was successful, but once I try
> > > > to program even one byte, communication with the flash chip fails
> > > > completely and I have to unplug and replug to fix that.
> > >
> > > I'm seeing exactly the same thing with the cn700+vt8237r on Jetway
> > > J7F2W, same flash chip, except that where it hangs seems random.
> >
> > I take this to mean that 8237 and the W39V040BPZ aren't really
> > compatible.
> >
> > I would investigate if the timing for the two is compatible on paper.
> > Also, maybe the 8237 has some speed setting like the ITE.
>
> The interesting thing is that the chip is the factory installed chip and 
> flashing it should (not yet tested) work fine with the vendor flash tool.

Yes, the 8237 has some LPC timing settings which could help here. See 
the data shett for details.

Regards,
Carl-Daniel