P2B - "Update data incorrect.. CPUID =..."

[Retro Bob]

P2B Bios message question:

I have some P2B 1.10 systems. Upgraded a few a year or so ago with
Slot-T and 1.4g Tat. kits from Straton Computer. DId the BIOS upgrade
to 1014.003. Everything works fine on those.

Problem:
I just got a couple new upgrade kits from Straton. The Slot-T rev is
the same, 1.4G Tat. again, although the chip is obviously a newer rev.
I did the BIOS upgrade to 1014.003, everything went well. Did the
"load setup defaults", etc. But, now when I boot, I get a message
right after the display of I/O addresses & disk settings, etc that
says: "BIOS update data incorrect - CPUID=000006B4 Update not
loaded".
I also noticed that although the "ESCD Updated" message
still pops, the "DMI Updated message" seems to be missing.

So, questions: Can I get fix this ? Does it have any effect on
running or performance (i.e. can I just ignore it) ?

Thanks,

 

[Paul]

P2B Bios message question:

I have some P2B 1.10 systems. Upgraded a few a year or so ago with
Slot-T and 1.4g Tat. kits from Straton Computer. DId the BIOS upgrade
to 1014.003. Everything works fine on those.

Problem:
I just got a couple new upgrade kits from Straton. The Slot-T rev is
the same, 1.4G Tat. again, although the chip is obviously a newer rev.
I did the BIOS upgrade to 1014.003, everything went well. Did the
"load setup defaults", etc. But, now when I boot, I get a message
right after the display of I/O addresses & disk settings, etc that
says: "BIOS update data incorrect - CPUID=000006B4 Update not
loaded".
I also noticed that although the "ESCD Updated" message
still pops, the "DMI Updated message" seems to be missing.

So, questions: Can I get fix this ? Does it have any effect on
running or performance (i.e. can I just ignore it) ?

Thanks,

It sounds like you need a microcode update. I used CTMC on
mine to do this, from CT magazine. Have a look here:

http://tipperlinne.com/bios6b4.htm

Basically, CTMC uses a BIOS hook specified and invented by
Intel. The BIOS contains code, whereby a 2KB microcode segment
can be burned into the flash BIOS chip, in an area dedicated
to caching the last microcode info. CTMC simply calls this
BIOS hook, passes the 2KB segment, and the BIOS does the work.

When you use CTMC, the results are not permanent, in the sense
that if you change processor types several times, the microcodes
for those processors flush out the cache, and the CTMC program
would have to be used again. But, the CTMC method has the
advantage that modifying an entire BIOS file is not necessary,
so it is a low risk means of solving your problem.

The method requires extracting a 2KB microcode segment from
an existing BIOS file (from another board where Asus included
microcode for 06B4), then using CTMC to cause it to be
stored in the flash chip.

HTH,
Paul

 

[Bolooser11]

Take a look here

http://www.mrufer.ch/pc/tualatin_e.html

 

[Paul]

It sounds like you need a microcode update. I used CTMC on
mine to do this, from CT magazine. Have a look here:

http://tipperlinne.com/bios6b4.htm

Basically, CTMC uses a BIOS hook specified and invented by
Intel. The BIOS contains code, whereby a 2KB microcode segment
can be burned into the flash BIOS chip, in an area dedicated
to caching the last microcode info. CTMC simply calls this
BIOS hook, passes the 2KB segment, and the BIOS does the work.

When you use CTMC, the results are not permanent, in the sense
that if you change processor types several times, the microcodes
for those processors flush out the cache, and the CTMC program
would have to be used again. But, the CTMC method has the
advantage that modifying an entire BIOS file is not necessary,
so it is a low risk means of solving your problem.

The method requires extracting a 2KB microcode segment from
an existing BIOS file (from another board where Asus included
microcode for 06B4), then using CTMC to cause it to be
stored in the flash chip.

HTH,
Paul

As for your question about running or performance, the purpose
of a microcode update is to incorporate any fixes issued by
Intel, for operation of their processor. The bugs in question
could be trivial corner cases (never happen in practice) or
they could be serious ones. Perhaps reading the "Spec Update"
for the processor from the Intel site, would give you some
idea of how many errata are in a given stepping of CPU. The
errata list for most processors is lengthy, and it is hard
to tell exactly which of the errata is important, and which
of them are patched by microcode.

At least WinXP has its own microcode loader built in. But, that
means you are running "unpatched", from POST, until Windows
has started using its drivers (probably when the desktop
appears). Something like Win98 probably doesn't have its own
microcode loader, so you would be exposed to the bugs for the
entire operating session. On a similar note, WinXP SP2 may hang,
if the microcode is not up to date (it did for Prescott processors).
So, that should give some ideas of what a missing microcode
might do.

As for the comment on the mrufer web page, about CTMC not working,
one of the first things CTMC does, is check for the BIOS hook
that supports burning a separate microcode segment. If it doesn't
find the hook, then CTMC will not operate. That could be what
happened in that case.

HTH,
Paul

 

[tk]

LOL just looked at the link, pretty cool but why go through all that
trouble? I had a P2B-F
Years back and have since had a cusl2-c, a P4t-533 and now a P4P800-E Deluxe
and
every one has been a stepup. The only one that had trouble was the p4t-533.
It was an early revision
that was defective and was promptly replaced.

In other words why go to all the trouble? Just curious.

 

[Bolooser11]

Perhaps because mobos don't rust ;-)

LOL just looked at the link, pretty cool but why go through all that
trouble? I had a P2B-F
Years back and have since had a cusl2-c, a P4t-533 and now a P4P800-E Deluxe
and
every one has been a stepup. The only one that had trouble was the p4t-533.
It was an early revision
that was defective and was promptly replaced.

In other words why go to all the trouble? Just curious.

 

[Retro Bob]

As for the comment on the mrufer web page, about CTMC not working,
one of the first things CTMC does, is check for the BIOS hook
that supports burning a separate microcode segment. If it doesn't
find the hook, then CTMC will not operate. That could be what
happened in that case.

Lemme ask a dumb question... when I run aflash and do the BIOS
update, the only option is "Update BIOS without boot block and ESCD".

Is this related to the problem ? Would there be another way of doing
the update or:
a. it's totally unrelated
b. the "hook" is missing so it can't do the update although it
would like to
c. the 1014.003 BIOS doesn't know about this CPU rev so it
skips the update ?

What's a little confusing is that the other 1.4 tat's (last year's
rev) are doing fine and worked OK on this mobo too - no errors.

Bob

 

[P2B]

Lemme ask a dumb question... when I run aflash and do the BIOS
update, the only option is "Update BIOS without boot block and ESCD".

If you use pflash, there's also an option to "Update BIOS including boot
block and ESCD" - however that's not relevant to microcode updates.

Is this related to the problem ? Would there be another way of doing
the update or:
a. it's totally unrelated
b. the "hook" is missing so it can't do the update although it
would like to
c. the 1014.003 BIOS doesn't know about this CPU rev so it
skips the update ?

c.

The BIOS contains a table of CPU step codes and corresponding microcode
updates. The 1014.003 table includes an entry for tA1 stepping
Tualatins, but not for tB1 steppings - so the BIOS finds a tB1 CPU
installed, looks it up in the table, doesn't find it, and displays the
update error message.

 

[Paul]

Lemme ask a dumb question... when I run aflash and do the BIOS
update, the only option is "Update BIOS without boot block and ESCD".

Is this related to the problem ? Would there be another way of doing
the update or:
a. it's totally unrelated
b. the "hook" is missing so it can't do the update although it
would like to
c. the 1014.003 BIOS doesn't know about this CPU rev so it
skips the update ?

What's a little confusing is that the other 1.4 tat's (last year's
rev) are doing fine and worked OK on this mobo too - no errors.

Bob

That is unrelated. I was referring to a command that CTMC uses.
Flashing the BIOS is a similar operation, but doesn't use the
Intel microcode update routine.

Try executing "aflash \?" .

ASUS ACPI BIOSFLASH MEMORY WRITER V2.21%s
Copyright (C) 1994-2002, ASUSTeK COMPUTER INC.
USAGE: AFLASH [Options] [ROMFile]
/AUTO ROMFile Automated update BIOS without Boot Block and ESCD
/BOOT Update BIOS including Boot Block and ESCD

It almost sounds like the /Auto option is being invoked somehow.
Are you calling aflash directly, or using some other program
to call aflash ?

Paul

 

[Retro Bob]

That is unrelated. I was referring to a command that CTMC uses.
Flashing the BIOS is a similar operation, but doesn't use the
Intel microcode update routine.
Understood

Try executing "aflash \?" .

ASUS ACPI BIOSFLASH MEMORY WRITER V2.21%s
Copyright (C) 1994-2002, ASUSTeK COMPUTER INC.
USAGE: AFLASH [Options] [ROMFile]
/AUTO ROMFile Automated update BIOS without Boot Block and ESCD
/BOOT Update BIOS including Boot Block and ESCD

It almost sounds like the /Auto option is being invoked somehow.
Are you calling aflash directly, or using some other program
to call aflash ?

boot win95,
c:> aflash.exe

Interesting. I tried it with /BOOT and it does offer to update them
So, followup question... should I go back and update them all to
1014.003 with the /BOOT option selected ?

 

[Retro Bob]

The BIOS contains a table of CPU step codes and corresponding microcode
updates. The 1014.003 table includes an entry for tA1 stepping
Tualatins, but not for tB1 steppings - so the BIOS finds a tB1 CPU
installed, looks it up in the table, doesn't find it, and displays the
update error message.

OK... sort of what I thought.
TA,
BOb

 

[P2B]

That is unrelated. I was referring to a command that CTMC uses.
Flashing the BIOS is a similar operation, but doesn't use the
Intel microcode update routine.

Understood


Try executing "aflash \?" .

ASUS ACPI BIOSFLASH MEMORY WRITER V2.21%s
Copyright (C) 1994-2002, ASUSTeK COMPUTER INC.
USAGE: AFLASH [Options] [ROMFile]
/AUTO ROMFile Automated update BIOS without Boot Block and ESCD
/BOOT Update BIOS including Boot Block and ESCD

It almost sounds like the /Auto option is being invoked somehow.
Are you calling aflash directly, or using some other program
to call aflash ?

boot win95,
c:> aflash.exe

Interesting. I tried it with /BOOT and it does offer to update them
So, followup question... should I go back and update them all to
1014.003 with the /BOOT option selected ?

No, since none of the P2B-series BIOS updates required a boot block update.

P2B