P2B-DS BIOS request
P2B
P2B-DS, Slot-Ts, and PIII-S cpus which work in slot 1 but not in slot 2.
One thing I'd really like to try is a BIOS for this board which has had
microcode for cpuid 06B4 added. I could probably do this myself - the
information is out there - but I'd have to climb the learning curve and
would probably make mistakes along the way.
I'm hoping there is a kind soul out there, who already has the requisite
tools and knowledge, and would be willing to help.
P2B-DS BIOS 1014 beta 3 with the cpucode.exe file replaced by one from
another Asus BIOS containing 06B4 microcode is all I need for an initial
test but, if it works, I would like one that adds 06B4 to the existing
cpu support. I won't be upset if someone tries to help but provides a
bad BIOS as I have spare BIOS chips and don't mind hot flashing to fix
any failures.
Any takers? Emailed replies would be fine too.
TIA
KURIAKI
ftp://ftp.asuscom.de/pub/ASUSCOM/BIOS/Socket_370/SIS_Chipset/SIS_630et/TUSI-M/without_OnBoard_Lan/1017tmn.zip
--
KURIAKI
84.397 BIOS strings detected and growing in my database everyday.
Follow this link: http://www.unicore.com/biosagent/ba.exe
Run Bios Agent, safe the results to bios.txt and send me to kuria...@prodigy.net.mx
Help me to maintain a updated database with all BIOS motherboards. Thank you.
"P2B" <p...@sympatico.ca> wrote in message news:xUpla.9093$N94.8...@news20.bellglobal.com...
Paul
<p...@sympatico.ca> wrote:
I've been working on the microcode thing, had a nice post with the
results all ready to go, and a crash wiped it out. I am not a happy
camper right now :-(
So, this will be a truncated version.
I've been investigating various ways to do a microcode update over
the last couple of days -
Some options are:
1) Visit WIMS bios website. A guy in Russia has a patcher that
will fix all sorts of things in a BIOS. I downloaded the package
but will leave it for some other time. The patcher fixes a number
of things, including something called "tualatin cache init bug".
The four files you need to use his package are here:
http://k6-2.narod.ru/award/patcher/real_microcodes.rar
http://k6-2.narod.ru/award/patcher/LHA.RAR
http://k6-2.narod.ru/award/patcher/CBROM.RAR
http://k6-2.narod.ru/award/patcher/bp-4rc_F.rar
The k6-2.narod.ru is an alternate download address to rom.by
As I'm a newbie at this, I have no idea what tool decompresses .rar
2) There was mention of awardmod on sourceforge.net, but when I
downloaded it, it didn't install properly, and I do what I do
with all such packages --delete--. If you cannot make an installer
that works, then you probably cannot write code either...
3) Tried a couple of versions of CBROM, but the problem is, to use
this as a method, I'd need a BIOS Savior, because the odds are
against it working the first time.
4) So, this leaves CTMC...
The CTMC package doesn't work exactly as the packaging suggests.
When you unzip the package, you get CTMC.EXE, LHA255.EXE,
SPLITAWD.EXE. Executing LHA255.EXE will give you LHA.EXE (the
executable) plus a lot of other little files you don't need.
1) Take SPLITAWD.EXE and the "source" BIOS file, and in an MSDOS
window, execute "splitawd.exe 1018tc.004". This will give you
a directory called $MCTEMP, with 7 *.bin files. Watch the
screen output carefully, as the second line says "cpucode.exe"
and this corresponds to 00000001.bin. SPLITAWD will complain
that there is no LHA.EXE in the path, but this is desirable,
because if you give it a copy of LHA, it will hang :-(
2) Move the copy of LHA.EXE into $MCTEMP.
Execute "lha x 00000001.bin". This should put a copy of
cpucode.exe in $MCTEMP. Move cpucode.exe out of $MCTEMP and
forget about $MCTEMP.
3) In your current directory, put a copy of CTMC.EXE, cpucode.exe
and create a directory called "goodies".
Execute "set MCDATABASE=goodies"
Execute "CMTC.exe cpucode.exe /store"
The result should be a bunch of *.bin files will be placed
in the "goodies" directory. Move the 2KB file 06B41001.bin out
of the "goodies" directory and forget about "goodies". This is
your microcode file.
4) Everyone has a bootable floppy with a copy of aflash.exe and
1014S.003, right ? Well, place a copy of CTMC.EXE on there,
plus the microcode file 06B41001.bin. You are now ready for
your experiment. Since aflash is on the floppy, you can use
it after running CTMC, to either archive the new BIOS image,
and/or flash the "old" 1014S.003.
5) Install the new processor CPUID 06B4 on its adapter (the
microcode update process only works if the current processor
is a 06B4.) Boot the computer with the new processor with the floppy.
Using the CTMC.EXE and 06B41001.bin files on the floppy, execute:
"ctmc.exe 06B41001.bin /write"
The program displays the info in the microcode file and prompts
you for permission to write. For this to work, "Update BIOS" must
be enabled in the BIOS. CTMC will then suggest you reboot.
At this point, you can use aflash to backup the "new" BIOS, and
reflash 1014S.003 if you want. Otherwise, if you have a BIOS
Savior or a spare chip, just reboot.
This is all from memory, as my notes were all stored in the post I
lost. I hope I haven't gotten any of the workarounds I used wrong.
I also spent time mapping the BIOS. The summary of that work is,
CTMC does not update "cpucode.bin" in the BIOS. Instead, there is
a "volatile" area in the BIOS, which store DMI pool, and the like.
There are two 2KB "slots" to store the most recently used microcode
patches, in that volatile area. I guess normally, when the BIOS
finds a new processor, it pulls a microcode patch out of
cpucode.bin and stores it in one of the slots, at the same time
as it is updating the DMI and/or ESCD. CTMC, using the INT18 BIOS
call, is simply doing the same thing.
Note that you cannot use the new BIOS image as a donor for another
CBROM or CTMC transplant operation, as the microcode patch you've just
done, is stored in the volatile part of the EEPROM, and that part of
the EEPROM is not visible to either program. So, keep the 06B41001.bin
microcode file around, if you ever need to repeat this process.
Note: I haven't had the guts yet, to POST with the new image. As
you probably have options for recovering from a bad flash, I'll
leave it to you to try the find step -- reboot.
HTH,
Paul (cursing the huge post that is now lost...)
P2B
Sorry to hear about your crash, but many thanks for the truncated
version - saved me the time to figure all that out.
I actually started by trying to use CBROM to update cpucode.exe. None of
the versions I tried were able to read and display the Asus-format
microcode update file, but they at least extracted something that CTMC
could chop up into 2KB microcode files.
I then used a hex editor to add the 6B4 microcode to cpucode.exe,
according to the file format documentation here:
http://www.mrufer.ch/pc/asus_microcode.pdf
CTMC was able to parse the edited file, so I crossed my fingers and used
CBROM to inject it into the BIOS image and flashed the result.
The system still booted (!), but the BIOS couldn't find microcode for
*any* cpu now. I'm quite sure I edited cpucode.exe correctly, and CTMC
seemed to agree, so I suspect CBROM failed to install it properly.
I gave up on this route (for now, anyway), reflashed 1014beta3, and
followed your steps 4 and 5 to install 06B41001.bin.
That worked! The BIOS now recognises the 6B4 microcode update, but it
still won't boot a 6B4 PIII-S in slot 2 :-(
I think you are pretty safe to try rebooting your system with the 6B4
update, but if you'd prefer I'll repeat the process on a P2B-S and post
a tested BIOS image where you can download it.
Paul
<p...@sympatico.ca> wrote:
> Sorry to hear about your crash, but many thanks for the truncated
> version - saved me the time to figure all that out.
>
> I actually started by trying to use CBROM to update cpucode.exe. None of
> the versions I tried were able to read and display the Asus-format
> microcode update file, but they at least extracted something that CTMC
> could chop up into 2KB microcode files.
>
> I then used a hex editor to add the 6B4 microcode to cpucode.exe,
> according to the file format documentation here:
>
> http://www.mrufer.ch/pc/asus_microcode.pdf
>
> CTMC was able to parse the edited file, so I crossed my fingers and used
> CBROM to inject it into the BIOS image and flashed the result.
>
> The system still booted (!), but the BIOS couldn't find microcode for
> *any* cpu now. I'm quite sure I edited cpucode.exe correctly, and CTMC
> seemed to agree, so I suspect CBROM failed to install it properly.
>
> I gave up on this route (for now, anyway), reflashed 1014beta3, and
> followed your steps 4 and 5 to install 06B41001.bin.
>
> That worked! The BIOS now recognises the 6B4 microcode update, but it
> still won't boot a 6B4 PIII-S in slot 2 :-(
>
> I think you are pretty safe to try rebooting your system with the 6B4
> update, but if you'd prefer I'll repeat the process on a P2B-S and post
> a tested BIOS image where you can download it.
Maybe you could try out the "apple_rom" patcher. The thread describing it
is here:
I got the files here:
http://k6-2.narod.ru/award/patcher/real_microcodes.rar
http://k6-2.narod.ru/award/patcher/LHA.RAR
http://k6-2.narod.ru/award/patcher/CBROM.RAR
http://k6-2.narod.ru/award/patcher/bp-4rc_F.rar
I'm still looking for a command line decompresser, as I prefer them
to double clicking self extracting archives (if that is what these are).
Paul
David Maynard
Those are RAR compressed (I know, obvious). Similar to ZIP but with a
few additional features; in particular, automatic file spanning and
recombining. Most people use WinRAR (like WinZIP) for Windows but they
do have a DOS, command line, version. Linux and MacOS too.
Paul
Thanks! I'm RARing as we speak :-)
Paul
Paul
(Paul) wrote:
> Maybe you could try out the "apple_rom" patcher. The thread describing it
> is here:
>
>
http://www.wimsbios.com/phpBB2/viewtopic.php?t=3173&start=0&postdays=0&postorder=asc&highlight=microcode
>
> I got the files here:
> http://k6-2.narod.ru/award/patcher/real_microcodes.rar
> http://k6-2.narod.ru/award/patcher/LHA.RAR
> http://k6-2.narod.ru/award/patcher/CBROM.RAR
> http://k6-2.narod.ru/award/patcher/bp-4rc_F.rar
>
> I'm still looking for a command line decompresser, as I prefer them
> to double clicking self extracting archives (if that is what these are).
>
> Paul
This is a dead end. The rom.by patcher cannot parse the P2B-S 1014S.003
BIOS either. It is "bp-4rc_F" itself that quits with error 0012
"No more matching files". The author claims to understand Asus BIOS,
but it is just possible he is working with a moving target, and perhaps
has never tested against 1014S.003. The version of CBROM he uses is 2.07
and I threw a copy of 1.24C into the working directory instead and got
the same results. (1.24C doesn't understand the file either, but it only
trips up while parsing cpucode.exe).
Well, at least the CTMC recipe works.
Paul
David Maynard
It's very popular for sending large (or collections of) files via
newsgroups, or email, because it can recombine the separate RAR files
(limited message length) of a set in one stroke (plus they're
compressed, of course). A companion program, PAR, is a 'CRC'ish error
correction mechanism whereby one PAR can recreate any single corrupted
RAR out of the RAR set.
I.E.
file-part1.rar
file-part2.rar
file-part3.rar
file-part1.par
file-part1.p01
(the numbering convention can vary. E.g. you'll also see file.rar,
file.r01, file.r02, etc)
RAR will reconstruct whatever is the contents of parts 1 thru 3,
combined, in one fell swoop, including directory structure if it was so
created. If one of the parts is corrupted then PAR can reconstruct one
of them since there is one PAR part (the .p01 file. The .PAR file is a
small, not really needed, file listing the complete set of files for PAR
to use as a reference [or the user- tells you if you got all the RARs].)
If there were two PARs then any two corrupted RARs could be 'fixed', or
recreated from scratch. It too is automatic.
A 'usable' PAR file will be the same size as the RAR files (which are a
set size [the last one might be smaller if there's not enough data to fill])
An older 'check' mechanism was SFV, but that only told you if the files
were good or bad (so you could ask for a resend).
Typical use is to download all the (known) RARs, the .sfv, or preferably
.PAR, and then check the set. If using PAR, and an error is found, then
you can download one or more PARs, depending on how many the sender
provided, to repair what's broken or missing. If using SFV you have to
request a resend of the specific broken, or missing, part.
Not everyone uses either SFV or PAR though. Of course, RAR itself will
tell you if the file set is good, or not, when you try to decompress.