maxtherabbit wrote on 2023-05-18, 19:06:

My slocket8 P3B-F system is also still going strong!

Zerthimon wrote on 2023-05-18, 20:54:

I also run two p3b-bf 's with your awesome bios. Thanks a lot for your work @DenizOezmen. Any chance you would create a modded bios for p2b-ds ?

Great! 😀

I'l probably not be starting a mod on a completely different build any time soon. Have you checked out BIOS patcher from rom.by? It should be able to make various improvements to the P2B-DS BIOS.

DenizOezmen wrote on 2023-05-21, 20:20:

Have you checked out BIOS patcher from rom.by? It should be able to make various improvements to the P2B-DS BIOS.

There is one obscure trade-off when using the BIOS patcher from rom.by though: The 8x8 font that is stored in the BIOS ROM (located at F000:FA6E) gets removed and some DOS program uses it in a hardcoded way. Although the patcher has a command line switch to prevent the removal, it doesn't work properly.

An easy way to reproduce it is by running the first StarMines game which uses the font to display text.

Would you please consider adding Via C3 support to the Asus P2B BIOS (1014.003)?

I would like to try running an Ezra or Ezra-T in it.

I have attached two variants of the BIOS from here.

Thanks

feature request 😀
Is it possible to replace the anti-virus module with some kind of boot manager with the ability to boot from USB?
For example, PLOP.
https://www.plop.at/en/bootmanager/index.html

Gmlb256 wrote on 2023-05-21, 23:10:

There is one obscure trade-off when using the BIOS patcher from rom.by though: The 8x8 font that is stored in the BIOS ROM (located at F000:FA6E) gets removed and some DOS program uses it in a hardcoded way. Although the patcher has a command line switch to prevent the removal, it doesn't work properly.

Interesting, didn't know that. Thanks!

mockingbird wrote on 2023-06-13, 04:19:

Would you please consider adding Via C3 support to the Asus P2B BIOS (1014.003)?

I would like to try running an Ezra or Ezra-T in it.

Hmm. I tried booting two C3 chips in a P2B-F (since I assume this board would behave similarly). The board would post right away with an Ezra CPU, but would hang with a Nehemiah. (Not sure what the reason for the latter is yet, though it seems to occur somewhere around MTRR/cache configuration.)

Have you tried your CPU in the P2B? Is there anything specifically amiss (aside from the fact that the POST screen will call it a "Pentium III")?

shevalier wrote on 2023-06-16, 05:39:

feature request :-) Is it possible to replace the anti-virus module with some kind of boot manager with the ability to boot from […]
Show full quote

feature request 😀
Is it possible to replace the anti-virus module with some kind of boot manager with the ability to boot from USB?
For example, PLOP.
https://www.plop.at/en/bootmanager/index.html

Interesting idea. 😀

In general, this should already be possible; fellow user darry has done someting similar with XT-IDE on the P3B-F. We replaced the Symbios option ROM in that case. (Haven't tried the anti-virus module yet.)

From a quick test, this method also seems to work for PLOP. I had to use the "-forceINT" parameter when configuring the ROM module, otherwise the boot manager wouldn't initialize:

1plpbtrom.exe -forceINT plpbtrom.bin plpbt.rom

I didn't try actually booting anything from the boot manager screen, though.

Here's the test sample for the P3B-F CUBX-E/L. (You need to switch the Symbios option from "Auto" to "Enabled".)

DenizOezmen wrote on 2023-06-16, 09:18:

Interesting idea. 😀

Honestly stolen from these guys
http://www.x86-secret.com/articles/tweak/i815twken.htm

DenizOezmen wrote on 2023-06-16, 09:18:

Here's the test sample for the P3B-F. (You need to switch the Symbios option from "Auto" to "Enabled".)

I have CUBX-L which is in the process of being adapted to work with Tualatin.

I don't have working hardware to test yet 🙁

DenizOezmen wrote on 2023-06-16, 09:18:

Hmm. I tried booting two C3 chips in a P2B-F (since I assume this board would behave similarly). The board would post right away with an Ezra CPU, but would hang with a Nehemiah. (Not sure what the reason for the latter is yet, though it seems to occur somewhere around MTRR/cache configuration.)

Have you tried your CPU in the P2B? Is there anything specifically amiss (aside from the fact that the POST screen will call it a "Pentium III")?

Yes, it boots perfectly fine and shows as you say like a PIII. The issue is that when I try to overclock it to 133Mhz FSB, the speed shows incorrectly.

Also, I tested it with my P3B-F, it doesn't work at all (or barely works, I only got it to post once, 99% of the time my diagnostic card shows "----". ) I think the P3B-F has issues initializing it. Thanks

shevalier wrote on 2023-06-16, 10:38:

http://www.x86-secret.com/articles/tweak/i815twken.htm

Didn't know of that, thanks!

shevalier wrote on 2023-06-16, 10:38:

I have CUBX-L which is in the process of being adapted to work with Tualatin.

I've replaced the version above with a build for CUBX-E/L, though untested. Hope the mod goes well.

mockingbird wrote on 2023-06-16, 10:42:

Yes, it boots perfectly fine and shows as you say like a PIII. The issue is that when I try to overclock it to 133Mhz FSB, the speed shows incorrectly.

Is the speed really way off or "just" by a few MHz (something like 797 instead of 800)? If it's the latter, the BIOS probably just fails to prettify it. (The logic of version 4.51 is pretty old and messy compared to the more recent 6.0 builds.)

mockingbird wrote on 2023-06-16, 10:42:

Also, I tested it with my P3B-F, it doesn't work at all (or barely works, I only got it to post once, 99% of the time my diagnostic card shows "----". ) I think the P3B-F has issues initializing it. Thanks

Sounds like it. That's either a hardware problem, or the initialization fails so early that the BIOS didn't get the chance to output the first POST code. (I'm not yet aware of CPU-specific code that early in the boot process, though.)

DenizOezmen wrote on 2023-06-16, 09:18:

From a quick test, this method also seems to work for PLOP. I had to use the "-forceINT" parameter when configuring the ROM mod […]
Show full quote

From a quick test, this method also seems to work for PLOP. I had to use the "-forceINT" parameter when configuring the ROM module, otherwise the boot manager wouldn't initialize:

Copy code to clipboard

1plpbtrom.exe -forceINT plpbtrom.bin plpbt.rom

I didn't try actually booting anything from the boot manager screen, though.

Allow me to report 😀
As a proof of concept - very good.
PLOP loads Windows 2000 and DOS from USB and HDD.

For some reason, It`s didn’t identify a 256MB flash drive, 2GB is just perfect.

Of the shortcomings.
- Non-obvious option in BIOS to enable PLOP.
- Graphic mode with a star field on the background - knocks out of the style of those times. I liked text better.
- Font - better from BIOS.
- Loading PLOP by default, cancel - by pressing the button. It would be better the other way around.

And so - very even nothing. Exactly what is missing in the old motherboard.
Thank you!

DOS partition size is correct, maxes out at 2GB per partition for total of four partitions.

Windows 2000 supports larger partitions as well.

Cheers,

pentiumspeed wrote on 2023-06-22, 16:17:

DOS partition size is correct, maxes out at 2GB per partition for total of four partitions.

Windows 2000 supports larger partitions as well.

Cheers,

Undecided ancient flash drive 256 MB FAT32.
Which was been read correct in Windows 2000 itself.
But, this is most likely a nuance of PLOP USB driver.

And this is definitely not a question for @DenizOezmen, who is not the author of PLOP.

Then this is not correct, for USB stick, if 256MB, DOS should be slightly less than 256MB partition, replace the USB stick. You should be running newer USB sticks, get new usb 3.0 stick and set it up for DOS bootable partition. It will work properly even on vintage computers unless there is another compatibility issue with large capacity even partitioned to 2GB for one DOS partition.

Also, ancient USB sticks is less compatible too even on the vintage computers. Back in the day, it was a driver because I remember that when I had a 32MB USB stick and was not quite compatible.

Cheers

DenizOezmen
You can check the BIOS bug from ASUS, at least it's in the official beta for CUBX-L (and not related to your mod).

If you disable "Post on boot", then instead of quickly determining the memory size, a cyclic check of the Memory test occurs until you press "esc".
If you press esc, then the choice of booting from ASUS boot maneger appears instead of the normal boot.
PS. I didn’t even know that there was a built-in boot manager, and I couldn’t determine which key to call it.

DenizOezmen wrote on 2019-09-23, 18:57:

Hi everybody, […]
Show full quote

Hi everybody,

recently I toyed around with an ASUS P3B-F, and obviously I came across the modified "Noa" BIOS for this board. While effective, I disliked the way some of the changes were done, so I decided to re-do everything (and add a little bit more). Maybe somebody else can make use of it. The mod is based on the last official beta BIOS 1008.004, has been ported to some other ASUS boards with similar BIOS versions and includes the following changes:

• support of multipliers 12.5x through 16.0x on the "Advanced" tab of the BIOS Setup Utility, including display of suggested CPU speeds for standard FSB speeds
• support of displaying normalized frequencies at bootup for multipliers 12.5x through 16.0x at standard FSB speeds
• fix for the display of processor brand names at bootup, including support for Pentium III-S processors
• fix for the display of frequency suffixes (A, E, B, EB) at bootup according to Intel datasheets
• support of additional voltage ranges to suppress Hardware Monitor errors at bootup
• updated processor microcodes
• fixed display of multiplier 2.0x for processors with Klamath core
• unlocked core voltages for undervolting
• added Setup Utility "Enable" option for Symbios SCSI BIOS to force loading the ROM module even if no matching adapter is present
• CUBX: updated CMD Ultra66 ROM to version 1.9.16
• CUBX-E/L: updated Promise Ultra100 ROM to version 2.20.0.14
• POST fixes for Pentium Pro processors
• fix for the "Processor Version" and "Processor Family" fields in SMBIOS
• support for monitoring slow-spinning fans in Hardware Monitor (down to ca. 700 RPM, warnings at ca. 900 RPM)
• experimental support for Via C3/Cyrix III processors
• fixed Hardware Monitor errors and erroneous readings ("-6.14V") when using modern PSUs without a -5 V rail

Builds:

1008cu004_20230107.zip
1008cul004_20230107.zip
p3bf1084_20230107.zip
pv133132_20230107.zip
pv4x1064_20230107.zip

Ports for P3C-E, CUV4X-D and CUV4X-E here.

All BIOSes seem to work fine here. Still, make sure you have a working crisis recovery procedure before attempting to use them.

Edit 2019-10-30: Fixed suffix for Celeron 900; ported changes to ASUS CUBX-E/L, beta revision 1008.004.
Edit 2020-05-12: Fixed display of multiplier 2.0x for Klamath processors on P3B-F.
Edit 2020-05-17: Enabled selection of VCore voltages lower than stock voltage for P3B-F.
Edit 2020-05-20: Fixed a bug that caused the VCore voltage menu to stay hidden.
Edit 2o20-05-21: Added "Enable" option for Symbios SCSI BIOS on P3B-F.
Edit 2020-05-30: Uploaded latest versions for P3V4X and CUBX-E/L after testing.
Edit 2020-06-18: Uplodad latest version for CUBX after testing.
Edit 2021-06-23: Incorporated POST fixes for Pentium Pro processors on P3B-F.
Edit 2021-07-11: Uploaded latest versions with SMBIOS fixes and (untested) Pentium Pro fixes for P3V4X.
Edit 2022-02-05: Uploaded latest versions with slow fan support.
Edit 2022-02-20: Uploaded versions with experimental Via C3 support, fix for Katmai voltage monitor error and port for P3V133
Edit 2022-06-25: Disabled CPU temperature display in Hardware Monitor for Via CPUs; made upper VCore limit for CPUs requesting 1.50 V dependent on CPU core (1.70 V for Tualatin, 1.80 V for other cores)
Edit 2022-12-04: Unlocked all lower VCore voltages, added ports for P3C-E.
Edit 2022-12-11: Uploaded versions with -5 V rail fix; the boards should now play nice with modern PSUs without having to ignore the -5 V monitoring.
Edit 2023-01-07: Temporarily removed mods until VCore limit bug has been resolved.
Edit 2023-01-12: Restored mods.
Edit 2023-03-26: Added port for CUV4X-D.
Edit 2023-03-30: Added port for CUV4X-E.

How do you get the BIOS installed? I've tried AFLASH from Asus but no dice.

shevalier wrote on 2023-06-22, 13:47:

Allow me to report 😀

Cool, thanks. So it's just a matter of finding the perfect boot manager. 😉

shevalier wrote on 2023-06-26, 05:43:

You can check the BIOS bug from ASUS, at least it's in the official beta for CUBX-L (and not related to your mod). […]
Show full quote

You can check the BIOS bug from ASUS, at least it's in the official beta for CUBX-L (and not related to your mod).

If you disable "Post on boot", then instead of quickly determining the memory size, a cyclic check of the Memory test occurs until you press "esc".

If you press esc, then the choice of booting from ASUS boot maneger appears instead of the normal boot.
PS. I didn’t even know that there was a built-in boot manager, and I couldn’t determine which key to call it.

You mean "Quick Power On Self Test"?

Interesting, I wasn't aware of that either.
Seems like the key to invoke it is indeed ESC. This also works if quick POST is enabled, but the time slot is naturally shorter.

I can see where the boot menu is called and which flag gets checked, but I don't yet see where it gets set.

Edit: Okay, found it. It's located in the default Int 09h handler. One possibility would be to try and mask the interrupt during the memory test, but since it handles other scan codes as well, I'm not sure whether this might have adverse effects.

A second possibility would be to simply remap the boot menu to a different key. I tested it with F8 (since modern AMI BIOSes like to use this for the boot menu), and this seems to work. Any other ideas?

By the way: The memory "check" during quick POST doesn't seem to do anything. The system knows the memory size beforehand and just counts towards it in 256 K blocks. So I guess this is just there to look cool(?) ...

Geeknik wrote on 2023-06-30, 10:28:

How do you get the BIOS installed? I've tried AFLASH from Asus but no dice.

AFLASH should usually do the trick. Is there any specific error message?

I did get the 1008f.004 BIOS updated. No errors. Its just all the other stuff that came with the zip. How do you use that stuff? I tried the *.exe(s) in the TOOL folder, but most said they were WIN32 applications in DOS and in Windows didn't do anything I could see.

DenizOezmen wrote on 2023-07-01, 21:30:

I tested it with F8 (since modern AMI BIOSes like to use this for the boot menu), and this seems to work. Any other ideas?

Please, not F8.
AMI developers should burn in hell for this.
F8 is safe mode for older versions of Windows.
F11, like of Gigabyte, would be great - it does not intersect with anything.

DenizOezmen wrote on 2023-07-01, 21:30:

By the way: The memory "check" during quick POST doesn't seem to do anything. The system knows the memory size beforehand and just counts towards it in 256 K blocks. So I guess this is just there to look cool(?) ...

BX does not always determine the correct memory size with 4 occupied slots and a 133MHz bus.
I believe that "memory testing" in one pass can be useful when "quick POST" is disabled and disabled completely, in another.
Memory is best tested from the OS with other programs.

And please bring back the voltage regulation for Tualatin...

Geeknik wrote on 2023-06-30, 10:28:

How do you get the BIOS installed? I've tried AFLASH from Asus but no dice.

http://web.archive.org/web/20011204121509/htt … usupd32806e.exe

Geeknik wrote on 2023-07-02, 08:17:

I did get the 1008f.004 BIOS updated. No errors. Its just all the other stuff that came with the zip. How do you use that stuff? I tried the *.exe(s) in the TOOL folder, but most said they were WIN32 applications in DOS and in Windows didn't do anything I could see.

Its a semiautomatic builds and patch applying
See the .cmd file
@echo off

set BUILD_TARGET=TARGET_CUBXE
set BUILD_SYSTEM_ROM=stn.bin
set BUILD_IMAGE=1008cul.004
set BUILD_ARCHIVE_PREFIX=1008cul004
for /f "usebackq tokens=1,2 delims==" %%i in (`wmic os get LocalDateTime /value 2^>nul`) do if /%%i/==/LocalDateTime/ set BUILD_DATE=%%j
set BUILD_DATE=%BUILD_DATE:~0,8%

set NASM_LOCATION="C:\Program Files (x86)\NASM\nasm.exe"
set WINRAR_LOCATION="C:\Program Files\WinRAR\WinRAR.exe"

Geeknik wrote on 2023-07-02, 08:17:

Its just all the other stuff that came with the zip. How do you use that stuff? I tried the *.exe(s) in the TOOL folder, but most said they were WIN32 applications in DOS and in Windows didn't do anything I could see.

Ah. As shevalier already wrote, that's for building the BIOS mod from source, in case anyone wants to toy around with that. If you just want to use the modified BIOS, you can ignore all of it.

shevalier wrote on 2023-07-02, 08:52:

F11, like of Gigabyte, would be great - it does not intersect with anything.

I'll try that later today. Note that F11 is outside of the original XT key set (up to F10), so this might not work with every keyboard.

shevalier wrote on 2023-07-02, 08:52:

And please bring back the voltage regulation for Tualatin...

What's missing there?

DenizOezmen wrote on 2023-07-02, 09:52:

up to F10

IIRC, it's HP key for choose boot device.
I would say it's a good choice.

DenizOezmen wrote on 2023-07-02, 09:52:

shevalier wrote on 2023-07-02, 08:52:

And please bring back the voltage regulation for Tualatin...

What's missing there?

From first topic page
1008cul004_20230107.zip
File size 536.53 KiB
Edit 2023-01-07: Temporarily removed mods until VCore limit bug has been resolved.

Version "Edit 2023-01-12: Restored mods." I did not find.
With the file from the first page, there is a change in voltage, but monitoring shows that it remains the same (1.45V)