XP/W2k3 x86 on Modern Hardware

[George King]

Please continue here in original Win-Raid discussion. Thanks to WayBack machine - seems like many usefull informations are archived here!

 

Also thanks for topic archive to @iyut!

Download

https://www.upload.ee/files/13683020/WinXP_2003_x32_on_modern_hardware.rar.html

or

https://www.mediafire.com/file/7cs86b216kfj2rt/WinXP_2003_x32_on_modern_hardware.rar/file

Edited December 28, 2021 by George King

[Dietmar]

@George King

Can you repeat your last post to me,

because forum was closed before I can read it

Dietmar

PS: I notice, that the new Asus z690 boards can run Windows XP SP3 with full acpi support.

Anybody here, who tested another z690 board under XP?

 

 

 

Edited November 16, 2021 by Dietmar

[Mov AX, 0xDEAD]

Windows XP/2003 32-Bit on Modern Intel Hardware
[spoiler="Missing ACPI 2.0 support, BSODs 0xA5(...,  ...,  ...,  ...)"]
 Windows XP/2003 support only ACPI 1.0b syntax, last generation bioses uses ACPI 2.0+
 To support ACPI 2.0 syntax need replace acpi.sys and apply additional patches to avoid known BSODs :

...

 

Edited November 16, 2021 by Mov AX, 0xDEAD

[Dietmar]

 

Here is your content of spoiler

Edited November 20, 2021 by Dietmar

[infuscomus]

@Mov AX, 0xDEAD

What would be the best way to tell ntoskrnl extender to wait a few seconds in C code?

I want to see if waiting a few seconds might help the problematic 149C USB controller to start correctly.

like have an if statement at the start checking

if PCI\VEN_1022&DEV_149C is present

then wait 5 seconds

else continue as normal.

Edited November 23, 2021 by infuscomus

[infuscomus]

@Dietmar@George King

Do either of you know how to tell a driver to look for a specific device ID in C code?

[Mov AX, 0xDEAD]

On 11/23/2021 at 10:41 AM, infuscomus said:

if PCI\VEN_1022&DEV_149C is present

then wait 5 seconds

else continue as normal.

Hi infuscomus

In kernel mode you can use KeStallExecutionProcessor(), it hangs current kernel thread for requested time, but remain kernel threads still be executing. Anyway this will not help you with usb controller because you need to know from what moment you need to make pause. Usb3 driver do heavy reset of usb hardware(it waits some declared time after reset), also bios may has special acpi init code for controller. I recommend first to try any LiveCD with Win8.0 and Win8.1 to see how controller initialized with native OS and ported MS USB3 driver. 5 seconds is too much, usb controller has much smaller waiting times

 

[ExtremeGrief]

I have PCI\VEN_1022&DEV_149C, is there any fix to make it work?

[ExtremeGrief]

@Mov AX, 0xDEAD

@infuscomus

[infuscomus]

@ExtremeGrief

No fix is available yet unfortunately.

@Mov AX, 0xDEAD

I'll give KeStallExecutionProcessor() a try anyway since I know the driver works for this controller in XP by disabling and re-enabling it, it's just something peculiar at boot time that is making it not work.

To use KeStallExecutionProcessor() would putting this in DriverEntry of ntoskrn8.sys work?

    OldTime.QuadPart = Int32x32To64(1, -(10 * 1000 * 1000));
    Status = KeStallExecutionProcessor(                );
	

Windows 8.0 LiveCD - unmodded gives A5 BSOD - with patched acpi.sys it boots - 149C controller starts correctly.

Windows 8.0 LiveCD - no A5 BSOD - 149C controller starts correctly.

Edited November 28, 2021 by infuscomus

[Dietmar]

@infuscomus

From @Mov AX, 0xDEAD

"also bios may has special acpi init code for controller" , here 149C.

May be it is enough, to look at DSDT in a Bios from a board for AMD with 149C USB, that works, and replace it in DSDT

Dietmar

 

 

[ExtremeGrief]

39 minutes ago, Dietmar said:

@infuscomus

From @Mov AX, 0xDEAD

"also bios may has special acpi init code for controller" , here 149C.

May be it is enough, to look at DSDT in a Bios from a board for AMD with 149C USB, that works, and replace it in DSDT

Dietmar

 

 

I can give info for 149C or test cuz i want to get it working

[Dietmar]

I notice,

that

 EDIT: Mistake from me          

 

with all its settings for each USB port

is missed in "modern AMD DSDT" for USB

Dietmar

Edited November 28, 2021 by Dietmar

[infuscomus]

@Dietmar

In DSDT it may have a different name - To determine BIOS name under windows 10 device manager, select USB controller, go to details tab, select BIOS device name.

eg. - it could be something like

\_SB.PCI0.GPP2.PTXH

 

I'm skeptical that a DSDT modification will work, windows 7 has the same problem with the 149C controller, but something in the windows 8.0 ACPI driver makes it work at boot.

Nonetheless I can try DSDT modification anyway thanks to grub2 if I need to.

 

Edited November 28, 2021 by infuscomus

[infuscomus]

A reminder for myself -

DSDT GUID is: C118F50D-391D-45F4-B3D3-11BC931AA56D

[Dallen]

Hi @Dietmar

I’m trying to make a build around the z390 dark, but a lot of the references seem to be lost with the thread removal. Do you possibly still have the files/changes available for reference?

[Dietmar]

@Dallen

I have the z390 dark for myself with all drivers for XP.  Very fast board.

Only no Sound, so I use the Soundcard Audigy RX for XP SP3.

With the nice XP from Ramsey all works

https://www.zone94.com/downloads/software/operating-systems/123-windows-xp-professional-sp3-x86-integral-edition

Dietmar

[infuscomus]

@Mov AX, 0xDEAD

Over at the MDL forums another user has reported problems, this time now with an Intel USB controller.

I don't suppose you have any motivation to try and port over the 8.1 version of the USB 3.0 driver?

Edited December 1, 2021 by infuscomus

[iyut]

Hello colleagues,
here is an archive of forum's topic "Windows XP, 2003 x32 on modern hardware" (2018-2021) from win-raid.com, for the benefit of WinXP' admirers.

https://www.upload.ee/files/13683020/WinXP_2003_x32_on_modern_hardware.rar.html

EDIT:  above is a new link, with small corrections, and divided into 5 parts.  If one-page, consumes 190 mb of RAM in Firefox.
View in Firefox, because Chrome doesn't support spoilers.

(If you have downloaded the previous file, please delete, this one is better).

Edited December 2, 2021 by iyut

[nevelaev]

Hi to all I was absent a lot of time on win-raid. So, I lost a chain of events  Does somebody remember, maybe working driver for Bay Trail-T graphics under XP is exist?) I tried many versions from Ivy Bridge etc- no working one found =( 

Second question: except modern hardware Windows XP have another important problem for being used widely: some system libraries or modern environments like .Net Framework 4.5+, etc are not available for XP. If I have right another information at this time. Maybe there are some good news about this situation?) Thanx  

Edited December 1, 2021 by nevelaev

[iyut]

21 minutes ago, nevelaev said:

Second question

If programs require DotNet 4.5+, use Windows 7.

[nevelaev]

17 minutes ago, iyut said:

If programs require DotNet 4.5+, use Windows 7.

This is a simple and understandable solution  I meant maybe any solution exist for XP but a few people knows about this) 

[Mov AX, 0xDEAD]

14 hours ago, infuscomus said:

@Mov AX, 0xDEAD

Over at the MDL forums another user has reported problems, this time now with an Intel USB controller.

I don't suppose you have any motivation to try and port over the 8.1 version of the USB 3.0 driver?

I cannot reg on MDL due paranoid registration page (emal was banned/you are bot/ field  '29a0fc0c119ec7dee09d2b1af41f0a0e' was not recognised, wtf ?)

Yes, i still dont have motivation to port 8.1 usb3 driver because there is no evidence to prove that it will works with existing DSDT tables/WinXP ACPI driver. You wrote before:

Quote

Windows 8.0 LiveCD - no A5 BSOD - 149C controller starts correctly

So 8.0 usb3 driver works properly with native acpi 8.0 driver but not with WinXP/Win7 acpi driver ?

[Mov AX, 0xDEAD]

13 hours ago, iyut said:

Hello colleagues,
here is an archive of forum's topic "Windows XP, 2003 x32 on modern hardware" (2018-2021) from win-raid.com, for the benefit of WinXP' admirers.

iyut, thanks a lot !

[infuscomus]

1 hour ago, Mov AX, 0xDEAD said:

So 8.0 usb3 driver works properly with native acpi 8.0 driver but not with WinXP/Win7 acpi driver ?

@Mov AX, 0xDEAD

I made a typo there - only Windows 8.1 worked without patching. Windows 8.0 gave an A5 BSOD, after an acpi.sys patch from @Dietmar it was able to boot and USB 3.0 worked correctly.

I had not considered booting Windows 8.0 with XP and 7 acpi.sys - I will test this and report my findings back to you.