Dietmar

@casdanic My experience with Gigabyte motherboards is: When they introduce a new board, its Bios is faulty as much as possible. But with time, when some newer Bios versions are out for the same board, it becomes really nice. So, when you are not afraid of flashing Bios, you can make a try also, if there is a newer Bios for your board. But in past I kill a board from Asus, just with updating its original Bios to a newer version. For me, this story with much better Bios now is true for the Gigabyte z690 UD DDR4, which at the beginning even dont recognice any CD-Rom, no working Serial Port etc. Now it is really good and I make all Debug via its Serial Port of XP SP2 bit64 Dietmar

@casdanic I would not buy another motherboard. I am quite sure, that your problem can be solved, may be just when you set up a brandnew w2k3 bit32 install Dietmar

@casdanic Now I get everything to work for w2k3 bit32 on the Asrock z690 Extreme with 12900k cpu. No conflict in Device Manager. USB3 via @daniel_k mod for w2k3 of its *.inf Dietmar https://ufile.io/isho2boc

@casdanic Yepp, hal.dll works, I test with timer, but intelppm.sys not. I install W2k3 bit32 R2, USA version. acpi.sys works. It shows more than 75 Watt doing nothing on the Asrock Extreme z690 board. With the intelppm.sys modded from XP SP3 , this goes down to 25 Watt, but it is still much bigger than it should be. This means, that to change the 2 offsets to C1 states in intelppm.sys is not enough. And also, that the powermanagment in W2k3 bit32 is other than in XP SP3. I get everything to work, I have GT 730 graphikcard, no conflict with any device. USB3 I do not get to work, only xhci, HUB3 not. I tried @daniel_k USB3 AMD 145 driver, the first USB3 driver from @Mov AX, 0xDEAD and the very last USB3 driver from Ramsey, which gives DEADBEEF Bsod, I think because incompatible ntoskrnl.exe from w2k3 bit32 Dietmar

I just notice, that the variant of hal.dll 5.2.3790.3959 (srv03_sp2_rtm.070216-1710) and its ntoskrnl.exe is very much different from XP SP3 version, because it has dependencies on ExiReleaseFastMutex ExiAcquireFastMutex ExiTryToAcquireFastMutex KiCheckForSListAddress which the XP SP3 version does not have Dietmar

@un user Please write here the exact Bsod with XP SP3 acpi.sys version with OS fakes from 21 December Dietmar PS: You have post. Please write everything here, what happens.

@casdanic I succeed to patch intelppm.sys for Win2k3 bit32 for to overcome the big power consumption, by hand only using Winhex Dietmar https://ufile.io/zl1fu6c2 1. F:\win2003filesbit32\files\intelppmOriginal.sys: 40.960 Bytes 2. F:\win2003filesbit32\files\intelppmModd.sys: 40.960 Bytes Offsets: hexadez. 128: CE FE 129: 71 4C 4DBB: B4 90 4E5C: 60 90 4E5D: 2D 2C

@casdanic Please give me the exact(!) name of your motherboard, may be even versions number. The Bios then I fetch from Webside and I modd DSDT. At once we see, if it is the same conflict VGa<=> IOTR or other Dietmar PS: At the weekend I can set up the same W2K3 bit32 on an z690 board. And I make one try more to edit the intelppm.sys by hand. I am quite sure, that on one of this boards will also show the error with missing videocard. And this I can debug Dietmar

@casdanic To edit Bios DSDT in this case is easy for me. But always is a big risk, when you flash Bios. But may be, that @Mov AX, 0xDEAD has another idea, how to overcome this problem with conflict between IOTR and graphic card. For the acpi.sys for XP SP3 it is solved and also for XP SP2 bit 64 I think. For me it is a little bit strange, that this error can be seen now in W2k3 bit32 , it can be overcome with mod in this acpi.sys also Dietmar

@casdanic Does your motherboard have Bios Flashback? In this case you can use the patch for DSDT from @Mov AX, 0xDEAD Dietmar --- DSDT.dsl +++ DSDT.dsl @@ -1,36 +1,13 @@ AMI1 = ITA1 /* \ITA1 */ AMA1 = ITA1 /* \ITA1 */ AMI2 = ITA2 /* \ITA2 */ AMA2 = ITA2 /* \ITA2 */ AMI3 = ITA3 /* \ITA3 */ AMA3 = ITA3 /* \ITA3 */ - If ((ITS0 == One)) - { - ConcatenateResTemplate (Local0, BUF0, Local1) - Local0 = Local1 - } - - If ((ITS1 == One)) - { - ConcatenateResTemplate (Local0, BUF1, Local1) - Local0 = Local1 - } - - If ((ITS2 == One)) - { - ConcatenateResTemplate (Local0, BUF2, Local1) - Local0 = Local1 - } - - If ((ITS3 == One)) - { - ConcatenateResTemplate (Local0, BUF3, Local1) - Local0 = Local1 - } Return (Local0) } } } Scope (_SB.PCI0.LPCB)

@casdanic I remember, that also with XP SP3 we have the problem with missing video card (long time ago). After a patch from @Mov AX, 0xDEAD suddently the graphic card can be seen in Device Manager. But I dont know, how he did this Dietmar

@casdanic Here is the Timer modded hal.dll w2k bit32. Even the hack for your intelppm.sys is also very easy, I do not succeed with it with my Ida version, see post above Dietmar https://ufile.io/kex4sohj

@un user Here is the XP SP3 acpi.sys version with OS fakes from 21 December. This is the one that Ramsey uses with its *.pdb. Without OS fakes I can do later, I need to save the files in whole folder Acpi with all new entries before I can do this Dietmar https://ufile.io/hfxc70iv

@Mov AX, 0xDEAD The hal.dll w2k3 bit32 from @casdanic I succeed to modd for to use always the hardware clock. But for the the intelppm.sys I dont know, how with Ida Pro to write back the changes to the original file in _InitializeAcpi2IoSpaceCstates offsets AcpiC2Idle and AcpiC3ArbdisIdle to offset AcpiC1Idle each. When I click with Ida offset AcpiC2Idle , how can I change this function to offset AcpiC1Idle and the same for offset AcpiC3ArbdisIdle to offset AcpiC1Idle ? Is it via right click, manual, enter new string? When I do this, in Ida the name is changed to AcpiC1Idle but nothing in real is changed for example in Hex view with Ida Dietmar

@Dibya 2030 is not a joke from me. Most models say, that in 2030 AI will be more powerful than any human. When you read about chatGPT (now with v3.5 but the same for coming 4.0 this year), the restriction not to save any contents from chat is only for safety reasons Dietmar

@casdanic For industrial production I would ONLY choose the x58 platform. I have such a lot of different motherboards here and this is the only platform together with i920 cpu, that does not crash a single time in 3 years. Stable under XP SP3 (win2k3 bit32 I do not test) is also the Asrock z690 Extreme board. I have 6 z690 boards and I test them all like crazy. No crash only with the Asrock z690 Extreme board with 12900k cpu Dietmar

@casdanic This may be a problem of settings in Bios. You have to decide, which videocard you want to use. And this acpi.sys fakes the OS-version. This means, you see all the devices, which you also will see in win10. Some you can just disable. I dont have an intelppm.sys and hal.dll from NT 5.2. Can you send me this files? Then I check, if there is any difference to the XP SP3 version. In any case you need a modded intelppm.sys and hal.dll or you run into Power and timer- problems Dietmar

@casdanic Yes, this will work. Most easy is, to install the Sata driver from Kai on your B250 compi. Then you can copy at once everything to the 1200 board Dietmar

@casdanic Take this Sata driver from Kai Schtrom. When you have a newer storport.sys on your install of Win2003 bit32, you dont need the storport.sys in this package. Use Nlite for to integrate the Sata driver Dietmar https://ufile.io/ent598bk

@casdanic Any way your post is very interesting, because it means, that the Source Code for XP SP1 is better than for any win2003 Dietmar

@casdanic I never tried this. But make a try with this one, just fresh build and hacked Dietmar https://ufile.io/oo7ihfjy

@K4sum1 You have post Dietmar

@K4sum1 The Bsod, that you show picture from, cant be reached any more with my acpi.sys because it is cancelled via 90 90 Dietmar

@K4sum1 This is the only place, from where DDB was called. Are you sure, that this is the new acpi.sys? May be, that the System Restore put back the old acpi.sys in its place Dietmar

@K4sum1 Here is acpi hack DDB "8" for XP SP2 bit64 Dietmar https://ufile.io/c2bc9ntv