Mov AX, 0xDEAD

Sorry, i don't read this thread because i feel NDSI6 can not be ported this way, main problem is netio.sys, it offers some new fuctionality

@Dietmar Connection() fix, change ParseField() to Tested on DSDT test code: SHS3 and FLD0/FLD1 are dummy with undefined type, if DSDT uses it lately in some math calculation, i don't know what happens

@Dietmar Same way as double devce definition was fixed, jump opcode pointer to next object Just wait fix in few days

Easy-Peasy way: skip GpioIo definition and assume SHS3 as memory field To: I know this is wrong, OperationRegion can be NULL or dummy, but we don't have any GPIO support in acpi.sys anyway GeneralPurposeIo = 0x08, missed in XP acpi.sys

Ignoring Connection broke access to SHS3 from other part of DSDT to parse this new opcode, Win8 acpi.sys has new ParseFieldConnection((_ctxt *pctxt, XXX *pConnection)) to store connection config, _FieldDesc has new fields, last field has variable len: I think these new fileds used in some field functions read/write/access to connection alias SHS3

Thanks, usual acpi.sys compiled with deep optimization, unknow mixed register/stack calling convection, hard to decompile

Hi Can someone share ACPI.SYS Windows 8.0. or Windows 8.1 x32 checked version ? (from checked installation ISO)

XP lacks Connection (0x02) opcode in ParseField() Basicaly GPIO will never(ACPI v5.x standart) work in XP, we will have only interpreter bypass fix

Thanks I think acpi.sys get crazy on Connection or GpioIo decoding, these two comands are "macro" and coded by some way DefField := FieldOp PkgLength NameString FieldFlags FieldList FieldOp := ExtOpPrefix 0x81 5B 81 FieldOp 33 PkgLength 47 50 4F 32 GPO2 01 FieldFlags (ByteAcc, NoLock, Preserve) 02 FieldList: FieldElement := NamedField | ReservedField | AccessField | ExtendedAccessField | ConnectField ReservedField := 0x00 PkgLength AccessField := 0x01 AccessType AccessAttrib ConnectField := <0x02 NameString> | <0x02 BufferData> ExtendedAccessField := 0x03 AccessType ExtendedAccessAttrib AccessLength 11 26 0A 23 8C 20 00 - Buffer ()

Raw table please, not decompiled

@Dietmar GPO2 and next object seems as AML garbage, need DSDT/SSDT dumps to clarify

@LanXing Manual to setup remote debug connection to save crashlog: https://github.com/MovAX0xDEAD/KDNET

A5(0x3,"TZ10", STATUS_OBJECT_NAME_NOT_FOUND/AMLIERR_OBJ_NOT_FOUND, "_INI") I don't see any problem with this code, so only remote debugging can say more this B450 has realtek LAN chip, it is possible, need time and wish :)

@mockingbird It is OK, service device for Windows 10 only, it doesnt require any driver Just ignore or install manually from Chipset Driver by Intel (i think need special adapted to XP modded version)

You can use COM port (slow, ports are rare on most motherboard) or special USB3 debug cable (rare, expensive, i made DIY from noname PCIE Riser 2$ cable) Unfortunately LAN card is only possibllity to make quick connection

As Daniel said, i don't include_OSI unlocker. Just to minimize possible negative effects of additional DSDT/SSDT code execution When will the "final" version be ready, i will add unlocker as option

everything this is important step, we don't need original big public folder

BINPLACE errors ? You didn't prepared environment setup, I do not use original compiler/libs/headers Repeat exactly steps at first post to prepare environment and files

I handle only busdrv\acpi\driver folder = acpi.sys, you compiled other dirs, they don't need for driver

try build /Zcegbw

@George King Sorry, these errors occur due mismatch my local W2003 tree with origin sources link updated to v4.2, i hope this is full sync

fixed to v4.1

And arblib.c too, problem is diff patch, it doesnt include two new files, just wait for fixed version

@Damnation Now it is finished, you can try compile w2003, but i'm not sure about setup/compiler match with original files If you got compilation error, it is OK, i will check what i missed to reproduce my local setup

This is BSOD I was asking about on prev page I think PDRC conflict with GPI0, PDRC has _UID, GPI0 doesn't have if GPHD =1, PDRC will set _HID = PNP0C02, but without any _UID Again, error in DSDT, no _UID method inside device Dietmar has similar board, his "otto2" logfile has same device definition mistake, but without BSOD cheap = good, thanks to asrock