Compiling ACPI v2.0 driver for Windows XP SP3 and Windows 2003 SP2 (x32/x64)

[Mov AX, 0xDEAD]

8 hours ago, Dietmar said:

I get this Bsod also 1 time with debug acpi.sys, debug ntoskrnl and debug hal.

I think, the information is not enough, so I try the loong version also )

@Dietmar

yes, this lite log doesn't point to exact problem dsdt code:

Quote

FFFFFADCE3E8E760 5d071100 ACPIBuildProcessDevicePhasePr2: Status = 00000000
FFFFFADCE3E8E760 5d071100 ACPIBuildProcessDeviceGenericEval: Phase12 Status = 00000000
FFFFFADCE3E8E760 5d071100 (0x00000000): ACPIDeviceInternalDelayedDeviceRequest - Transition to D0
FFFFFADCE3E8E760 5d071100 ACPIBuildProcessDevicePhasePsc: Status = 00000103
FFFFFADCE3E8E470 ACPIBuildProcessDevicePhaseUid: Status = 00000103

*** Fatal System Error: 0x000000a5
                       (0x0000000000000003,0xFFFFFADCE3F2CC30,0xFFFFFFFFC0000034,0x000000004449485F)

Break instruction exception - code 80000003 (first chance)

need full log, but with COM speed it can be very long

!amli set spewon verboseon logon traceon

[Dietmar]

@Mov AX, 0xDEAD

I am in the still running Debugger.

*.txt is now about 50 Mbyte.

Endless loop with

HAL: RTC interrupt flag is not cleared by first read.
| RTC Status Register C = 0xc0

 

Any idea what I can put in commandline of Windbg, for to check or overcome this?

Dietmar

[Dietmar]

@Mov AX, 0xDEAD

Before Windbg reaches a Breakpoint, telling this, I think because of this endless loop

 

AMLI:| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | HAL: RTC interrupt flag is not cleared by first read.
| RTC Status Register C = 0xc0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | HAL: RTC interrupt flag is not cleared by first read.
| RTC Status Register C = 0xc0
| | | | | HAL: RTC interrupt flag is not cleared by first read.
| RTC Status Register C = 0xc0
| | | | | HAL: RTC interrupt flag is not cleared by first read.
| RTC Status Register C = 0xc0
| | | | | | | | | | HAL: RTC interrupt flag is not cleared by first read.
| RTC Status Register C = 0xc0
| | | | | HAL: RTC interrupt flag is not cleared by first read.
| RTC Status Register C = 0xc0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ValidateTarget=0 (pdataTarget=e3a32338)
=0x0000000000000014HAL: RTC interrupt flag is not cleared by first read.
)RTC Status Register C = 0xc0

fffffadce408d855: })
fffffadce408f4c0: }HAL: RTC interrupt flag is not cleared by first read.
RTC Status Register C = 0xc0
*** DPC execution time exceeds system limit
    This is NOT a break in update time
    This is a BUG in a DPC routine
    Perform a stack trace to find the culprit
Break instruction exception - code 80000003 (first chance)
nt!DbgBreakPoint:
fffff800`011a02c0 cc              int     3

 

 

 

 

[Dietmar]

@Mov AX, 0xDEAD

Here is the loong log file (until now).

I am still in the running Windbg but dont know what to do.

It looks, as if it hangs because of xhci (USB3) definition in DSDT

Dietmar

https://ufile.io/rwe6pkdz

 

Edited January 1, 2023 by Dietmar

[Mov AX, 0xDEAD]

6 minutes ago, Dietmar said:

@Mov AX, 0xDEAD

Before Windbg reaches a Breakpoint, telling this, I think because of this endless loop

AMLI:| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | HAL: RTC interrupt flag is not cleared by first read.
| RTC Status Register C = 0xc0

these messages are from checked hal.dll

6 minutes ago, Dietmar said:

*** DPC execution time exceeds system limit
    This is NOT a break in update time
    This is a BUG in a DPC routine
    Perform a stack trace to find the culprit

i think this is OK because slow COM connectiuon

[Mov AX, 0xDEAD]

I stuck with kernel debugging XP/W2003 x64 on VirtualBox, WinDBG hangs on virtual com1 port connection.

VirtualKD connection also useless because it doesnt work with VBoxHardenedLoader required for custom dsdt/ssdt tables.

QEMU also bad, it doesnt have good documentation for win32 platform,  i dont know how to enable virtual com port, there is a lot of options and nothing to works

Last hope is VMWare and VirtualPC...

[Dietmar]

@Mov AX, 0xDEAD

I can edit the DSDT (for example kill all USB) as much as I want from this Gigabyte UD z690 DDR4 board,

because it has Bios Flashback.

Here is its DSDT

Dietmar

https://ufile.io/foslv00u

Edited January 1, 2023 by Dietmar

[George King]

5 minutes ago, Mov AX, 0xDEAD said:

I stuck with kernel debugging XP/W2003 x64 on VirtualBox, WinDBG hangs on virtual com1 port connection.

VirtualKD connection also useless because it doesnt work with VBoxHardenedLoader required for custom dsdt/ssdt tables.

QEMU also bad, it doesnt have good documentation for win32 platform,  i dont know how to enable virtual com port, there is a lot of options and nothing to works

Last hope is VMWare and VirtualPC...

@UsefulAGKHelper You managed VirtualBox and XP x64 debugging right?

[reboot12]

@Mov AX, 0xDEAD

PC1: Try a virtual machine QEMU (virt-manager) on Linux e.g. Debian9 and Serial TCP net console Client Mode - set IP 192.168.0.2:4555

PC2: IP 192.168.0.2 install Fabulatech Serial Port Redirector, add virtual port COM Server eg. port TCP 4555, port COM12, speed 115200, Raw Data

Run Windbg -> Kernel Debug on COM12 115200

Edited January 1, 2023 by reboot12

[Dietmar]

@Mov AX, 0xDEAD

It looks, as if the acpi debug output itself hangs

Dietmar

PS: Now, compi crashes.

This is last output

https://ufile.io/9e7q4ho9

[Mov AX, 0xDEAD]

9 minutes ago, Dietmar said:

@Mov AX, 0xDEAD

It looks, as if the acpi debug output itself hangs

@Dietmar

At least we know v6666 way to fix it - mask ACPIBuildCompleteMustSucceed

of couse this way is not perfect, interpreter stopped on some dsdt code and never reach code after

[Dietmar]

@Mov AX, 0xDEAD

Hihi,

is there any chance to find out with Windbg, which device it is?

I can delete it in DSDT and check then, if all works

Dietmar

[Dietmar]

@Mov AX, 0xDEAD

Just now I make a new try:

I use the free ntoskrnl, the free hal.dll

but still the DEBUG acpi.sys

Dietmar

PS: Output is different.

[Mov AX, 0xDEAD]

29 minutes ago, Dietmar said:

@Mov AX, 0xDEAD

is there any chance to find out with Windbg, which device it is?

I can delete it in DSDT and check then, if all works

let's try with improved debug logging:

1) back to lite debug mode ed Kd_ACPI_Mask 0xFFFFFFFF
2) patch amliapi.c to partialy enable full debug lines:

Quote

      #ifdef DEBUGGER
        if (gDebugger.dwfDebugger & DBGF_VERBOSE_ON)
        {
            PRINTF(MODNAME ": %p: AsyncEvalObject(%s)\n",
                   KeGetCurrentThread(), GetObjectPath(pns));
        }
      #endif

to:

Quote

      //#ifdef DEBUGGER
        if (1)
        {
            PRINTF(MODNAME ": %p: AsyncEvalObject(%s)\n",
                   KeGetCurrentThread(), GetObjectPath(pns));
        }
      //#endif

Now you must get many "AMLI: xxxxx: AsyncEvalObject(\_SB.PC00.MC._ADR)" in lite debuglog, probably it will point to exact dsdt place before BSOD, as i understand we look for definition with bugged/missed _UID

Edited January 1, 2023 by Mov AX, 0xDEAD

[Dietmar]

@Mov AX, 0xDEAD

The Bsod happens mostly with the free version of acpi.sys of XP bit64.

Does it make sense, to edit amliapi.c the same way also for the free acpi.sys,

or is there another way to bring the free acpi.sys to give more output

Dietmar