how long can I hold a spin lock?

OSR_Community_User · October 11, 2004, 11:20pm

Is there any suggesting time? My driver seems to hang a HT system after
running a while(UP is fine), and I cannot ping it. It looks like a deadlock
to me. But I cannot break into the windbg. Any idea what to do?

Thanks

AH

Check out Election 2004 for up-to-date election news, plus voter tools and
more! http://special.msn.com/msn/election2004.armx

OSR_Community_User · October 12, 2004, 1:56am

You should hold onto a lock for the least amount of time as possible.
If you hold onto the lock too long, the user will see the affects. What
might be happening on your machine is that your driver is acquiring a
lock, but not releasing it and then return to the caller outside of your
driver. Turn on driver verifier and deadlock detection. If you think
this is what is going on, driver verifier should find it rather quickly.

d

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
Sent: Monday, October 11, 2004 8:18 PM
To: Windows System Software Devs Interest List
Subject: [ntdev] how long can I hold a spin lock?

Is there any suggesting time? My driver seems to hang a HT system after

running a while(UP is fine), and I cannot ping it. It looks like a
deadlock
to me. But I cannot break into the windbg. Any idea what to do?

Thanks

AH

Check out Election 2004 for up-to-date election news, plus voter tools
and
more! http://special.msn.com/msn/election2004.armx

Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256

You are currently subscribed to ntdev as: xxxxx@windows.microsoft.com
To unsubscribe send a blank email to xxxxx@lists.osr.com

OSR_Community_User · October 12, 2004, 9:34am

Here’s an interesting thread related to this:

http://www.osronline.com/lists_archive/ntdev/thread11347.html

0.75 second seems like a long time, I think I’ve read 10ms as a “good
practices” maximum somewhere around here.

Mat

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Doron Holan
Sent: Tuesday, October 12, 2004 1:51 AM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] how long can I hold a spin lock?

You should hold onto a lock for the least amount of time as possible.
If you hold onto the lock too long, the user will see the affects. What
might be happening on your machine is that your driver is acquiring a
lock, but not releasing it and then return to the caller outside of your
driver. Turn on driver verifier and deadlock detection. If you think
this is what is going on, driver verifier should find it rather quickly.

d

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
Sent: Monday, October 11, 2004 8:18 PM
To: Windows System Software Devs Interest List
Subject: [ntdev] how long can I hold a spin lock?

Is there any suggesting time? My driver seems to hang a HT system after

running a while(UP is fine), and I cannot ping it. It looks like a
deadlock
to me. But I cannot break into the windbg. Any idea what to do?

Thanks

AH

Check out Election 2004 for up-to-date election news, plus voter tools
and
more! http://special.msn.com/msn/election2004.armx

Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256

You are currently subscribed to ntdev as: xxxxx@windows.microsoft.com
To unsubscribe send a blank email to xxxxx@lists.osr.com

Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256

You are currently subscribed to ntdev as: unknown lmsubst tag argument: ‘’
To unsubscribe send a blank email to xxxxx@lists.osr.com

Jake_Oshins · October 12, 2004, 1:02pm

Good practices depend on what you’re doing with the machine. Today’s
machines tend to be asked to play audio. In a machine playing audio, I
would suggest a maximum well below a millisecond.

–
Jake Oshins
Windows Kernel Group

This posting is provided “AS IS” with no warranties, and confers no rights.

“Mathieu Routhier” wrote in message
news:xxxxx@ntdev…
> Here’s an interesting thread related to this:
>
> http://www.osronline.com/lists_archive/ntdev/thread11347.html
>
> 0.75 second seems like a long time, I think I’ve read 10ms as a “good
> practices” maximum somewhere around here.
>
> Mat
>
>
> -----Original Message-----
> From: xxxxx@lists.osr.com
> [mailto:xxxxx@lists.osr.com] On Behalf Of Doron Holan
> Sent: Tuesday, October 12, 2004 1:51 AM
> To: Windows System Software Devs Interest List
> Subject: RE: [ntdev] how long can I hold a spin lock?
>
> You should hold onto a lock for the least amount of time as possible.
> If you hold onto the lock too long, the user will see the affects. What
> might be happening on your machine is that your driver is acquiring a
> lock, but not releasing it and then return to the caller outside of your
> driver. Turn on driver verifier and deadlock detection. If you think
> this is what is going on, driver verifier should find it rather quickly.
>
> d
>
> -----Original Message-----
> From: xxxxx@lists.osr.com
> [mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> Sent: Monday, October 11, 2004 8:18 PM
> To: Windows System Software Devs Interest List
> Subject: [ntdev] how long can I hold a spin lock?
>
> Is there any suggesting time? My driver seems to hang a HT system after
>
> running a while(UP is fine), and I cannot ping it. It looks like a
> deadlock
> to me. But I cannot break into the windbg. Any idea what to do?
>
> Thanks
>
> AH
>
> _________________________________________________________________
> Check out Election 2004 for up-to-date election news, plus voter tools
> and
> more! http://special.msn.com/msn/election2004.armx
>
>
>
> —
> Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
>
> You are currently subscribed to ntdev as: xxxxx@windows.microsoft.com
> To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>
> —
> Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
>
> You are currently subscribed to ntdev as: unknown lmsubst tag argument: ‘’
> To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>
>

OSR_Community_User · October 12, 2004, 4:57pm

Thanks.

I turn on the driver verifier with deadlock detection, and I get a bug
check. Then I follow with “!deadlock” in the windbg, but I get “no deadlock
verifier issues”. How can I continue from there?

AH

From: “Doron Holan”
>Reply-To: “Windows System Software Devs Interest List”
>
>To: “Windows System Software Devs Interest List”
>Subject: RE: [ntdev] how long can I hold a spin lock?
>Date: Mon, 11 Oct 2004 22:51:09 -0700
>
>You should hold onto a lock for the least amount of time as possible.
>If you hold onto the lock too long, the user will see the affects. What
>might be happening on your machine is that your driver is acquiring a
>lock, but not releasing it and then return to the caller outside of your
>driver. Turn on driver verifier and deadlock detection. If you think
>this is what is going on, driver verifier should find it rather quickly.
>
>d
>
>-----Original Message-----
>From: xxxxx@lists.osr.com
>[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
>Sent: Monday, October 11, 2004 8:18 PM
>To: Windows System Software Devs Interest List
>Subject: [ntdev] how long can I hold a spin lock?
>
>Is there any suggesting time? My driver seems to hang a HT system after
>
>running a while(UP is fine), and I cannot ping it. It looks like a
>deadlock
>to me. But I cannot break into the windbg. Any idea what to do?
>
>Thanks
>
>AH
>
>
>Check out Election 2004 for up-to-date election news, plus voter tools
>and
>more! http://special.msn.com/msn/election2004.armx
>
>
>
>—
>Questions? First check the Kernel Driver FAQ at
>http://www.osronline.com/article.cfm?id=256
>
>You are currently subscribed to ntdev as: xxxxx@windows.microsoft.com
>To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>
>—
>Questions? First check the Kernel Driver FAQ at
>http://www.osronline.com/article.cfm?id=256
>
>You are currently subscribed to ntdev as: unknown lmsubst tag argument: ‘’
>To unsubscribe send a blank email to xxxxx@lists.osr.com

On the road to retirement? Check out MSN Life Events for advice on how to
get there! http://lifeevents.msn.com/category.aspx?cid=Retirement

OSR_Community_User · October 12, 2004, 6:00pm

Try debugging the crash - start with “!analyze -v”. Just because you
turn on an option and the system crashed you can’t assume that it will
crash because of the option you enabled.

You might also want to show us the bugcheck data. That always makes it
easier to suggest what you might have done wrong.

Regards,

Tony Mason
Consulting Partner
OSR Open Systems Resources, Inc.
http://www.osr.com

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
Sent: Tuesday, October 12, 2004 4:55 PM
To: ntdev redirect
Subject: RE: [ntdev] how long can I hold a spin lock?

Thanks.

I turn on the driver verifier with deadlock detection, and I get a bug
check. Then I follow with “!deadlock” in the windbg, but I get “no
deadlock
verifier issues”. How can I continue from there?

AH

From: “Doron Holan”
>Reply-To: “Windows System Software Devs Interest List”
>
>To: “Windows System Software Devs Interest List”
>Subject: RE: [ntdev] how long can I hold a spin lock?
>Date: Mon, 11 Oct 2004 22:51:09 -0700
>
>You should hold onto a lock for the least amount of time as possible.
>If you hold onto the lock too long, the user will see the affects.
What
>might be happening on your machine is that your driver is acquiring a
>lock, but not releasing it and then return to the caller outside of
your
>driver. Turn on driver verifier and deadlock detection. If you
think
>this is what is going on, driver verifier should find it rather
quickly.
>
>d
>
>-----Original Message-----
>From: xxxxx@lists.osr.com
>[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
>Sent: Monday, October 11, 2004 8:18 PM
>To: Windows System Software Devs Interest List
>Subject: [ntdev] how long can I hold a spin lock?
>
>Is there any suggesting time? My driver seems to hang a HT system
after
>
>running a while(UP is fine), and I cannot ping it. It looks like a
>deadlock
>to me. But I cannot break into the windbg. Any idea what to do?
>
>Thanks
>
>AH
>
>
>Check out Election 2004 for up-to-date election news, plus voter tools
>and
>more! http://special.msn.com/msn/election2004.armx
>
>
>
>—
>Questions? First check the Kernel Driver FAQ at
>http://www.osronline.com/article.cfm?id=256
>
>You are currently subscribed to ntdev as: xxxxx@windows.microsoft.com
>To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>
>—
>Questions? First check the Kernel Driver FAQ at
>http://www.osronline.com/article.cfm?id=256
>
>You are currently subscribed to ntdev as: unknown lmsubst tag argument:
‘’
>To unsubscribe send a blank email to xxxxx@lists.osr.com

On the road to retirement? Check out MSN Life Events for advice on how
to
get there! http://lifeevents.msn.com/category.aspx?cid=Retirement

—
Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256

You are currently subscribed to ntdev as: xxxxx@osr.com
To unsubscribe send a blank email to xxxxx@lists.osr.com

OSR_Community_User · October 12, 2004, 7:17pm

Thanks. Please see below. It is a HP/HT machine running XP Home.

*** Fatal System Error: 0x000000c4
(0x00000090,0xF870F120,0x00000000,0x00000000)

Break instruction exception - code 80000003 (first chance)

A fatal system error has occurred.
Debugger entered on first try; Bugcheck callbacks have not been invoked.

A fatal system error has occurred.

Connected to Windows XP 2600 x86 compatible target, ptr64 FALSE
Loading Kernel Symbols
…
Loading unloaded module list
…
Loading User Symbols
…
*******************************************************************************
*
*
* Bugcheck Analysis
*
*
*
*******************************************************************************

Use !analyze -v to get detailed debugging information.

BugCheck C4, {90, f870f120, 0, 0}

*** ERROR: Module load completed but symbols could not be loaded for
SAVRT.SYS
Probably caused by : SAVRT.SYS ( SAVRT+7dad )

Followup: MachineOwner

nt!RtlpBreakWithStatusInstruction:
8051d778 cc int 3
1: kd> !analyze -v
*******************************************************************************
*
*
* Bugcheck Analysis
*
*
*
*******************************************************************************

DRIVER_VERIFIER_DETECTED_VIOLATION (c4)
A device driver attempting to corrupt the system has been caught. This is
because the driver was specified in the registry as being suspect (by the
administrator) and the kernel has enabled substantial checking of this
driver.
If the driver attempts to corrupt the system, bugchecks 0xC4, 0xC1 and 0xA
will
be among the most commonly seen crashes.
Parameter 1 = 0x1000 … 0x1020 - deadlock verifier error codes.
Typically the code is 0x1001 (deadlock detected) and you can
issue a ‘!deadlock’ KD command to get more information.
Arguments:
Arg1: 00000090, A driver switched stacks. The current stack is neither a
thread
stack nor a DPC stack. Typically the driver doing this should be
on the stack obtained from `kb’ command.
Arg2: f870f120
Arg3: 00000000
Arg4: 00000000

Debugging Details:

BUGCHECK_STR: 0xc4_90

DEFAULT_BUCKET_ID: DRIVER_FAULT

LAST_CONTROL_TRANSFER: from 804f444e to 8051d778

STACK_TEXT:
824a693c 804f444e 00000003 824a6c6c 00000090
nt!RtlpBreakWithStatusInstruction
824a6988 804f4ed3 00000003 8276bd70 00000000 nt!KiBugCheckDebugBreak+0x19
824a6d54 804f5471 000000c4 00000090 f870f120 nt!KeBugCheck2+0x46d
824a6d74 8063ab77 000000c4 00000090 f870f120 nt!KeBugCheckEx+0x19
824a6d90 8063ae47 8063b3db 00000000 00000000
nt!ViDeadlockCheckStackLimits+0x6b
824a6d94 8063b3db 00000000 00000000 00000000 nt!ViDeadlockCanProceed+0x25
824a6db8 80538367 8276bd78 00000020 82675770
nt!VfDeadlockDeleteMemoryRange+0x11
824a6df4 8052618c 8276bd78 00000000 828add18 nt!ExFreePoolWithTag+0x95
824a6e24 805264e5 824a6e3c f00929ac 824a6ec0 nt!ExpFindCurrentThread+0x184
824a6e48 eecb5dad eece8028 00000000 eecb96ef
nt!ExAcquireResourceSharedLite+0x4f
WARNING: Stack unwind information not available. Following frames may be
wrong.
00000410 00009902 00240000 50e00024 1c0d1c0d SAVRT+0x7dad

FOLLOWUP_IP:
SAVRT+7dad
eecb5dad 84c0 test al,al

SYMBOL_STACK_INDEX: a

FOLLOWUP_NAME: MachineOwner

SYMBOL_NAME: SAVRT+7dad

MODULE_NAME: SAVRT

IMAGE_NAME: SAVRT.SYS

DEBUG_FLR_IMAGE_TIMESTAMP: 3d40d9c3

STACK_COMMAND: kb

BUCKET_ID: 0xc4_90_SAVRT+7dad

Followup: MachineOwner

1: kd> !deadlock

Resources: 1
Nodes: 0
Threads: 29

Nodes trimmed based on age 0.
Nodes trimmed based on count 0.
Analyze calls 0.
Maximum nodes searched 0.

No deadlock verifier issues.

From: “Tony Mason”
>Reply-To: “Windows System Software Devs Interest List”
>
>To: “Windows System Software Devs Interest List”
>Subject: RE: [ntdev] how long can I hold a spin lock?
>Date: Tue, 12 Oct 2004 17:54:26 -0400
>
>Try debugging the crash - start with “!analyze -v”. Just because you
>turn on an option and the system crashed you can’t assume that it will
>crash because of the option you enabled.
>
>You might also want to show us the bugcheck data. That always makes it
>easier to suggest what you might have done wrong.
>
>Regards,
>
>Tony Mason
>Consulting Partner
>OSR Open Systems Resources, Inc.
>http://www.osr.com
>
>
>-----Original Message-----
>From: xxxxx@lists.osr.com
>[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
>Sent: Tuesday, October 12, 2004 4:55 PM
>To: ntdev redirect
>Subject: RE: [ntdev] how long can I hold a spin lock?
>
>Thanks.
>
>I turn on the driver verifier with deadlock detection, and I get a bug
>check. Then I follow with “!deadlock” in the windbg, but I get “no
>deadlock
>verifier issues”. How can I continue from there?
>
>AH
>
> >From: “Doron Holan”
> >Reply-To: “Windows System Software Devs Interest List”
> >
> >To: “Windows System Software Devs Interest List”
> >Subject: RE: [ntdev] how long can I hold a spin lock?
> >Date: Mon, 11 Oct 2004 22:51:09 -0700
> >
> >You should hold onto a lock for the least amount of time as possible.
> >If you hold onto the lock too long, the user will see the affects.
>What
> >might be happening on your machine is that your driver is acquiring a
> >lock, but not releasing it and then return to the caller outside of
>your
> >driver. Turn on driver verifier and deadlock detection. If you
>think
> >this is what is going on, driver verifier should find it rather
>quickly.
> >
> >d
> >
> >-----Original Message-----
> >From: xxxxx@lists.osr.com
> >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> >Sent: Monday, October 11, 2004 8:18 PM
> >To: Windows System Software Devs Interest List
> >Subject: [ntdev] how long can I hold a spin lock?
> >
> >Is there any suggesting time? My driver seems to hang a HT system
>after
> >
> >running a while(UP is fine), and I cannot ping it. It looks like a
> >deadlock
> >to me. But I cannot break into the windbg. Any idea what to do?
> >
> >Thanks
> >
> >AH
> >
> >
> >Check out Election 2004 for up-to-date election news, plus voter tools
> >and
> >more! http://special.msn.com/msn/election2004.armx
> >
> >
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=256
> >
> >You are currently subscribed to ntdev as: xxxxx@windows.microsoft.com
> >To unsubscribe send a blank email to xxxxx@lists.osr.com
> >
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=256
> >
> >You are currently subscribed to ntdev as: unknown lmsubst tag argument:
>‘’
> >To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>
>On the road to retirement? Check out MSN Life Events for advice on how
>to
>get there! http://lifeevents.msn.com/category.aspx?cid=Retirement
>
>
>
>—
>Questions? First check the Kernel Driver FAQ at
>http://www.osronline.com/article.cfm?id=256
>
>You are currently subscribed to ntdev as: xxxxx@osr.com
>To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>—
>Questions? First check the Kernel Driver FAQ at
>http://www.osronline.com/article.cfm?id=256
>
>You are currently subscribed to ntdev as: unknown lmsubst tag argument: ‘’
>To unsubscribe send a blank email to xxxxx@lists.osr.com

_________________________________________________________________
Check out Election 2004 for up-to-date election news, plus voter tools and
more! http://special.msn.com/msn/election2004.armx

OSR_Community_User · October 12, 2004, 7:25pm

I’d guess that SVART is trying to call a kernel API (like
ExAcquireResourceSharedLite) while at raised IRQL. That’s probably a
good place for you to start debugging.

-p

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
Sent: Tuesday, October 12, 2004 4:13 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] how long can I hold a spin lock?

Thanks. Please see below. It is a HP/HT machine running XP Home.

*** Fatal System Error: 0x000000c4
(0x00000090,0xF870F120,0x00000000,0x00000000)

Break instruction exception - code 80000003 (first chance)

A fatal system error has occurred.
Debugger entered on first try; Bugcheck callbacks have not
been invoked.

A fatal system error has occurred.

Connected to Windows XP 2600 x86 compatible target, ptr64
FALSE Loading Kernel Symbols
…
…
…
Loading unloaded module list
…
Loading User Symbols
…
**************************************************************
*****************
*

*
* Bugcheck Analysis

*
*

*
**************************************************************
*****************

Use !analyze -v to get detailed debugging information.

BugCheck C4, {90, f870f120, 0, 0}

*** ERROR: Module load completed but symbols could not be
loaded for SAVRT.SYS Probably caused by : SAVRT.SYS ( SAVRT+7dad )

Followup: MachineOwner

nt!RtlpBreakWithStatusInstruction:
8051d778 cc int 3
1: kd> !analyze -v
**************************************************************
*****************
*

*
* Bugcheck Analysis

*
*

*
**************************************************************
*****************

DRIVER_VERIFIER_DETECTED_VIOLATION (c4)
A device driver attempting to corrupt the system has been
caught. This is because the driver was specified in the
registry as being suspect (by the
administrator) and the kernel has enabled substantial
checking of this driver.
If the driver attempts to corrupt the system, bugchecks 0xC4,
0xC1 and 0xA will be among the most commonly seen crashes.
Parameter 1 = 0x1000 … 0x1020 - deadlock verifier
error codes.
Typically the code is 0x1001 (deadlock
detected) and you can
issue a ‘!deadlock’ KD command to get more information.
Arguments:
Arg1: 00000090, A driver switched stacks. The current stack
is neither a thread
stack nor a DPC stack. Typically the driver doing this should be
on the stack obtained from `kb’ command.
Arg2: f870f120
Arg3: 00000000
Arg4: 00000000

Debugging Details:

BUGCHECK_STR: 0xc4_90

DEFAULT_BUCKET_ID: DRIVER_FAULT

LAST_CONTROL_TRANSFER: from 804f444e to 8051d778

STACK_TEXT:
824a693c 804f444e 00000003 824a6c6c 00000090
nt!RtlpBreakWithStatusInstruction
824a6988 804f4ed3 00000003 8276bd70 00000000
nt!KiBugCheckDebugBreak+0x19
824a6d54 804f5471 000000c4 00000090 f870f120 nt!KeBugCheck2+0x46d
824a6d74 8063ab77 000000c4 00000090 f870f120
nt!KeBugCheckEx+0x19 824a6d90 8063ae47 8063b3db 00000000
00000000 nt!ViDeadlockCheckStackLimits+0x6b
824a6d94 8063b3db 00000000 00000000 00000000
nt!ViDeadlockCanProceed+0x25
824a6db8 80538367 8276bd78 00000020 82675770
nt!VfDeadlockDeleteMemoryRange+0x11
824a6df4 8052618c 8276bd78 00000000 828add18 nt!ExFreePoolWithTag+0x95
824a6e24 805264e5 824a6e3c f00929ac 824a6ec0
nt!ExpFindCurrentThread+0x184
824a6e48 eecb5dad eece8028 00000000 eecb96ef
nt!ExAcquireResourceSharedLite+0x4f
WARNING: Stack unwind information not available. Following
frames may be wrong.
00000410 00009902 00240000 50e00024 1c0d1c0d SAVRT+0x7dad

FOLLOWUP_IP:
SAVRT+7dad
eecb5dad 84c0 test al,al

SYMBOL_STACK_INDEX: a

FOLLOWUP_NAME: MachineOwner

SYMBOL_NAME: SAVRT+7dad

MODULE_NAME: SAVRT

IMAGE_NAME: SAVRT.SYS

DEBUG_FLR_IMAGE_TIMESTAMP: 3d40d9c3

STACK_COMMAND: kb

BUCKET_ID: 0xc4_90_SAVRT+7dad

Followup: MachineOwner

1: kd> !deadlock

Resources: 1
Nodes: 0
Threads: 29

Nodes trimmed based on age 0.
Nodes trimmed based on count 0.
Analyze calls 0.
Maximum nodes searched 0.

No deadlock verifier issues.

>From: “Tony Mason”
> >Reply-To: “Windows System Software Devs Interest List”
> >
> >To: “Windows System Software Devs Interest List”
>
> >Subject: RE: [ntdev] how long can I hold a spin lock?
> >Date: Tue, 12 Oct 2004 17:54:26 -0400
> >
> >Try debugging the crash - start with “!analyze -v”. Just
> because you
> >turn on an option and the system crashed you can’t assume
> that it will
> >crash because of the option you enabled.
> >
> >You might also want to show us the bugcheck data. That
> always makes it
> >easier to suggest what you might have done wrong.
> >
> >Regards,
> >
> >Tony Mason
> >Consulting Partner
> >OSR Open Systems Resources, Inc.
> >http://www.osr.com
> >
> >
> >-----Original Message-----
> >From: xxxxx@lists.osr.com
> >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> >Sent: Tuesday, October 12, 2004 4:55 PM
> >To: ntdev redirect
> >Subject: RE: [ntdev] how long can I hold a spin lock?
> >
> >Thanks.
> >
> >I turn on the driver verifier with deadlock detection, and I
> get a bug
> >check. Then I follow with “!deadlock” in the windbg, but I get “no
> >deadlock verifier issues”. How can I continue from there?
> >
> >AH
> >
> > >From: “Doron Holan”
> > >Reply-To: “Windows System Software Devs Interest List”
> > >
> > >To: “Windows System Software Devs Interest List”
> > >
> > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > >Date: Mon, 11 Oct 2004 22:51:09 -0700
> > >
> > >You should hold onto a lock for the least amount of time
> as possible.
> > >If you hold onto the lock too long, the user will see the affects.
> >What
> > >might be happening on your machine is that your driver is
> acquiring a
> > >lock, but not releasing it and then return to the caller outside of
> >your
> > >driver. Turn on driver verifier and deadlock detection. If you
> >think
> > >this is what is going on, driver verifier should find it rather
> >quickly.
> > >
> > >d
> > >
> > >-----Original Message-----
> > >From: xxxxx@lists.osr.com
> > >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> > >Sent: Monday, October 11, 2004 8:18 PM
> > >To: Windows System Software Devs Interest List
> > >Subject: [ntdev] how long can I hold a spin lock?
> > >
> > >Is there any suggesting time? My driver seems to hang a HT system
> >after
> > >
> > >running a while(UP is fine), and I cannot ping it. It
> looks like a
> > >deadlock to me. But I cannot break into the windbg. Any
> idea what
> > >to do?
> > >
> > >Thanks
> > >
> > >AH
> > >
> > >
> > >Check out Election 2004 for up-to-date election news, plus voter
> > >tools and more! http://special.msn.com/msn/election2004.armx
> > >
> > >
> > >
> > >—
> > >Questions? First check the Kernel Driver FAQ at
> > >http://www.osronline.com/article.cfm?id=256
> > >
> > >You are currently subscribed to ntdev as:
> > >xxxxx@windows.microsoft.com To unsubscribe send a blank email to
> > >xxxxx@lists.osr.com
> > >
> > >
> > >—
> > >Questions? First check the Kernel Driver FAQ at
> > >http://www.osronline.com/article.cfm?id=256
> > >
> > >You are currently subscribed to ntdev as: unknown lmsubst
> tag argument:
> >‘’
> > >To unsubscribe send a blank email to
> xxxxx@lists.osr.com
> >
> >
> >On the road to retirement? Check out MSN Life Events for
> advice on how
> >to get there! http://lifeevents.msn.com/category.aspx?cid=Retirement
> >
> >
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=256
> >
> >You are currently subscribed to ntdev as: xxxxx@osr.com To
> unsubscribe
> >send a blank email to xxxxx@lists.osr.com
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=256
> >
> >You are currently subscribed to ntdev as: unknown lmsubst
> tag argument: ‘’
> >To unsubscribe send a blank email to xxxxx@lists.osr.com
>
> _________________________________________________________________
> Check out Election 2004 for up-to-date election news, plus
> voter tools and more! http://special.msn.com/msn/election2004.armx
>
>
>
> —
> Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
>
> You are currently subscribed to ntdev as:
> xxxxx@windows.microsoft.com
> To unsubscribe send a blank email to xxxxx@lists.osr.com
>

OSR_Community_User · October 12, 2004, 7:26pm

Unless you’re working for Symantec, this isn’t your bug. Symantec DOES
switch stacks.

I would suggest testing without an anti-virus product installed.

Regards,

Tony Mason
Consulting Partner
OSR Open Systems Resources, Inc.
http://www.osr.com

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
Sent: Tuesday, October 12, 2004 7:13 PM
To: ntdev redirect
Subject: RE: [ntdev] how long can I hold a spin lock?

Thanks. Please see below. It is a HP/HT machine running XP Home.

*** Fatal System Error: 0x000000c4
(0x00000090,0xF870F120,0x00000000,0x00000000)

Break instruction exception - code 80000003 (first chance)

A fatal system error has occurred.
Debugger entered on first try; Bugcheck callbacks have not been invoked.

A fatal system error has occurred.

Connected to Windows XP 2600 x86 compatible target, ptr64 FALSE
Loading Kernel Symbols
…
…
Loading unloaded module list
…
Loading User Symbols
…
************************************************************************
*******
*

*
* Bugcheck Analysis

*
*

*
************************************************************************
*******

Use !analyze -v to get detailed debugging information.

BugCheck C4, {90, f870f120, 0, 0}

*** ERROR: Module load completed but symbols could not be loaded for
SAVRT.SYS
Probably caused by : SAVRT.SYS ( SAVRT+7dad )

Followup: MachineOwner

nt!RtlpBreakWithStatusInstruction:
8051d778 cc int 3
1: kd> !analyze -v
************************************************************************
*******
*

*
* Bugcheck Analysis

*
*

*
************************************************************************
*******

DRIVER_VERIFIER_DETECTED_VIOLATION (c4)
A device driver attempting to corrupt the system has been caught. This
is
because the driver was specified in the registry as being suspect (by
the
administrator) and the kernel has enabled substantial checking of this
driver.
If the driver attempts to corrupt the system, bugchecks 0xC4, 0xC1 and
0xA
will
be among the most commonly seen crashes.
Parameter 1 = 0x1000 … 0x1020 - deadlock verifier error codes.
Typically the code is 0x1001 (deadlock detected) and you
can
issue a ‘!deadlock’ KD command to get more information.
Arguments:
Arg1: 00000090, A driver switched stacks. The current stack is neither a

thread
stack nor a DPC stack. Typically the driver doing this should be
on the stack obtained from `kb’ command.
Arg2: f870f120
Arg3: 00000000
Arg4: 00000000

Debugging Details:

BUGCHECK_STR: 0xc4_90

DEFAULT_BUCKET_ID: DRIVER_FAULT

LAST_CONTROL_TRANSFER: from 804f444e to 8051d778

STACK_TEXT:
824a693c 804f444e 00000003 824a6c6c 00000090
nt!RtlpBreakWithStatusInstruction
824a6988 804f4ed3 00000003 8276bd70 00000000
nt!KiBugCheckDebugBreak+0x19
824a6d54 804f5471 000000c4 00000090 f870f120 nt!KeBugCheck2+0x46d
824a6d74 8063ab77 000000c4 00000090 f870f120 nt!KeBugCheckEx+0x19
824a6d90 8063ae47 8063b3db 00000000 00000000
nt!ViDeadlockCheckStackLimits+0x6b
824a6d94 8063b3db 00000000 00000000 00000000
nt!ViDeadlockCanProceed+0x25
824a6db8 80538367 8276bd78 00000020 82675770
nt!VfDeadlockDeleteMemoryRange+0x11
824a6df4 8052618c 8276bd78 00000000 828add18 nt!ExFreePoolWithTag+0x95
824a6e24 805264e5 824a6e3c f00929ac 824a6ec0
nt!ExpFindCurrentThread+0x184
824a6e48 eecb5dad eece8028 00000000 eecb96ef
nt!ExAcquireResourceSharedLite+0x4f
WARNING: Stack unwind information not available. Following frames may be

wrong.
00000410 00009902 00240000 50e00024 1c0d1c0d SAVRT+0x7dad

FOLLOWUP_IP:
SAVRT+7dad
eecb5dad 84c0 test al,al

SYMBOL_STACK_INDEX: a

FOLLOWUP_NAME: MachineOwner

SYMBOL_NAME: SAVRT+7dad

MODULE_NAME: SAVRT

IMAGE_NAME: SAVRT.SYS

DEBUG_FLR_IMAGE_TIMESTAMP: 3d40d9c3

STACK_COMMAND: kb

BUCKET_ID: 0xc4_90_SAVRT+7dad

Followup: MachineOwner

1: kd> !deadlock

Resources: 1
Nodes: 0
Threads: 29

Nodes trimmed based on age 0.
Nodes trimmed based on count 0.
Analyze calls 0.
Maximum nodes searched 0.

No deadlock verifier issues.

From: “Tony Mason”
>Reply-To: “Windows System Software Devs Interest List”
>
>To: “Windows System Software Devs Interest List”
>Subject: RE: [ntdev] how long can I hold a spin lock?
>Date: Tue, 12 Oct 2004 17:54:26 -0400
>
>Try debugging the crash - start with “!analyze -v”. Just because you
>turn on an option and the system crashed you can’t assume that it will
>crash because of the option you enabled.
>
>You might also want to show us the bugcheck data. That always makes it
>easier to suggest what you might have done wrong.
>
>Regards,
>
>Tony Mason
>Consulting Partner
>OSR Open Systems Resources, Inc.
>http://www.osr.com
>
>
>-----Original Message-----
>From: xxxxx@lists.osr.com
>[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
>Sent: Tuesday, October 12, 2004 4:55 PM
>To: ntdev redirect
>Subject: RE: [ntdev] how long can I hold a spin lock?
>
>Thanks.
>
>I turn on the driver verifier with deadlock detection, and I get a bug
>check. Then I follow with “!deadlock” in the windbg, but I get “no
>deadlock
>verifier issues”. How can I continue from there?
>
>AH
>
> >From: “Doron Holan”
> >Reply-To: “Windows System Software Devs Interest List”
> >
> >To: “Windows System Software Devs Interest List”

> >Subject: RE: [ntdev] how long can I hold a spin lock?
> >Date: Mon, 11 Oct 2004 22:51:09 -0700
> >
> >You should hold onto a lock for the least amount of time as possible.
> >If you hold onto the lock too long, the user will see the affects.
>What
> >might be happening on your machine is that your driver is acquiring a
> >lock, but not releasing it and then return to the caller outside of
>your
> >driver. Turn on driver verifier and deadlock detection. If you
>think
> >this is what is going on, driver verifier should find it rather
>quickly.
> >
> >d
> >
> >-----Original Message-----
> >From: xxxxx@lists.osr.com
> >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> >Sent: Monday, October 11, 2004 8:18 PM
> >To: Windows System Software Devs Interest List
> >Subject: [ntdev] how long can I hold a spin lock?
> >
> >Is there any suggesting time? My driver seems to hang a HT system
>after
> >
> >running a while(UP is fine), and I cannot ping it. It looks like a
> >deadlock
> >to me. But I cannot break into the windbg. Any idea what to do?
> >
> >Thanks
> >
> >AH
> >
> >
> >Check out Election 2004 for up-to-date election news, plus voter
tools
> >and
> >more! http://special.msn.com/msn/election2004.armx
> >
> >
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=256
> >
> >You are currently subscribed to ntdev as:
xxxxx@windows.microsoft.com
> >To unsubscribe send a blank email to xxxxx@lists.osr.com
> >
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=256
> >
> >You are currently subscribed to ntdev as: unknown lmsubst tag
argument:
>‘’
> >To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>
>On the road to retirement? Check out MSN Life Events for advice on how
>to
>get there! http://lifeevents.msn.com/category.aspx?cid=Retirement
>
>
>
>—
>Questions? First check the Kernel Driver FAQ at
>http://www.osronline.com/article.cfm?id=256
>
>You are currently subscribed to ntdev as: xxxxx@osr.com
>To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>—
>Questions? First check the Kernel Driver FAQ at
>http://www.osronline.com/article.cfm?id=256
>
>You are currently subscribed to ntdev as: unknown lmsubst tag argument:
‘’
>To unsubscribe send a blank email to xxxxx@lists.osr.com

_________________________________________________________________
Check out Election 2004 for up-to-date election news, plus voter tools
and
more! http://special.msn.com/msn/election2004.armx

—
Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256

You are currently subscribed to ntdev as: xxxxx@osr.com
To unsubscribe send a blank email to xxxxx@lists.osr.com

OSR_Community_User · October 12, 2004, 7:43pm

Symantec switched stacks. That is a feature. Turn off deadlock
verification.
Ravi

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
Sent: Tuesday, October 12, 2004 4:13 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] how long can I hold a spin lock?

Thanks. Please see below. It is a HP/HT machine running XP Home.

*** Fatal System Error: 0x000000c4
(0x00000090,0xF870F120,0x00000000,0x00000000)

Break instruction exception - code 80000003 (first chance)

A fatal system error has occurred.
Debugger entered on first try; Bugcheck callbacks have not been invoked.

A fatal system error has occurred.

Connected to Windows XP 2600 x86 compatible target, ptr64 FALSE Loading
Kernel Symbols
…
…
Loading unloaded module list
…
Loading User Symbols
…
************************************************************************
*******
*

*
* Bugcheck Analysis

*
*

*
************************************************************************
*******

Use !analyze -v to get detailed debugging information.

BugCheck C4, {90, f870f120, 0, 0}

*** ERROR: Module load completed but symbols could not be loaded for
SAVRT.SYS Probably caused by : SAVRT.SYS ( SAVRT+7dad )

Followup: MachineOwner

nt!RtlpBreakWithStatusInstruction:
8051d778 cc int 3
1: kd> !analyze -v
************************************************************************
*******
*

*
* Bugcheck Analysis

*
*

*
************************************************************************
*******

DRIVER_VERIFIER_DETECTED_VIOLATION (c4)
A device driver attempting to corrupt the system has been caught. This
is because the driver was specified in the registry as being suspect (by
the
administrator) and the kernel has enabled substantial checking of this
driver.
If the driver attempts to corrupt the system, bugchecks 0xC4, 0xC1 and
0xA will be among the most commonly seen crashes.
Parameter 1 = 0x1000 … 0x1020 - deadlock verifier error codes.
Typically the code is 0x1001 (deadlock detected) and you
can
issue a ‘!deadlock’ KD command to get more information.
Arguments:
Arg1: 00000090, A driver switched stacks. The current stack is neither a
thread
stack nor a DPC stack. Typically the driver doing this should be
on the stack obtained from `kb’ command.
Arg2: f870f120
Arg3: 00000000
Arg4: 00000000

Debugging Details:

BUGCHECK_STR: 0xc4_90

DEFAULT_BUCKET_ID: DRIVER_FAULT

LAST_CONTROL_TRANSFER: from 804f444e to 8051d778

STACK_TEXT:
824a693c 804f444e 00000003 824a6c6c 00000090
nt!RtlpBreakWithStatusInstruction
824a6988 804f4ed3 00000003 8276bd70 00000000
nt!KiBugCheckDebugBreak+0x19
824a6d54 804f5471 000000c4 00000090 f870f120 nt!KeBugCheck2+0x46d
824a6d74 8063ab77 000000c4 00000090 f870f120 nt!KeBugCheckEx+0x19
824a6d90 8063ae47 8063b3db 00000000 00000000
nt!ViDeadlockCheckStackLimits+0x6b
824a6d94 8063b3db 00000000 00000000 00000000
nt!ViDeadlockCanProceed+0x25
824a6db8 80538367 8276bd78 00000020 82675770
nt!VfDeadlockDeleteMemoryRange+0x11
824a6df4 8052618c 8276bd78 00000000 828add18 nt!ExFreePoolWithTag+0x95
824a6e24 805264e5 824a6e3c f00929ac 824a6ec0
nt!ExpFindCurrentThread+0x184
824a6e48 eecb5dad eece8028 00000000 eecb96ef
nt!ExAcquireResourceSharedLite+0x4f
WARNING: Stack unwind information not available. Following frames may be
wrong.
00000410 00009902 00240000 50e00024 1c0d1c0d SAVRT+0x7dad

FOLLOWUP_IP:
SAVRT+7dad
eecb5dad 84c0 test al,al

SYMBOL_STACK_INDEX: a

FOLLOWUP_NAME: MachineOwner

SYMBOL_NAME: SAVRT+7dad

MODULE_NAME: SAVRT

IMAGE_NAME: SAVRT.SYS

DEBUG_FLR_IMAGE_TIMESTAMP: 3d40d9c3

STACK_COMMAND: kb

BUCKET_ID: 0xc4_90_SAVRT+7dad

Followup: MachineOwner

1: kd> !deadlock

Resources: 1
Nodes: 0
Threads: 29

Nodes trimmed based on age 0.
Nodes trimmed based on count 0.
Analyze calls 0.
Maximum nodes searched 0.

No deadlock verifier issues.

From: “Tony Mason”
>Reply-To: “Windows System Software Devs Interest List”
>
>To: “Windows System Software Devs Interest List”
>Subject: RE: [ntdev] how long can I hold a spin lock?
>Date: Tue, 12 Oct 2004 17:54:26 -0400
>
>Try debugging the crash - start with “!analyze -v”. Just because you
>turn on an option and the system crashed you can’t assume that it will
>crash because of the option you enabled.
>
>You might also want to show us the bugcheck data. That always makes it

>easier to suggest what you might have done wrong.
>
>Regards,
>
>Tony Mason
>Consulting Partner
>OSR Open Systems Resources, Inc.
>http://www.osr.com
>
>
>-----Original Message-----
>From: xxxxx@lists.osr.com
>[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
>Sent: Tuesday, October 12, 2004 4:55 PM
>To: ntdev redirect
>Subject: RE: [ntdev] how long can I hold a spin lock?
>
>Thanks.
>
>I turn on the driver verifier with deadlock detection, and I get a bug
>check. Then I follow with “!deadlock” in the windbg, but I get “no
>deadlock verifier issues”. How can I continue from there?
>
>AH
>
> >From: “Doron Holan”
> >Reply-To: “Windows System Software Devs Interest List”
> >
> >To: “Windows System Software Devs Interest List”
> >
> >Subject: RE: [ntdev] how long can I hold a spin lock?
> >Date: Mon, 11 Oct 2004 22:51:09 -0700
> >
> >You should hold onto a lock for the least amount of time as possible.
> >If you hold onto the lock too long, the user will see the affects.
>What
> >might be happening on your machine is that your driver is acquiring a

> >lock, but not releasing it and then return to the caller outside of
>your
> >driver. Turn on driver verifier and deadlock detection. If you
>think
> >this is what is going on, driver verifier should find it rather
>quickly.
> >
> >d
> >
> >-----Original Message-----
> >From: xxxxx@lists.osr.com
> >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> >Sent: Monday, October 11, 2004 8:18 PM
> >To: Windows System Software Devs Interest List
> >Subject: [ntdev] how long can I hold a spin lock?
> >
> >Is there any suggesting time? My driver seems to hang a HT system
>after
> >
> >running a while(UP is fine), and I cannot ping it. It looks like a
> >deadlock to me. But I cannot break into the windbg. Any idea what
> >to do?
> >
> >Thanks
> >
> >AH
> >
> >
> >Check out Election 2004 for up-to-date election news, plus voter
> >tools and more! http://special.msn.com/msn/election2004.armx
> >
> >
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=256
> >
> >You are currently subscribed to ntdev as:
> >xxxxx@windows.microsoft.com To unsubscribe send a blank email to
> >xxxxx@lists.osr.com
> >
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=256
> >
> >You are currently subscribed to ntdev as: unknown lmsubst tag
argument:
>‘’
> >To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>
>On the road to retirement? Check out MSN Life Events for advice on how
>to get there! http://lifeevents.msn.com/category.aspx?cid=Retirement
>
>
>
>—
>Questions? First check the Kernel Driver FAQ at
>http://www.osronline.com/article.cfm?id=256
>
>You are currently subscribed to ntdev as: xxxxx@osr.com To unsubscribe
>send a blank email to xxxxx@lists.osr.com
>
>—
>Questions? First check the Kernel Driver FAQ at
>http://www.osronline.com/article.cfm?id=256
>
>You are currently subscribed to ntdev as: unknown lmsubst tag argument:
‘’
>To unsubscribe send a blank email to xxxxx@lists.osr.com

_________________________________________________________________
Check out Election 2004 for up-to-date election news, plus voter tools
and more! http://special.msn.com/msn/election2004.armx

—
Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256

You are currently subscribed to ntdev as: xxxxx@windows.microsoft.com
To unsubscribe send a blank email to xxxxx@lists.osr.com

OSR_Community_User · October 12, 2004, 11:09pm

Many thanks!

Once I uninstall the Norton SW, everything works beautifully. I have been
scratching my head for a week. Should I warn my customers or make a
complaint to Symantec?!

AH

From: “Tony Mason”
>Reply-To: “Windows System Software Devs Interest List”
>
>To: “Windows System Software Devs Interest List”
>Subject: RE: [ntdev] how long can I hold a spin lock?
>Date: Tue, 12 Oct 2004 19:23:43 -0400
>
>Unless you’re working for Symantec, this isn’t your bug. Symantec DOES
>switch stacks.
>
>I would suggest testing without an anti-virus product installed.
>
>Regards,
>
>Tony Mason
>Consulting Partner
>OSR Open Systems Resources, Inc.
>http://www.osr.com
>
>
>-----Original Message-----
>From: xxxxx@lists.osr.com
>[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
>Sent: Tuesday, October 12, 2004 7:13 PM
>To: ntdev redirect
>Subject: RE: [ntdev] how long can I hold a spin lock?
>
>Thanks. Please see below. It is a HP/HT machine running XP Home.
>================================================
>Fatal System Error: 0x000000c4
> (0x00000090,0xF870F120,0x00000000,0x00000000)
>
>Break instruction exception - code 80000003 (first chance)
>
>A fatal system error has occurred.
>Debugger entered on first try; Bugcheck callbacks have not been invoked.
>
>A fatal system error has occurred.
>
>Connected to Windows XP 2600 x86 compatible target, ptr64 FALSE
>Loading Kernel Symbols
>…
>…
>Loading unloaded module list
>…
>Loading User Symbols
>…
>****
>
>
>
>
> Bugcheck Analysis
>
>
>
>
>
>
> ****
>
>Use !analyze -v to get detailed debugging information.
>
>BugCheck C4, {90, f870f120, 0, 0}
>
> ERROR: Module load completed but symbols could not be loaded for
>SAVRT.SYS
>Probably caused by : SAVRT.SYS ( SAVRT+7dad )
>
>Followup: MachineOwner
>---------
>
>nt!RtlpBreakWithStatusInstruction:
>8051d778 cc int 3
>1: kd> !analyze -v
> ****************************************************************
>
>
>
>
> Bugcheck Analysis
>
>
>
>
> *
> *****************************************************************
>
>
>DRIVER_VERIFIER_DETECTED_VIOLATION (c4)
>A device driver attempting to corrupt the system has been caught. This
>is
>because the driver was specified in the registry as being suspect (by
>the
>administrator) and the kernel has enabled substantial checking of this
>driver.
>If the driver attempts to corrupt the system, bugchecks 0xC4, 0xC1 and
>0xA
>will
>be among the most commonly seen crashes.
> Parameter 1 = 0x1000 … 0x1020 - deadlock verifier error codes.
> Typically the code is 0x1001 (deadlock detected) and you
>can
> issue a ‘!deadlock’ KD command to get more information.
>Arguments:
>Arg1: 00000090, A driver switched stacks. The current stack is neither a
>
>thread
> stack nor a DPC stack. Typically the driver doing this should be
> on the stack obtained from `kb’ command.
>Arg2: f870f120
>Arg3: 00000000
>Arg4: 00000000
>
>Debugging Details:
>------------------
>
>
>BUGCHECK_STR: 0xc4_90
>
>DEFAULT_BUCKET_ID: DRIVER_FAULT
>
>LAST_CONTROL_TRANSFER: from 804f444e to 8051d778
>
>STACK_TEXT:
>824a693c 804f444e 00000003 824a6c6c 00000090
>nt!RtlpBreakWithStatusInstruction
>824a6988 804f4ed3 00000003 8276bd70 00000000
>nt!KiBugCheckDebugBreak+0x19
>824a6d54 804f5471 000000c4 00000090 f870f120 nt!KeBugCheck2+0x46d
>824a6d74 8063ab77 000000c4 00000090 f870f120 nt!KeBugCheckEx+0x19
>824a6d90 8063ae47 8063b3db 00000000 00000000
>nt!ViDeadlockCheckStackLimits+0x6b
>824a6d94 8063b3db 00000000 00000000 00000000
>nt!ViDeadlockCanProceed+0x25
>824a6db8 80538367 8276bd78 00000020 82675770
>nt!VfDeadlockDeleteMemoryRange+0x11
>824a6df4 8052618c 8276bd78 00000000 828add18 nt!ExFreePoolWithTag+0x95
>824a6e24 805264e5 824a6e3c f00929ac 824a6ec0
>nt!ExpFindCurrentThread+0x184
>824a6e48 eecb5dad eece8028 00000000 eecb96ef
>nt!ExAcquireResourceSharedLite+0x4f
>WARNING: Stack unwind information not available. Following frames may be
>
>wrong.
>00000410 00009902 00240000 50e00024 1c0d1c0d SAVRT+0x7dad
>
>
>FOLLOWUP_IP:
>SAVRT+7dad
>eecb5dad 84c0 test al,al
>
>SYMBOL_STACK_INDEX: a
>
>FOLLOWUP_NAME: MachineOwner
>
>SYMBOL_NAME: SAVRT+7dad
>
>MODULE_NAME: SAVRT
>
>IMAGE_NAME: SAVRT.SYS
>
>DEBUG_FLR_IMAGE_TIMESTAMP: 3d40d9c3
>
>STACK_COMMAND: kb
>
>BUCKET_ID: 0xc4_90_SAVRT+7dad
>
>Followup: MachineOwner
>---------
>
>1: kd> !deadlock
>
>Resources: 1
>Nodes: 0
>Threads: 29
>
>Nodes trimmed based on age 0.
>Nodes trimmed based on count 0.
>Analyze calls 0.
>Maximum nodes searched 0.
>
>No deadlock verifier issues.
>
>
>
> >From: “Tony Mason”
> >Reply-To: “Windows System Software Devs Interest List”
> >
> >To: “Windows System Software Devs Interest List”
> >Subject: RE: [ntdev] how long can I hold a spin lock?
> >Date: Tue, 12 Oct 2004 17:54:26 -0400
> >
> >Try debugging the crash - start with “!analyze -v”. Just because you
> >turn on an option and the system crashed you can’t assume that it will
> >crash because of the option you enabled.
> >
> >You might also want to show us the bugcheck data. That always makes it
> >easier to suggest what you might have done wrong.
> >
> >Regards,
> >
> >Tony Mason
> >Consulting Partner
> >OSR Open Systems Resources, Inc.
> >http://www.osr.com
> >
> >
> >-----Original Message-----
> >From: xxxxx@lists.osr.com
> >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> >Sent: Tuesday, October 12, 2004 4:55 PM
> >To: ntdev redirect
> >Subject: RE: [ntdev] how long can I hold a spin lock?
> >
> >Thanks.
> >
> >I turn on the driver verifier with deadlock detection, and I get a bug
> >check. Then I follow with “!deadlock” in the windbg, but I get “no
> >deadlock
> >verifier issues”. How can I continue from there?
> >
> >AH
> >
> > >From: “Doron Holan”
> > >Reply-To: “Windows System Software Devs Interest List”
> > >
> > >To: “Windows System Software Devs Interest List”
>
> > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > >Date: Mon, 11 Oct 2004 22:51:09 -0700
> > >
> > >You should hold onto a lock for the least amount of time as possible.
> > >If you hold onto the lock too long, the user will see the affects.
> >What
> > >might be happening on your machine is that your driver is acquiring a
> > >lock, but not releasing it and then return to the caller outside of
> >your
> > >driver. Turn on driver verifier and deadlock detection. If you
> >think
> > >this is what is going on, driver verifier should find it rather
> >quickly.
> > >
> > >d
> > >
> > >-----Original Message-----
> > >From: xxxxx@lists.osr.com
> > >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> > >Sent: Monday, October 11, 2004 8:18 PM
> > >To: Windows System Software Devs Interest List
> > >Subject: [ntdev] how long can I hold a spin lock?
> > >
> > >Is there any suggesting time? My driver seems to hang a HT system
> >after
> > >
> > >running a while(UP is fine), and I cannot ping it. It looks like a
> > >deadlock
> > >to me. But I cannot break into the windbg. Any idea what to do?
> > >
> > >Thanks
> > >
> > >AH
> > >
> > >
> > >Check out Election 2004 for up-to-date election news, plus voter
>tools
> > >and
> > >more! http://special.msn.com/msn/election2004.armx
> > >
> > >
> > >
> > >—
> > >Questions? First check the Kernel Driver FAQ at
> > >http://www.osronline.com/article.cfm?id=256
> > >
> > >You are currently subscribed to ntdev as:
>xxxxx@windows.microsoft.com
> > >To unsubscribe send a blank email to xxxxx@lists.osr.com
> > >
> > >
> > >—
> > >Questions? First check the Kernel Driver FAQ at
> > >http://www.osronline.com/article.cfm?id=256
> > >
> > >You are currently subscribed to ntdev as: unknown lmsubst tag
>argument:
> >‘’
> > >To unsubscribe send a blank email to xxxxx@lists.osr.com
> >
> >
> >On the road to retirement? Check out MSN Life Events for advice on how
> >to
> >get there! http://lifeevents.msn.com/category.aspx?cid=Retirement
> >
> >
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=256
> >
> >You are currently subscribed to ntdev as: xxxxx@osr.com
> >To unsubscribe send a blank email to xxxxx@lists.osr.com
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=256
> >
> >You are currently subscribed to ntdev as: unknown lmsubst tag argument:
>‘’
> >To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>
>Check out Election 2004 for up-to-date election news, plus voter tools
>and
>more! http://special.msn.com/msn/election2004.armx
>
>
>
>—
>Questions? First check the Kernel Driver FAQ at
>http://www.osronline.com/article.cfm?id=256
>
>You are currently subscribed to ntdev as: xxxxx@osr.com
>To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>—
>Questions? First check the Kernel Driver FAQ at
>http://www.osronline.com/article.cfm?id=256
>
>You are currently subscribed to ntdev as: unknown lmsubst tag argument: ‘’
>To unsubscribe send a blank email to xxxxx@lists.osr.com

Is your PC infected? Get a FREE online computer virus scan from McAfee®
Security. http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963

OSR_Community_User · October 13, 2004, 4:41am

You can try making a complaint to Symantec, but I expect that they have=
a
very good reason to violate the MS rules, or they wouldn’t be doing tha=
t.

Of course, why your software isn’t working with Symantec is probably no=
t
the same reason as this bugcheck, so maybe you want to try to figure ou=
t
why your driver isn’t working together with Symantec’s AV software stil=
l,
as I suspect Symantec isn’t going to fix whatever the problem is
themselves…

Whether you want to do this or not depends on whether you think it’s
important that your driver works together with Symantec AV software or =
not.
It’s easy to write in the README-file that it doesn’t work with Symante=
c
AV. It is, however, much harder to convince the consumer of the softwar=
e
that it’s actually NOT YOUR FAULT (assuming it isn’t, of course), and m=
ay
be even harder yet to convince the consumer that rather than not using =
your
driver, they should change the brand of AV software. Changing the AV
software if you work in a small company/home environment is trivial, bu=
t
large companies with many hundreds or thousands of users, it’s often a
corporate policy that is very strict, and you need really good reasons =
to
get an exception to this policy.

With this in mind, I would recommend that you still try to fix your dri=
ver.
At least make some effort. I understand that it’s not trivial. And it m=
ay
be a good idea to register a complaint with Symantec, and see what help=

they may be able to offer you.

Also, you will want to try several different brands of AV software with=

your driver, to make sure that you’re working well with those drivers.

So, the conclusion is that you can’t use Deadlock detection if you’re u=
sing
Symantec/Norton AV, but that doesn’t mean that you haven’t got another =
bug
somewhere…

–
Mats

xxxxx@lists.osr.com wrote on 10/13/2004 04:07:16 AM:

Many thanks!

Once I uninstall the Norton SW, everything works beautifully. I have=

been

scratching my head for a week. Should I warn my customers or make a
complaint to Symantec?!

AH

>From: “Tony Mason”
> >Reply-To: “Windows System Software Devs Interest List”
> >
> >To: “Windows System Software Devs Interest List” m>
> >Subject: RE: [ntdev] how long can I hold a spin lock?
> >Date: Tue, 12 Oct 2004 19:23:43 -0400
> >
> >Unless you’re working for Symantec, this isn’t your bug. Symantec D=
OES
> >switch stacks.
> >
> >I would suggest testing without an anti-virus product installed.
> >
> >Regards,
> >
> >Tony Mason
> >Consulting Partner
> >OSR Open Systems Resources, Inc.
> >http://www.osr.com
> >
> >
> >-----Original Message-----
> >From: xxxxx@lists.osr.com
> >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> >Sent: Tuesday, October 12, 2004 7:13 PM
> >To: ntdev redirect
> >Subject: RE: [ntdev] how long can I hold a spin lock?
> >
> >Thanks. Please see below. It is a HP/HT machine running XP Home.
> >=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D
> >Fatal System Error: 0x000000c4
> > (0x00000090,0xF870F120,0x00000000,0x00000000=
)
> >
> >Break instruction exception - code 80000003 (first chance)
> >
> >A fatal system error has occurred.
> >Debugger entered on first try; Bugcheck callbacks have not been invo=
ked.
> >
> >A fatal system error has occurred.
> >
> >Connected to Windows XP 2600 x86 compatible target, ptr64 FALSE
> >Loading Kernel Symbols
> >…=
…
> >…
> >Loading unloaded module list
> >…
> >Loading User Symbols
> >…
> >*****************************************************************=

> >***
> >*
> >
> >
> > Bugcheck Analysis
> >
> >
> >
> >
> > *
> > ****************************************************************=

> > ****
> >
> >Use !analyze -v to get detailed debugging information.
> >
> >BugCheck C4, {90, f870f120, 0, 0}
> >
> > ERROR: Module load completed but symbols could not be loaded for=

> >SAVRT.SYS
> >Probably caused by : SAVRT.SYS ( SAVRT+7dad )
> >
> >Followup: MachineOwner
> >---------
> >
> >nt!RtlpBreakWithStatusInstruction:
> >8051d778 cc int 3
> >1: kd> !analyze -v
> > =

> >
> >
> >
> >
> > Bugcheck Analysis
> >
> >
> >
> >
> >
> > =

> >
> >
> >DRIVER_VERIFIER_DETECTED_VIOLATION (c4)
> >A device driver attempting to corrupt the system has been caught. T=
his
> >is
> >because the driver was specified in the registry as being suspect (b=
y
> >the
> >administrator) and the kernel has enabled substantial checking of th=
is
> >driver.
> >If the driver attempts to corrupt the system, bugchecks 0xC4, 0xC1 a=
nd
> >0xA
> >will
> >be among the most commonly seen crashes.
> > Parameter 1 =3D 0x1000 … 0x1020 - deadlock verifier error =
codes.
> > Typically the code is 0x1001 (deadlock detected) and=
you
> >can
> > issue a ‘!deadlock’ KD command to get more informati=
on.
> >Arguments:
> >Arg1: 00000090, A driver switched stacks. The current stack is neith=
er a
> >
> >thread
> > stack nor a DPC stack. Typically the driver doing this should be
> > on the stack obtained from `kb’ command.
> >Arg2: f870f120
> >Arg3: 00000000
> >Arg4: 00000000
> >
> >Debugging Details:
> >------------------
> >
> >
> >BUGCHECK_STR: 0xc4_90
> >
> >DEFAULT_BUCKET_ID: DRIVER_FAULT
> >
> >LAST_CONTROL_TRANSFER: from 804f444e to 8051d778
> >
> >STACK_TEXT:
> >824a693c 804f444e 00000003 824a6c6c 00000090
> >nt!RtlpBreakWithStatusInstruction
> >824a6988 804f4ed3 00000003 8276bd70 00000000
> >nt!KiBugCheckDebugBreak+0x19
> >824a6d54 804f5471 000000c4 00000090 f870f120 nt!KeBugCheck2+0x46d
> >824a6d74 8063ab77 000000c4 00000090 f870f120 nt!KeBugCheckEx+0x19
> >824a6d90 8063ae47 8063b3db 00000000 00000000
> >nt!ViDeadlockCheckStackLimits+0x6b
> >824a6d94 8063b3db 00000000 00000000 00000000
> >nt!ViDeadlockCanProceed+0x25
> >824a6db8 80538367 8276bd78 00000020 82675770
> >nt!VfDeadlockDeleteMemoryRange+0x11
> >824a6df4 8052618c 8276bd78 00000000 828add18 nt!ExFreePoolWithTag+0x=
95
> >824a6e24 805264e5 824a6e3c f00929ac 824a6ec0
> >nt!ExpFindCurrentThread+0x184
> >824a6e48 eecb5dad eece8028 00000000 eecb96ef
> >nt!ExAcquireResourceSharedLite+0x4f
> >WARNING: Stack unwind information not available. Following frames ma=
y be
> >
> >wrong.
> >00000410 00009902 00240000 50e00024 1c0d1c0d SAVRT+0x7dad
> >
> >
> >FOLLOWUP_IP:
> >SAVRT+7dad
> >eecb5dad 84c0 test al,al
> >
> >SYMBOL_STACK_INDEX: a
> >
> >FOLLOWUP_NAME: MachineOwner
> >
> >SYMBOL_NAME: SAVRT+7dad
> >
> >MODULE_NAME: SAVRT
> >
> >IMAGE_NAME: SAVRT.SYS
> >
> >DEBUG_FLR_IMAGE_TIMESTAMP: 3d40d9c3
> >
> >STACK_COMMAND: kb
> >
> >BUCKET_ID: 0xc4_90_SAVRT+7dad
> >
> >Followup: MachineOwner
> >---------
> >
> >1: kd> !deadlock
> >
> >Resources: 1
> >Nodes: 0
> >Threads: 29
> >
> >Nodes trimmed based on age 0.
> >Nodes trimmed based on count 0.
> >Analyze calls 0.
> >Maximum nodes searched 0.
> >
> >No deadlock verifier issues.
> >
> >
> >
> > >From: “Tony Mason”
> > >Reply-To: “Windows System Software Devs Interest List”
> > >
> > >To: “Windows System Software Devs Interest List” com>
> > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > >Date: Tue, 12 Oct 2004 17:54:26 -0400
> > >
> > >Try debugging the crash - start with “!analyze -v”. Just because =
you
> > >turn on an option and the system crashed you can’t assume that it =
will
> > >crash because of the option you enabled.
> > >
> > >You might also want to show us the bugcheck data. That always mak=
es
it
> > >easier to suggest what you might have done wrong.
> > >
> > >Regards,
> > >
> > >Tony Mason
> > >Consulting Partner
> > >OSR Open Systems Resources, Inc.
> > >http://www.osr.com
> > >
> > >
> > >-----Original Message-----
> > >From: xxxxx@lists.osr.com
> > >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> > >Sent: Tuesday, October 12, 2004 4:55 PM
> > >To: ntdev redirect
> > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > >
> > >Thanks.
> > >
> > >I turn on the driver verifier with deadlock detection, and I get a=
bug
> > >check. Then I follow with “!deadlock” in the windbg, but I get “n=
o
> > >deadlock
> > >verifier issues”. How can I continue from there?
> > >
> > >AH
> > >
> > > >From: “Doron Holan”
> > > >Reply-To: “Windows System Software Devs Interest List”
> > > >
> > > >To: “Windows System Software Devs Interest List”
> >
> > > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > > >Date: Mon, 11 Oct 2004 22:51:09 -0700
> > > >
> > > >You should hold onto a lock for the least amount of time as
possible.
> > > >If you hold onto the lock too long, the user will see the affect=
s.
> > >What
> > > >might be happening on your machine is that your driver is acquir=
ing
a
> > > >lock, but not releasing it and then return to the caller outside=
of
> > >your
> > > >driver. Turn on driver verifier and deadlock detection. If y=
ou
> > >think
> > > >this is what is going on, driver verifier should find it rather
> > >quickly.
> > > >
> > > >d
> > > >
> > > >-----Original Message-----
> > > >From: xxxxx@lists.osr.com
> > > >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> > > >Sent: Monday, October 11, 2004 8:18 PM
> > > >To: Windows System Software Devs Interest List
> > > >Subject: [ntdev] how long can I hold a spin lock?
> > > >
> > > >Is there any suggesting time? My driver seems to hang a HT syst=
em
> > >after
> > > >
> > > >running a while(UP is fine), and I cannot ping it. It looks lik=
e a
> > > >deadlock
> > > >to me. But I cannot break into the windbg. Any idea what to do=
?
> > > >
> > > >Thanks
> > > >
> > > >AH
> > > >
> > > > =

> > > >Check out Election 2004 for up-to-date election news, plus voter=

> >tools
> > > >and
> > > >more! http://special.msn.com/msn/election2004.armx
> > > >
> > > >
> > > >
> > > >—
> > > >Questions? First check the Kernel Driver FAQ at
> > > >http://www.osronline.com/article.cfm?id=3D256
> > > >
> > > >You are currently subscribed to ntdev as:
> >xxxxx@windows.microsoft.com
> > > >To unsubscribe send a blank email to
xxxxx@lists.osr.com
> > > >
> > > >
> > > >—
> > > >Questions? First check the Kernel Driver FAQ at
> > > >http://www.osronline.com/article.cfm?id=3D256
> > > >
> > > >You are currently subscribed to ntdev as: unknown lmsubst tag
> >argument:
> > >‘’
> > > >To unsubscribe send a blank email to
xxxxx@lists.osr.com
> > >
> > >__
> > >On the road to retirement? Check out MSN Life Events for advice on=
how
> > >to
> > >get there! http://lifeevents.msn.com/category.aspx?cid=3DRetiremen=
t
> > >
> > >
> > >
> > >—
> > >Questions? First check the Kernel Driver FAQ at
> > >http://www.osronline.com/article.cfm?id=3D256
> > >
> > >You are currently subscribed to ntdev as: xxxxx@osr.com
> > >To unsubscribe send a blank email to xxxxx@lists.osr.=
com
> > >
> > >—
> > >Questions? First check the Kernel Driver FAQ at
> > >http://www.osronline.com/article.cfm?id=3D256
> > >
> > >You are currently subscribed to ntdev as: unknown lmsubst tag
argument:
> >‘’
> > >To unsubscribe send a blank email to xxxxx@lists.osr.=
com
> >
> >
> >Check out Election 2004 for up-to-date election news, plus voter too=
ls
> >and
> >more! http://special.msn.com/msn/election2004.armx
> >
> >
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=3D256
> >
> >You are currently subscribed to ntdev as: xxxxx@osr.com
> >To unsubscribe send a blank email to xxxxx@lists.osr.co=
m
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=3D256
> >
> >You are currently subscribed to ntdev as: unknown lmsubst tag argume=
nt:
‘’
> >To unsubscribe send a blank email to xxxxx@lists.osr.co=
m
>
>
> Is your PC infected? Get a FREE online computer virus scan from McAfe=
e=AE
> Security. http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3D396=
3
>
>
>
> —
> Questions? First check the Kernel Driver FAQ at http://www.
> osronline.com/article.cfm?id=3D256
>
> You are currently subscribed to ntdev as: xxxxx@3dlabs.com
> To unsubscribe send a blank email to xxxxx@lists.osr.com=

> ForwardSourceID:NT0000544A=

Don_Burn_1 · October 13, 2004, 8:38am

Mats,

Almost any driver will get that bugcheck with Symantec and driver
verifier. The question is there other problems with his driver and Symantec
or just this one.

–
Don Burn (MVP, Windows DDK)
Windows 2k/XP/2k3 Filesystem and Driver Consulting
Remove StopSpam from the email to reply

“Mats PETERSSON” wrote in message
news:xxxxx@ntdev…

You can try making a complaint to Symantec, but I expect that they have a
very good reason to violate the MS rules, or they wouldn’t be doing that.

Of course, why your software isn’t working with Symantec is probably not
the same reason as this bugcheck, so maybe you want to try to figure out
why your driver isn’t working together with Symantec’s AV software still,
as I suspect Symantec isn’t going to fix whatever the problem is
themselves…

Whether you want to do this or not depends on whether you think it’s
important that your driver works together with Symantec AV software or not.
It’s easy to write in the README-file that it doesn’t work with Symantec
AV. It is, however, much harder to convince the consumer of the software
that it’s actually NOT YOUR FAULT (assuming it isn’t, of course), and may
be even harder yet to convince the consumer that rather than not using your
driver, they should change the brand of AV software. Changing the AV
software if you work in a small company/home environment is trivial, but
large companies with many hundreds or thousands of users, it’s often a
corporate policy that is very strict, and you need really good reasons to
get an exception to this policy.

With this in mind, I would recommend that you still try to fix your driver.
At least make some effort. I understand that it’s not trivial. And it may
be a good idea to register a complaint with Symantec, and see what help
they may be able to offer you.

Also, you will want to try several different brands of AV software with
your driver, to make sure that you’re working well with those drivers.

So, the conclusion is that you can’t use Deadlock detection if you’re using
Symantec/Norton AV, but that doesn’t mean that you haven’t got another bug
somewhere…

–
Mats

xxxxx@lists.osr.com wrote on 10/13/2004 04:07:16 AM:

> Many thanks!
>
> Once I uninstall the Norton SW, everything works beautifully. I have
been
> scratching my head for a week. Should I warn my customers or make a
> complaint to Symantec?!
>
> AH
>
>
> >From: “Tony Mason”
> >Reply-To: “Windows System Software Devs Interest List”
> >
> >To: “Windows System Software Devs Interest List”
> >Subject: RE: [ntdev] how long can I hold a spin lock?
> >Date: Tue, 12 Oct 2004 19:23:43 -0400
> >
> >Unless you’re working for Symantec, this isn’t your bug. Symantec DOES
> >switch stacks.
> >
> >I would suggest testing without an anti-virus product installed.
> >
> >Regards,
> >
> >Tony Mason
> >Consulting Partner
> >OSR Open Systems Resources, Inc.
> >http://www.osr.com
> >
> >
> >-----Original Message-----
> >From: xxxxx@lists.osr.com
> >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> >Sent: Tuesday, October 12, 2004 7:13 PM
> >To: ntdev redirect
> >Subject: RE: [ntdev] how long can I hold a spin lock?
> >
> >Thanks. Please see below. It is a HP/HT machine running XP Home.
> >================================================
> >Fatal System Error: 0x000000c4
> > (0x00000090,0xF870F120,0x00000000,0x00000000)
> >
> >Break instruction exception - code 80000003 (first chance)
> >
> >A fatal system error has occurred.
> >Debugger entered on first try; Bugcheck callbacks have not been invoked.
> >
> >A fatal system error has occurred.
> >
> >Connected to Windows XP 2600 x86 compatible target, ptr64 FALSE
> >Loading Kernel Symbols
> >…
> >…
> >Loading unloaded module list
> >…
> >Loading User Symbols
> >…
> >****
> >
> >
> >
> >
> > Bugcheck Analysis
> >
> >
> >
> >
> >
> >
> > ****
> >
> >Use !analyze -v to get detailed debugging information.
> >
> >BugCheck C4, {90, f870f120, 0, 0}
> >
> > ERROR: Module load completed but symbols could not be loaded for
> >SAVRT.SYS
> >Probably caused by : SAVRT.SYS ( SAVRT+7dad )
> >
> >Followup: MachineOwner
> >---------
> >
> >nt!RtlpBreakWithStatusInstruction:
> >8051d778 cc int 3
> >1: kd> !analyze -v
> > ****************************************************************
> >
> >
> >
> >
> > Bugcheck Analysis
> >
> >
> >
> >
> > *
> > *****************************************************************
> >
> >
> >DRIVER_VERIFIER_DETECTED_VIOLATION (c4)
> >A device driver attempting to corrupt the system has been caught. This
> >is
> >because the driver was specified in the registry as being suspect (by
> >the
> >administrator) and the kernel has enabled substantial checking of this
> >driver.
> >If the driver attempts to corrupt the system, bugchecks 0xC4, 0xC1 and
> >0xA
> >will
> >be among the most commonly seen crashes.
> > Parameter 1 = 0x1000 … 0x1020 - deadlock verifier error codes.
> > Typically the code is 0x1001 (deadlock detected) and you
> >can
> > issue a ‘!deadlock’ KD command to get more information.
> >Arguments:
> >Arg1: 00000090, A driver switched stacks. The current stack is neither a
> >
> >thread
> > stack nor a DPC stack. Typically the driver doing this should be
> > on the stack obtained from `kb’ command.
> >Arg2: f870f120
> >Arg3: 00000000
> >Arg4: 00000000
> >
> >Debugging Details:
> >------------------
> >
> >
> >BUGCHECK_STR: 0xc4_90
> >
> >DEFAULT_BUCKET_ID: DRIVER_FAULT
> >
> >LAST_CONTROL_TRANSFER: from 804f444e to 8051d778
> >
> >STACK_TEXT:
> >824a693c 804f444e 00000003 824a6c6c 00000090
> >nt!RtlpBreakWithStatusInstruction
> >824a6988 804f4ed3 00000003 8276bd70 00000000
> >nt!KiBugCheckDebugBreak+0x19
> >824a6d54 804f5471 000000c4 00000090 f870f120 nt!KeBugCheck2+0x46d
> >824a6d74 8063ab77 000000c4 00000090 f870f120 nt!KeBugCheckEx+0x19
> >824a6d90 8063ae47 8063b3db 00000000 00000000
> >nt!ViDeadlockCheckStackLimits+0x6b
> >824a6d94 8063b3db 00000000 00000000 00000000
> >nt!ViDeadlockCanProceed+0x25
> >824a6db8 80538367 8276bd78 00000020 82675770
> >nt!VfDeadlockDeleteMemoryRange+0x11
> >824a6df4 8052618c 8276bd78 00000000 828add18 nt!ExFreePoolWithTag+0x95
> >824a6e24 805264e5 824a6e3c f00929ac 824a6ec0
> >nt!ExpFindCurrentThread+0x184
> >824a6e48 eecb5dad eece8028 00000000 eecb96ef
> >nt!ExAcquireResourceSharedLite+0x4f
> >WARNING: Stack unwind information not available. Following frames may be
> >
> >wrong.
> >00000410 00009902 00240000 50e00024 1c0d1c0d SAVRT+0x7dad
> >
> >
> >FOLLOWUP_IP:
> >SAVRT+7dad
> >eecb5dad 84c0 test al,al
> >
> >SYMBOL_STACK_INDEX: a
> >
> >FOLLOWUP_NAME: MachineOwner
> >
> >SYMBOL_NAME: SAVRT+7dad
> >
> >MODULE_NAME: SAVRT
> >
> >IMAGE_NAME: SAVRT.SYS
> >
> >DEBUG_FLR_IMAGE_TIMESTAMP: 3d40d9c3
> >
> >STACK_COMMAND: kb
> >
> >BUCKET_ID: 0xc4_90_SAVRT+7dad
> >
> >Followup: MachineOwner
> >---------
> >
> >1: kd> !deadlock
> >
> >Resources: 1
> >Nodes: 0
> >Threads: 29
> >
> >Nodes trimmed based on age 0.
> >Nodes trimmed based on count 0.
> >Analyze calls 0.
> >Maximum nodes searched 0.
> >
> >No deadlock verifier issues.
> >
> >
> >
> > >From: “Tony Mason”
> > >Reply-To: “Windows System Software Devs Interest List”
> > >
> > >To: “Windows System Software Devs Interest List”
> > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > >Date: Tue, 12 Oct 2004 17:54:26 -0400
> > >
> > >Try debugging the crash - start with “!analyze -v”. Just because you
> > >turn on an option and the system crashed you can’t assume that it will
> > >crash because of the option you enabled.
> > >
> > >You might also want to show us the bugcheck data. That always makes
it
> > >easier to suggest what you might have done wrong.
> > >
> > >Regards,
> > >
> > >Tony Mason
> > >Consulting Partner
> > >OSR Open Systems Resources, Inc.
> > >http://www.osr.com
> > >
> > >
> > >-----Original Message-----
> > >From: xxxxx@lists.osr.com
> > >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> > >Sent: Tuesday, October 12, 2004 4:55 PM
> > >To: ntdev redirect
> > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > >
> > >Thanks.
> > >
> > >I turn on the driver verifier with deadlock detection, and I get a bug
> > >check. Then I follow with “!deadlock” in the windbg, but I get “no
> > >deadlock
> > >verifier issues”. How can I continue from there?
> > >
> > >AH
> > >
> > > >From: “Doron Holan”
> > > >Reply-To: “Windows System Software Devs Interest List”
> > > >
> > > >To: “Windows System Software Devs Interest List”
> >
> > > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > > >Date: Mon, 11 Oct 2004 22:51:09 -0700
> > > >
> > > >You should hold onto a lock for the least amount of time as
possible.
> > > >If you hold onto the lock too long, the user will see the affects.
> > >What
> > > >might be happening on your machine is that your driver is acquiring
a
> > > >lock, but not releasing it and then return to the caller outside of
> > >your
> > > >driver. Turn on driver verifier and deadlock detection. If you
> > >think
> > > >this is what is going on, driver verifier should find it rather
> > >quickly.
> > > >
> > > >d
> > > >
> > > >-----Original Message-----
> > > >From: xxxxx@lists.osr.com
> > > >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> > > >Sent: Monday, October 11, 2004 8:18 PM
> > > >To: Windows System Software Devs Interest List
> > > >Subject: [ntdev] how long can I hold a spin lock?
> > > >
> > > >Is there any suggesting time? My driver seems to hang a HT system
> > >after
> > > >
> > > >running a while(UP is fine), and I cannot ping it. It looks like a
> > > >deadlock
> > > >to me. But I cannot break into the windbg. Any idea what to do?
> > > >
> > > >Thanks
> > > >
> > > >AH
> > > >
> > > >
> > > >Check out Election 2004 for up-to-date election news, plus voter
> >tools
> > > >and
> > > >more! http://special.msn.com/msn/election2004.armx
> > > >
> > > >
> > > >
> > > >—
> > > >Questions? First check the Kernel Driver FAQ at
> > > >http://www.osronline.com/article.cfm?id=256
> > > >
> > > >You are currently subscribed to ntdev as:
> >xxxxx@windows.microsoft.com
> > > >To unsubscribe send a blank email to
xxxxx@lists.osr.com
> > > >
> > > >
> > > >—
> > > >Questions? First check the Kernel Driver FAQ at
> > > >http://www.osronline.com/article.cfm?id=256
> > > >
> > > >You are currently subscribed to ntdev as: unknown lmsubst tag
> >argument:
> > >‘’
> > > >To unsubscribe send a blank email to
xxxxx@lists.osr.com
> > >
> > >
> > >On the road to retirement? Check out MSN Life Events for advice on how
> > >to
> > >get there! http://lifeevents.msn.com/category.aspx?cid=Retirement
> > >
> > >
> > >
> > >—
> > >Questions? First check the Kernel Driver FAQ at
> > >http://www.osronline.com/article.cfm?id=256
> > >
> > >You are currently subscribed to ntdev as: xxxxx@osr.com
> > >To unsubscribe send a blank email to xxxxx@lists.osr.com
> > >
> > >—
> > >Questions? First check the Kernel Driver FAQ at
> > >http://www.osronline.com/article.cfm?id=256
> > >
> > >You are currently subscribed to ntdev as: unknown lmsubst tag
argument:
> >‘’
> > >To unsubscribe send a blank email to xxxxx@lists.osr.com
> >
> >
> >Check out Election 2004 for up-to-date election news, plus voter tools
> >and
> >more! http://special.msn.com/msn/election2004.armx
> >
> >
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=256
> >
> >You are currently subscribed to ntdev as: xxxxx@osr.com
> >To unsubscribe send a blank email to xxxxx@lists.osr.com
> >
> >—
> >Questions? First check the Kernel Driver FAQ at
> >http://www.osronline.com/article.cfm?id=256
> >
> >You are currently subscribed to ntdev as: unknown lmsubst tag argument:
‘’
> >To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>
> Is your PC infected? Get a FREE online computer virus scan from McAfee®
> Security. http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963
>
>
>
> —
> Questions? First check the Kernel Driver FAQ at http://www.
> osronline.com/article.cfm?id=256
>
> You are currently subscribed to ntdev as: xxxxx@3dlabs.com
> To unsubscribe send a blank email to xxxxx@lists.osr.com

> ForwardSourceID:NT0000544A

OSR_Community_User · October 13, 2004, 8:57am

When I first received this email I knew that I had read a white paper from
MS on this subject. Now I have found it:

(under the title of “Windows Logo Program Issues”)
“A driver or thread cannot perform any action-such as holding a spin lock,
executing a CLI or STI instruction, or masking an interrupt that blocks the
execution of other drivers for more than 25 microseconds. Doing so would
have the same effect as running an excessively long ISR or DPC.”
http://www.microsoft.com/whdc/driver/perform/mmdrv.mspx

Granted this is related to a future WHQL requirement, but perhaps that is
all the more reason to adhere to it.

Philip Lukidis < mailto:xxxxx@guillemot.com>
Software Engineer - Guillemot R&D inc. http://www.guillemot.com
Tel: (514) 279-9960 ext:4111 Fax: (514) 279-2558
5800 St-Denis, Suite 1102, Montreal, Quebec, H2S 3L5, Canada

-----Original Message-----
From: Ta H. [mailto:xxxxx@hotmail.com]
Sent: Monday, October 11, 2004 11:18 PM
To: Windows System Software Devs Interest List
Subject: [ntdev] how long can I hold a spin lock?

Is there any suggesting time? My driver seems to hang a HT system after
running a while(UP is fine), and I cannot ping it. It looks like a deadlock

to me. But I cannot break into the windbg. Any idea what to do?

Thanks

AH

Check out Election 2004 for up-to-date election news, plus voter tools and
more! http://special.msn.com/msn/election2004.armx

Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256

You are currently subscribed to ntdev as: xxxxx@guillemot.com
To unsubscribe send a blank email to xxxxx@lists.osr.com

OSR_Community_User · October 13, 2004, 10:11am

> Mats,

Almost any driver will get that bugcheck with Symantec and driv=
er
verifier. The question is there other problems with his driver and
Symantec
or just this one.

I agree with that, and that was my point (although I may have expressed=
it
less clearly than you did!).

Of course, the deadlock detection in Driver Verifier has to be turned o=
ff
since Symantec can’t be run with it. So, run Driver Verifier with deadl=
ock
detection without Symantec, and run without Deadlock Detection when run=
ning
with Symantec, and try to figure out in other ways why the driver doesn=
't
work with Symantec (assuming it actually locks up in this mode, which I=

still believe it does).

By the way, when we use a Simulator for a new model of graphics chip, w=
e
also do stack switching. This is because the C++ code for the simulator=

needs an order of magnitude more stack than the standard kernel stack
supplies, and aside from really complicated solutions, we haven’t found=
a
way around this. It’s not that we’re storing big data structures on the=

stack, but rather that the simulator has an enormous call-depth that ki=
lls
us (sure, we’d get a bit closer to surviving if we didn’t have a load o=
f
fair-sized data structures on the stack). However, to be different from=

Symantec, this driver will never ever leave the internal testing situat=
ion
(no-one will want it anyways, as it’s orders of magnitude slower than e=
ven
the slowest cheap hardware solution).

–
Mats

–
Don Burn (MVP, Windows DDK)
Windows 2k/XP/2k3 Filesystem and Driver Consulting
Remove StopSpam from the email to reply

“Mats PETERSSON” wrote in message
> news:xxxxx@ntdev…
>
>
>
>
>
> You can try making a complaint to Symantec, but I expect that they ha=
ve a
> very good reason to violate the MS rules, or they wouldn’t be doing t=
hat.
>
> Of course, why your software isn’t working with Symantec is probably =
not
> the same reason as this bugcheck, so maybe you want to try to figure =
out
> why your driver isn’t working together with Symantec’s AV software st=
ill,
> as I suspect Symantec isn’t going to fix whatever the problem is
> themselves…
>
> Whether you want to do this or not depends on whether you think it’s
> important that your driver works together with Symantec AV software o=
r
not.
> It’s easy to write in the README-file that it doesn’t work with Syman=
tec
> AV. It is, however, much harder to convince the consumer of the softw=
are
> that it’s actually NOT YOUR FAULT (assuming it isn’t, of course), and=
may
> be even harder yet to convince the consumer that rather than not usin=
g
your
> driver, they should change the brand of AV software. Changing the AV
> software if you work in a small company/home environment is trivial, =
but
> large companies with many hundreds or thousands of users, it’s often =
a
> corporate policy that is very strict, and you need really good reason=
s to
> get an exception to this policy.
>
> With this in mind, I would recommend that you still try to fix your
driver.
> At least make some effort. I understand that it’s not trivial. And it=
may
> be a good idea to register a complaint with Symantec, and see what he=
lp
> they may be able to offer you.
>
> Also, you will want to try several different brands of AV software wi=
th
> your driver, to make sure that you’re working well with those drivers=
.
>
> So, the conclusion is that you can’t use Deadlock detection if you’re=

using
> Symantec/Norton AV, but that doesn’t mean that you haven’t got anothe=
r
bug
> somewhere…
>
> –
> Mats
>
> xxxxx@lists.osr.com wrote on 10/13/2004 04:07:16 AM:
>
> > Many thanks!
> >
> > Once I uninstall the Norton SW, everything works beautifully. I ha=
ve
> been
> > scratching my head for a week. Should I warn my customers or make =
a
> > complaint to Symantec?!
> >
> > AH
> >
> >
> > >From: “Tony Mason”
> > >Reply-To: “Windows System Software Devs Interest List”
> > >
> > >To: “Windows System Software Devs Interest List” com>
> > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > >Date: Tue, 12 Oct 2004 19:23:43 -0400
> > >
> > >Unless you’re working for Symantec, this isn’t your bug. Symantec=

DOES
> > >switch stacks.
> > >
> > >I would suggest testing without an anti-virus product installed.
> > >
> > >Regards,
> > >
> > >Tony Mason
> > >Consulting Partner
> > >OSR Open Systems Resources, Inc.
> > >http://www.osr.com
> > >
> > >
> > >-----Original Message-----
> > >From: xxxxx@lists.osr.com
> > >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> > >Sent: Tuesday, October 12, 2004 7:13 PM
> > >To: ntdev redirect
> > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > >
> > >Thanks. Please see below. It is a HP/HT machine running XP Home.=

> > >=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D
> > >Fatal System Error: 0x000000c4
> > > (0x00000090,0xF870F120,0x00000000,0x000000=
00)
> > >
> > >Break instruction exception - code 80000003 (first chance)
> > >
> > >A fatal system error has occurred.
> > >Debugger entered on first try; Bugcheck callbacks have not been
invoked.
> > >
> > >A fatal system error has occurred.
> > >
> > >Connected to Windows XP 2600 x86 compatible target, ptr64 FALSE
> > >Loading Kernel Symbols
> >
>…=
…
> > >…
> > >Loading unloaded module list
> > >…
> > >Loading User Symbols
> > >…
> >
>*************************************************************=

> > >
> > >
> > >
> > >
> > > Bugcheck Analysis
> > >
> > >
> > >
> > >
> > > *
> >
> ********************************************************************=

> > > ****
> > >
> > >Use !analyze -v to get detailed debugging information.
> > >
> > >BugCheck C4, {90, f870f120, 0, 0}
> > >
> > > ERROR: Module load completed but symbols could not be loaded f=
or
> > >SAVRT.SYS
> > >Probably caused by : SAVRT.SYS ( SAVRT+7dad )
> > >
> > >Followup: MachineOwner
> > >---------
> > >
> > >nt!RtlpBreakWithStatusInstruction:
> > >8051d778 cc int 3
> > >1: kd> !analyze -v
> >
> =

> > >
> > >
> > >
> > >
> > > Bugcheck Analysis
> > >
> > >
> > >
> > >
> > >
> >
> =

> > >
> > >
> > >DRIVER_VERIFIER_DETECTED_VIOLATION (c4)
> > >A device driver attempting to corrupt the system has been caught.
This
> > >is
> > >because the driver was specified in the registry as being suspect =
(by
> > >the
> > >administrator) and the kernel has enabled substantial checking of =
this
> > >driver.
> > >If the driver attempts to corrupt the system, bugchecks 0xC4, 0xC1=
and
> > >0xA
> > >will
> > >be among the most commonly seen crashes.
> > > Parameter 1 =3D 0x1000 … 0x1020 - deadlock verifier erro=
r
codes.
> > > Typically the code is 0x1001 (deadlock detected) a=
nd
you
> > >can
> > > issue a ‘!deadlock’ KD command to get more
information.
> > >Arguments:
> > >Arg1: 00000090, A driver switched stacks. The current stack is nei=
ther
a
> > >
> > >thread
> > > stack nor a DPC stack. Typically the driver doing this should b=
e
> > > on the stack obtained from `kb’ command.
> > >Arg2: f870f120
> > >Arg3: 00000000
> > >Arg4: 00000000
> > >
> > >Debugging Details:
> > >------------------
> > >
> > >
> > >BUGCHECK_STR: 0xc4_90
> > >
> > >DEFAULT_BUCKET_ID: DRIVER_FAULT
> > >
> > >LAST_CONTROL_TRANSFER: from 804f444e to 8051d778
> > >
> > >STACK_TEXT:
> > >824a693c 804f444e 00000003 824a6c6c 00000090
> > >nt!RtlpBreakWithStatusInstruction
> > >824a6988 804f4ed3 00000003 8276bd70 00000000
> > >nt!KiBugCheckDebugBreak+0x19
> > >824a6d54 804f5471 000000c4 00000090 f870f120 nt!KeBugCheck2+0x46d
> > >824a6d74 8063ab77 000000c4 00000090 f870f120 nt!KeBugCheckEx+0x19
> > >824a6d90 8063ae47 8063b3db 00000000 00000000
> > >nt!ViDeadlockCheckStackLimits+0x6b
> > >824a6d94 8063b3db 00000000 00000000 00000000
> > >nt!ViDeadlockCanProceed+0x25
> > >824a6db8 80538367 8276bd78 00000020 82675770
> > >nt!VfDeadlockDeleteMemoryRange+0x11
> > >824a6df4 8052618c 8276bd78 00000000 828add18 nt!ExFreePoolWithTag+=
0x95
> > >824a6e24 805264e5 824a6e3c f00929ac 824a6ec0
> > >nt!ExpFindCurrentThread+0x184
> > >824a6e48 eecb5dad eece8028 00000000 eecb96ef
> > >nt!ExAcquireResourceSharedLite+0x4f
> > >WARNING: Stack unwind information not available. Following frames =
may
be
> > >
> > >wrong.
> > >00000410 00009902 00240000 50e00024 1c0d1c0d SAVRT+0x7dad
> > >
> > >
> > >FOLLOWUP_IP:
> > >SAVRT+7dad
> > >eecb5dad 84c0 test al,al
> > >
> > >SYMBOL_STACK_INDEX: a
> > >
> > >FOLLOWUP_NAME: MachineOwner
> > >
> > >SYMBOL_NAME: SAVRT+7dad
> > >
> > >MODULE_NAME: SAVRT
> > >
> > >IMAGE_NAME: SAVRT.SYS
> > >
> > >DEBUG_FLR_IMAGE_TIMESTAMP: 3d40d9c3
> > >
> > >STACK_COMMAND: kb
> > >
> > >BUCKET_ID: 0xc4_90_SAVRT+7dad
> > >
> > >Followup: MachineOwner
> > >---------
> > >
> > >1: kd> !deadlock
> > >
> > >Resources: 1
> > >Nodes: 0
> > >Threads: 29
> > >
> > >Nodes trimmed based on age 0.
> > >Nodes trimmed based on count 0.
> > >Analyze calls 0.
> > >Maximum nodes searched 0.
> > >
> > >No deadlock verifier issues.
> > >
> > >
> > >
> > > >From: “Tony Mason”
> > > >Reply-To: “Windows System Software Devs Interest List”
> > > >
> > > >To: “Windows System Software Devs Interest List”

> > > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > > >Date: Tue, 12 Oct 2004 17:54:26 -0400
> > > >
> > > >Try debugging the crash - start with “!analyze -v”. Just becaus=
e
you
> > > >turn on an option and the system crashed you can’t assume that i=
t
will
> > > >crash because of the option you enabled.
> > > >
> > > >You might also want to show us the bugcheck data. That always m=
akes
> it
> > > >easier to suggest what you might have done wrong.
> > > >
> > > >Regards,
> > > >
> > > >Tony Mason
> > > >Consulting Partner
> > > >OSR Open Systems Resources, Inc.
> > > >http://www.osr.com
> > > >
> > > >
> > > >-----Original Message-----
> > > >From: xxxxx@lists.osr.com
> > > >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> > > >Sent: Tuesday, October 12, 2004 4:55 PM
> > > >To: ntdev redirect
> > > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > > >
> > > >Thanks.
> > > >
> > > >I turn on the driver verifier with deadlock detection, and I get=
a
bug
> > > >check. Then I follow with “!deadlock” in the windbg, but I get =
“no
> > > >deadlock
> > > >verifier issues”. How can I continue from there?
> > > >
> > > >AH
> > > >
> > > > >From: “Doron Holan”
> > > > >Reply-To: “Windows System Software Devs Interest List”
> > > > >
> > > > >To: “Windows System Software Devs Interest List”
> > >
> > > > >Subject: RE: [ntdev] how long can I hold a spin lock?
> > > > >Date: Mon, 11 Oct 2004 22:51:09 -0700
> > > > >
> > > > >You should hold onto a lock for the least amount of time as
> possible.
> > > > >If you hold onto the lock too long, the user will see the affe=
cts.
> > > >What
> > > > >might be happening on your machine is that your driver is
acquiring
> a
> > > > >lock, but not releasing it and then return to the caller outsi=
de
of
> > > >your
> > > > >driver. Turn on driver verifier and deadlock detection. If=
you
> > > >think
> > > > >this is what is going on, driver verifier should find it rathe=
r
> > > >quickly.
> > > > >
> > > > >d
> > > > >
> > > > >-----Original Message-----
> > > > >From: xxxxx@lists.osr.com
> > > > >[mailto:xxxxx@lists.osr.com] On Behalf Of Ta H.
> > > > >Sent: Monday, October 11, 2004 8:18 PM
> > > > >To: Windows System Software Devs Interest List
> > > > >Subject: [ntdev] how long can I hold a spin lock?
> > > > >
> > > > >Is there any suggesting time? My driver seems to hang a HT sy=
stem
> > > >after
> > > > >
> > > > >running a while(UP is fine), and I cannot ping it. It looks l=
ike
a
> > > > >deadlock
> > > > >to me. But I cannot break into the windbg. Any idea what to =
do?
> > > > >
> > > > >Thanks
> > > > >
> > > > >AH
> > > > >
> > > > > _________________________________________________________=

> > > > >Check out Election 2004 for up-to-date election news, plus vot=
er
> > >tools
> > > > >and
> > > > >more! http://special.msn.com/msn/election2004.armx
> > > > >
> > > > >
> > > > >
> > > > >—
> > > > >Questions? First check the Kernel Driver FAQ at
> > > > >http://www.osronline.com/article.cfm?id=3D256
> > > > >
> > > > >You are currently subscribed to ntdev as:
> > >xxxxx@windows.microsoft.com
> > > > >To unsubscribe send a blank email to
> xxxxx@lists.osr.com
> > > > >
> > > > >
> > > > >—
> > > > >Questions? First check the Kernel Driver FAQ at
> > > > >http://www.osronline.com/article.cfm?id=3D256
> > > > >
> > > > >You are currently subscribed to ntdev as: unknown lmsubst tag
> > >argument:
> > > >‘’
> > > > >To unsubscribe send a blank email to
> xxxxx@lists.osr.com
> > > >
> > > > =

> > > >On the road to retirement? Check out MSN Life Events for advice =
on
how
> > > >to
> > > >get there! http://lifeevents.msn.com/category.aspx?cid=3DRetirem=
ent
> > > >
> > > >
> > > >
> > > >—
> > > >Questions? First check the Kernel Driver FAQ at
> > > >http://www.osronline.com/article.cfm?id=3D256
> > > >
> > > >You are currently subscribed to ntdev as: xxxxx@osr.com
> > > >To unsubscribe send a blank email to
xxxxx@lists.osr.com
> > > >
> > > >—
> > > >Questions? First check the Kernel Driver FAQ at
> > > >http://www.osronline.com/article.cfm?id=3D256
> > > >
> > > >You are currently subscribed to ntdev as: unknown lmsubst tag
> argument:
> > >‘’
> > > >To unsubscribe send a blank email to
xxxxx@lists.osr.com
> > >
> > >
> > >Check out Election 2004 for up-to-date election news, plus voter t=
ools
> > >and
> > >more! http://special.msn.com/msn/election2004.armx
> > >
> > >
> > >
> > >—
> > >Questions? First check the Kernel Driver FAQ at
> > >http://www.osronline.com/article.cfm?id=3D256
> > >
> > >You are currently subscribed to ntdev as: xxxxx@osr.com
> > >To unsubscribe send a blank email to xxxxx@lists.osr.=
com
> > >
> > >—
> > >Questions? First check the Kernel Driver FAQ at
> > >http://www.osronline.com/article.cfm?id=3D256
> > >
> > >You are currently subscribed to ntdev as: unknown lmsubst tag
argument:
> ‘’
> > >To unsubscribe send a blank email to xxxxx@lists.osr.=
com
> >
> > _________________________________________________________________
> > Is your PC infected? Get a FREE online computer virus scan from McA=
fee=AE
> > Security. http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3D3=
963
> >
> >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at http://www.
> > osronline.com/article.cfm?id=3D256
> >
> > You are currently subscribed to ntdev as: xxxxx@3dlabs.com=

> > To unsubscribe send a blank email to xxxxx@lists.osr.c=
om
>
> > ForwardSourceID:NT0000544A
>
>
>
>
>
> —
> Questions? First check the Kernel Driver FAQ at http://www.
> osronline.com/article.cfm?id=3D256
>
> You are currently subscribed to ntdev as: xxxxx@3dlabs.com
> To unsubscribe send a blank email to xxxxx@lists.osr.com=

> ForwardSourceID:NT0000549E=