Look for some help about Windows pv driver

Look for some help about Windows pv driver

[fanliang]

[-- Attachment #1.1: Type: text/plain, Size: 477 bytes --]

Hi:
Is there any body have used GPL Windows pv driver ? Or WLK tools ? After I
installed GPL windows pv driver in my windows vm, I run WLK for a while, and
a few minutes later, I got a BSoD with error code STOP0X000000F4 . Used
windbg to debug, I got some call stack about  the system proccess .It is
strange that the call stack include nothing about the pv driver. I'm
wandering how can I fix this problem ? Many thanks if some one gave me some
suggestions. 
Tks. 
Lyman Van


[-- Attachment #1.2: Type: text/html, Size: 1056 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Re: Look for some help about Windows pv driver

[Paul Durrant]

fanliang wrote:
> Is there any body have used GPL Windows pv driver ? Or WLK tools ? After 
> I installed GPL windows pv driver in my windows vm, I run WLK for a 
> while, and a few minutes later, I got a BSoD with error code 
> STOP0X000000F4 . Used windbg to debug, I got some call stack about  the 
> system proccess .It is strange that the call stack include nothing about 
> the pv driver. I'm wandering how can I fix this problem ? Many thanks if 
> some one gave me some suggestions.

First, have a look at 
http://msdn.microsoft.com/en-us/library/ms789516.aspx and look up the 
bugcheck code. In your case it's CRITICAL_OBJECT_TERMINATION. If you 
look at http://msdn.microsoft.com/en-us/library/ms797140.aspx you'll see 
that parameter 1 of the bugcheck tells you the object type (process or 
thread), parameter 2 points to the object, parameter 3 points to the 
process image file name and parameter 4 points to an ASCII string 
telling you what happened.
If you didn't already, then run !analyze -v inside windbg as it will 
also tell you most of this and decode some of the info for you.

   Paul


-- 
===============================
Paul Durrant, Software Engineer

Citrix Systems (R&D) Ltd.
First Floor, Building 101
Cambridge Science Park
Milton Road
Cambridge CB4 0FY
United Kingdom
===============================

答复: [Xen-devel] Look for some help about Windows pv driver

[fanliang]

Thanks a lot. 
I can understand what you said , But what puzzled me is that I saw nothing
about my windows pv driver from the output of   !analyze -v inside windbg .I
am not sure that  the BSoD had nothing with the windows pv driver.  WLK goes
very well in other windows system without pv driver ,so I guess the pv
driver cases the BSod. 


 

-----邮件原件-----
发件人: Paul Durrant [mailto:paul.durrant@citrix.com] 
发送时间: 2009年12月22日 17:38
收件人: fanliang
抄送: xen-devel@lists.xensource.com
主题: Re: [Xen-devel] Look for some help about Windows pv driver

fanliang wrote:
> Is there any body have used GPL Windows pv driver ? Or WLK tools ? 
> After I installed GPL windows pv driver in my windows vm, I run WLK 
> for a while, and a few minutes later, I got a BSoD with error code
> STOP0X000000F4 . Used windbg to debug, I got some call stack about  
> the system proccess .It is strange that the call stack include nothing 
> about the pv driver. I'm wandering how can I fix this problem ? Many 
> thanks if some one gave me some suggestions.

First, have a look at
http://msdn.microsoft.com/en-us/library/ms789516.aspx and look up the
bugcheck code. In your case it's CRITICAL_OBJECT_TERMINATION. If you look at
http://msdn.microsoft.com/en-us/library/ms797140.aspx you'll see that
parameter 1 of the bugcheck tells you the object type (process or thread),
parameter 2 points to the object, parameter 3 points to the process image
file name and parameter 4 points to an ASCII string telling you what
happened.
If you didn't already, then run !analyze -v inside windbg as it will also
tell you most of this and decode some of the info for you.

   Paul


-- 
===============================
Paul Durrant, Software Engineer

Citrix Systems (R&D) Ltd.
First Floor, Building 101
Cambridge Science Park
Milton Road
Cambridge CB4 0FY
United Kingdom
===============================

Re: 答复: [Xen-devel] Look for some help about Windows pv driver

[Paul Durrant]

fanliang wrote:
> Thanks a lot. 
> I can understand what you said , But what puzzled me is that I saw nothing
> about my windows pv driver from the output of   !analyze -v inside windbg .I
> am not sure that  the BSoD had nothing with the windows pv driver.  WLK goes
> very well in other windows system without pv driver ,so I guess the pv
> driver cases the BSod. 
> 

Just because you did not get a direct stack backtrace into a PV driver
does not, of course, mean it's not at fault. You need to find out the
exact nature of the object that terminated to have a clue as to what
happened.

  Paul

-- 
===============================
Paul Durrant, Software Engineer

Citrix Systems (R&D) Ltd.
First Floor, Building 101
Cambridge Science Park
Milton Road
Cambridge CB4 0FY
United Kingdom
===============================

RE: Look for some help about Windows pv driver

[James Harper]

> 
> Hi:
> Is there any body have used GPL Windows pv driver ? Or WLK tools ?

I've run the NDIS test a while ago (a seriously good network test
suite!) to test some of the IP offload code I'd been working on. I
haven't run much of the other stuff though.

> After I
> installed GPL windows pv driver in my windows vm, I run WLK for a
while, and a
> few minutes later, I got a BSoD with error code STOP0X000000F4.

What version of GPLPV?

Parameter 4 of BSoD should be a pointer to a string with some more
information... can you tell me what the string contains? I assume it's
just the name of the task that crashed which won't tell us much.

> Used windbg
> to debug, I got some call stack about  the system proccess .It is
strange that
> the call stack include nothing about the pv driver. I'm wandering how
can I
> fix this problem ? Many thanks if some one gave me some suggestions.

Do you know what was being tested? If it was the disk then maybe there
is some corruption happening - the WLK is supposed to give things a
pretty extreme workout so it could be touching some corner case that
we've never seen in real use. One thing I can think of is how GPLPV
behaves when allocations to system resources fail (eg memory
allocation)... if I don't know how to handle a failure at a specific
point I should have at least put an ASSERT in there so that you get a
definite crash right at the point of failure, not a strange crash some
time later, but maybe I've missed some?

If you are not running a fairly recent version of GPLPV then there was a
bug that only ever showed up under Call Of Duty 4 in server mode on one
specific map that was caused by some buffer corruption on disk
read/write.

Depending on your version of Dom0 and the version of GPLPV, there should
be debug information written out to /var/log/xen/qemu-dm-<domu
name>.log. There might be something useful in there.

If you can reproduce the problem pretty quickly, I can make a very very
new version of GPLPV available to you for testing. I haven't tested it
enough for production use yet but it would be useful to run it through
the WLK. It uses the latest WinDDK (7600).

Also, if you want the GPLPV symbol files for debugging then I have them
for any given recent release, I just don't upload them unless someone
wants them as they are big and I don't have a particularly fast link, so
just ask if you want them.

James