The first step should be to get crashdump from customer and analyze it as Mark suggested. In most cases it is enough to identify problem and you may not need similar system at all. As a customer I really hate companies which need to reproduce problem themselves in order to make fix; it is quite unprofessional approach. As a developer I try to identify problem from the info customer makes which involves sending special driver versions to customer in the worst case. Of course, I don’t mean problems easily reproducible at any machine which is the simplest possibility.
As for the differences, if your driver runs under XP at different board, I’d expect a bug related to timing. Usually race conditions. Trying to reproduce such problem on different machine, however similar, may be very sad task because it may not be reproducible at all. Last year I had a deadlock problem which was reproducible on one customers machine only although we had equal machine here. It took a week of testing at customer side with several specially customized driver versions (with debug version and traces problem disappeared) and finally I found race conditions with about 10 us window.
BTW, what is the next what is the next bugcheck parameter after 0x2? It should be PDO address which caused bugcheck. From WinDbg help:
Invalid PDO: An API which requires a PDO has been called with random memory, or with an FDO, or with a PDO which hasn’t been initialized.
Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]
From: xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com] on behalf of Russell Poffenberger[SMTP:xxxxx@credence.com]
Reply To: Windows System Software Devs Interest List
Sent: Wednesday, June 15, 2005 7:26 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Driver works on Win2K, gives STOP 0x000000CA on XP
Thanks Mark,
I will do that once I have a system I can try it on, but until then (due to
a move of our engineering offices to another location in the next week, and
the red tape involved in purchasing a system similar to the customer’s, I
won’t be able to get access to a system that shows the problem for at least
2 to 3 weeks), I was hoping someone might be able to point out any
differences between W2K and XP that might explain it.
At 10:10 AM 6/15/2005, you wrote:
>The output from windbg ‘!analyze -v’ would help. Also - try running with
>verifier on.
>
>-----Original Message-----
>From: xxxxx@lists.osr.com
>[mailto:xxxxx@lists.osr.com] On Behalf Of Russell
>Poffenberger
>Sent: Wednesday, June 15, 2005 12:51 PM
>To: Windows System Software Devs Interest List
>Subject: [ntdev] Driver works on Win2K, gives STOP 0x000000CA on XP
>
>
>Hi,
>
>I have a driver that has worked fine on Windows 2000 (developed using
>Driver Works and Windows 2000 DDK) for some time (at least a year) on
>multiple platforms including dual CPU (P4, Xeon) and hyperthreading, and
Russ Poffenberger
Credence Systems Corp.
xxxxx@credence.com
Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256
You are currently subscribed to ntdev as: xxxxx@upek.com
To unsubscribe send a blank email to xxxxx@lists.osr.com