WHQL: for usbccgp.sys filter driver on windows 7

NTDEV
1

Hi,
I am running WHQL tests on a USB composite filter driver (Filter driver for changing usbccgp enumeration behavior).
DTM WLK 1.5,
Windows 7 32 Bit
FAILED TEST:Device Install Check for Other Device Stability.
Message in DTM Log: xyz device added newly / xyz device removed

I have a filter driver for usbccgp.sys to enumerate custom PDO(s).
(http://msdn.microsoft.com/en-us/library/aa476427.aspx)

The test results are same even if child function drivers are installed for the PDO or not.

As per the http://msdn.microsoft.com/en-us/library/dd834196.aspx the Sub test Run DeviceScan.exe with Diff Mode ought to fail if new devices are found (But filter driver along with usbccgp is doing this).

I installed usbccgp.sys alone for my device without filter driver, And the results were same ,Device Install Check for Other Device Stability failed.

Is there any errata/filter available for this?

To WHQL this type of filter driver (which enumerates custom PDOs) which is correct way to run :
a) install drivers for child devices also.
b)Drivers for child devices need not be installed, filter driver alone is enough.

any help?

Thanks in Advance

2

On the winqual site you can browse the exisitng filters. I was just
doing that for a similar issue. Though in my case it inolved a smart
card driver. I guess that the xyz device is yours. The issue might be
that you have a new device instance id and hence the test thinks that
a new device was added. Can you share what xyz is? It will give more
insight if this is the case with the instance id failure.

Cheers
Faik

3

Hi,

thanks for reply.

“xyz” is the the child PDO’s created by the usbccgp (with support of filter driver) on one system.

On another system “xyz” is HTREE\ROOT\0 or some other driver.

did your issue got solved?

4

Hi,

your xyz of the removed device and xyz of the added newly device
should be different, and I suspect that they are different, this is
the reason I wanted to know what xyz was. Anyway if you have a
submission allready then just update the filters and rerun the test,
the filters will probably mask out the issue.

/Faik

5

Hi,

xyz of the removed device and xyz of the added newly device are same.
I updated the errata/filter for WLK 1.5 and re run the tests but the problem prevails.

Is it a hardware problem

Can some guys from microsoft WHQL team throw some light on it.

http://msdn.microsoft.com/en-us/library/dd834196.aspx page says that

Run DeviceScan.exe with Diff Mode.
Fail Condition: At least one mismatch in the device status / missing / new devices found.

A bus driver will most probably enumerate some new child device, so they will never pass a WHQL.
then how come usbccgp.sys got certification?
Do i got this incorrect or overlooking something?

pleas update the errata/filters .

6

There is an auto-triage filter for this that directs you to a manual errata id. Refer to filter #1860 on winqual (https://winqual.microsoft.com/). It should be part of the latest filters cab as well.

7

thanks for replies.

I updated the filters ( #1860 is part of latest cab), still device stability is failing.

let me try to make it more clear.

I have a usb filter driver for usbccgp.sys which enumerates child devices (after combining some interfaces).

  1. I installed the filter driver only (and not installed drivers for child devices)and started WHQL test , Device Install Check for Other Device Stability.
    test fails, and when status is checked filter #1860 is applied but still it is counted as failed test.
    device instance Id for the child devices “added newly” and “removed” are same.

  2. if I re run the test with drivers for child devices installed situation is same except for one child driver device instance id is different and for others it is same in “added newly” and “removed”.

unable to figure what is going wrong

8

JItender Singh wrote:

I updated the filters ( #1860 is part of latest cab), still device
stability is failing.

Did you try without your filter in place? Does the test then pass?

9

yes, without filter driver the drivers for child device pass the test.

and if I install drivers for child devices over usbccgp.sys, test fails.

10

JItender Singh wrote:

yes, without filter driver the drivers for child device pass the
test.

and if I install drivers for child devices over usbccgp.sys, test
fails.

So just to clarify, did it occur to you that the problem is actually in your filter and that hunting for DTM errata’s is probably a non-starter? What exactly are you doing in your filter? What enumeration behaviour are you trying to change?

11

This was the first result, I also deduces from observation, that there might be a problem with my filter driver only.

Then to clarify it I installed usbccgp for my composite device, and let the usbccgp.sys to enumerate the child devices and then I run the test (“device stability”) on usbccgp.sys (both with and w/o child drivers installed) and the results was same … it also failed in “device stability” test.

In my filter driver I am combining two of the available interfaces into a single PDO.
(http://www.osronline.com/showThread.cfm?link=168755)

So now I am unable to locate where the problem is? :frowning: