Asynchronous IO problems on Win7

NTDEV
1

Hi! I am in the process of digging into the KMDF. Its been great so far,
however there is a mystery in the 1394 vdev_hybrid driver.

This is a duplicate that did not appear to get resolved. (would link it, but
searching this forum for “ASYNC_READ” will get you there.

In the kmdf sample for the 1394 driver, asyncrhonous io thru the kmdf1394_SubmitIrpAsyncCompletion call does not appear to work.
It returns STATUS_INVALID_PARAMETER.

If you create a duplicate WDFREQUEST it does work.

However, when attempting to release objects and keep it clean, the system crashes…
Since I am somewhat new to KMDF what is the correct way to handle this issue - apparently with WdfIoTargetFormatRequestForInternalIoctlOthers on a “reused” WDFREQUEST.

Thanks!

2

Hi Everyone,

Is this a simple question? I have a sense from other posts that it has already been solved, I just can’t find the solution.

The code follows the vdev_hybrid driver closely. In a prior post, It was asserted that this may be a 1394 bug.

Thanks!

3

Would be a great topic for an NTINSIDER! That whole wdrcreate for something else because something is broken inside… kind of like a hole in my head… (oh please someone from the other coast…your words are welcome)

Just confirmed that the ASYNC_READ and ASYNC_WRITE operations work great if I use the synchronous interface… I can even write patterns to the hardware and read them back.

Any insight if there are bugs in the KMDF 1394ochi stack related to this? (windows 7)

do I sound desperate?

For any insight there is much gratitude!

4

What exactly do you want from me? Just because the stack was written in kmdf dies not mean I am familiar with its ins and outs (and bugs)

d

dent from a phine with no keynoard

-----Original Message-----
From: xxxxx@schange.com
Sent: Monday, March 14, 2011 7:17 PM
To: Windows System Software Devs Interest List
Subject: RE:[ntdev] Asynchronous IO problems on Win7

Would be a great topic for an NTINSIDER! That whole wdrcreate for something else because something is broken inside… kind of like a hole in my head… (oh please someone from the other coast…your words are welcome)

Just confirmed that the ASYNC_READ and ASYNC_WRITE operations work great if I use the synchronous interface… I can even write patterns to the hardware and read them back.

Any insight if there are bugs in the KMDF 1394ochi stack related to this? (windows 7)

do I sound desperate?

For any insight there is much gratitude!

NTDEV is sponsored by OSR

For our schedule of WDF, WDM, debugging and other seminars visit:
http://www.osr.com/seminars

To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer

5

Hi!

You answered my question. Its not fixed… (you could have said “oh that… look at this hotfix…”)

I’m still trying to get my head around the WDFREQUEST - IRP relationship… you and those on your blog as well as this blog have produced some fantastic work with regard to the KMDF insight. Still reading lots.

Sorry for the scramble on my part. Really appreciate what you are doing.

6

> Any insight if there are bugs in the KMDF 1394ochi stack related to this? (windows 7)

It was already noticed to have bugs. Try replacing it with legacy stack (Win7 supports this).

–
Maxim S. Shatskih
Windows DDK MVP
xxxxx@storagecraft.comhttp://www.storagecraft.com

7

There is a desire to get away from the legacy driver.

Does anyone know if the legacy driver supports the faster speeds and other features that are new to the windows 7 kmdf implementation?

Does anyone at microsoft who is involved in the 1394ochi project know when an update or hotfix is coming out?

Thanks again!