Hi
We are designing an embedded system that sends data on USB link to PC
running windows XP. The data is sent every 4 milliseconds.

Is there a way we could ensure that the USB client driver is never 'starved'
of processor time slice? We want to make sure that we never loose data.

Thanks,
Shani

The above statement does not make sense in itself - the very term "CPU
startvation" applies only to threads, because scheduling is done on thread
basis,
and, unless you mean dedicated threads that have been created by your driver
and/or workitem routine that runs in context of some system thread, your
driver's code can run in context of any thread by the very definition....

Anton Bassov

"Shani" wrote:

nothing with relation to thread or process priorites will help you. In the
driver you should make sure that there are always pending transfers on the
endpoint (this translates to multiple IRP+URB pairs pending at once). this
way when the device has data to send, there is a transfer on the host
waiting for it. If you use KMDF to write your driver, this is taken care of
for you by using the continuous reader.

d

--
Please do not send e-mail directly to this alias. this alias is for
newsgroup purposes only.
This posting is provided "AS IS" with no warranties, and confers no rights.


"Shani" <Shani@discussions.microsoft.com> wrote in message
news:76BBA19D-F17A-47B9-9818-23A6E5C2DB84@microsoft.com...

Thanks to you all for your help.

One thing of which I am still unclear as to what gets priority at kernel
level. I am a novice in the world of driver/application development for
Windows. So if any could please give me pointers to articles/tutorials
describing how things happen at kernel level or user level modes, I will
greatly appreciate.

Thanks,
Shani



"Doron Holan [MSFT]" wrote:

The OSR class on developing drivers covers this subject.

"Shani" <Shani@discussions.microsoft.com> wrote in message
news:5E5198D9-63CB-4508-A098-FCD3F732C8D1@microsoft.com...