Hi,

I'm looking for a way to get reasonable 1-microsecond accurate timing
in a VxD (9x/Me) on one hand, and in a kernel-mode driver (NT/2000/XP)
on the other hand.

What I need is a way to measure some delays inside the driver. I don't
need to "schedule" anything. So it's just a measurement.

Re: microsecond precision timing in VxD's and kernel drivers? by Josh

Josh
Fri Oct 24 13:42:21 CDT 2003

I think you'll want to look at KeQueryPerformanceCounter.

"Guillaume" <grs-N-SPAM@NO-mail.com> wrote in message
news:uWO5U0lmDHA.3316@tk2msftngp13.phx.gbl...
> Hi,
>
> I'm looking for a way to get reasonable 1-microsecond accurate timing
> in a VxD (9x/Me) on one hand, and in a kernel-mode driver (NT/2000/XP)
> on the other hand.
>
> What I need is a way to measure some delays inside the driver. I don't
> need to "schedule" anything. So it's just a measurement.
>



Re: microsecond precision timing in VxD's and kernel drivers? by Guillaume

Guillaume
Fri Oct 24 14:01:40 CDT 2003

> I think you'll want to look at KeQueryPerformanceCounter.

Thanks.

I think this isn't an option in a VxD, though?


Re: microsecond precision timing in VxD's and kernel drivers? by Josh

Josh
Fri Oct 24 14:05:43 CDT 2003

I don't know. I do know it works in a driver on XP. If it isn't available
in a VxD, maybe you could use the user mode equivalent,
QueryPerformanceCounter.

"Guillaume" <grs-N-SPAM@NO-mail.com> wrote in message
news:OHxr$DmmDHA.744@tk2msftngp13.phx.gbl...
> > I think you'll want to look at KeQueryPerformanceCounter.
>
> Thanks.
>
> I think this isn't an option in a VxD, though?
>



Re: microsecond precision timing in VxD's and kernel drivers? by heinz_baer

heinz_baer
Sun Oct 26 22:53:29 CST 2003

> I think this isn't an option in a VxD, though?

Correct. Use VTD_Get_Real_Time