PB documentation says kernel calls oemidle if it doesn't have any more thread
to schedule.
But when I actually test it, it seems not like documentation.
I ran a wmp playing some contents and at the moment I could see oemidle code
has been called simultaneously. ( I added a serial message within the oemidle
code from platform directory )
Is anybody who can clear this question?

Re: oemidle()... by Valter

Valter
Mon Apr 14 04:28:32 PDT 2008

=?Utf-8?B?ZGFuaWVs?= <daniel@discussions.microsoft.com> wrote in
news:B983370B-E1A2-4A2E-83CD-734AE45F568D@microsoft.com:

> PB documentation says kernel calls oemidle if it doesn't have any
> more thread to schedule.
> But when I actually test it, it seems not like documentation.
> I ran a wmp playing some contents and at the moment I could see
> oemidle code has been called simultaneously. ( I added a serial
> message within the oemidle code from platform directory )
> Is anybody who can clear this question?

It may be that the player does not need 100% of you CPU to perform
playback and this leads to calls to OEMIdle.
To keep your CPU from being idle you may use this code:

for (;;);

Even running at the lowest priority level this code will eat all the
available CPU time, preventing the system from entering idle state.


--
Valter Minute
www.fortechembeddedlabs.it
Training, support and development for Windows CE
(the reply address of this message is invalid)

Re: oemidle()... by Chris

Chris
Mon Apr 14 11:05:42 PDT 2008

Providing the code was optimized of course. A Release build would likely
strip that code.


--

Chris Tacke, Embedded MVP
OpenNETCF Consulting
Giving back to the embedded community
http://community.OpenNETCF.com

"Valter Minute" <v_a_l_t_e_r.m_i_n_u_t_e@g_m_a_i_l.com> wrote in message
news:Xns9A808913F8060VALTERMINUTE@207.46.248.16...
> =?Utf-8?B?ZGFuaWVs?= <daniel@discussions.microsoft.com> wrote in
> news:B983370B-E1A2-4A2E-83CD-734AE45F568D@microsoft.com:
>
>> PB documentation says kernel calls oemidle if it doesn't have any
>> more thread to schedule.
>> But when I actually test it, it seems not like documentation.
>> I ran a wmp playing some contents and at the moment I could see
>> oemidle code has been called simultaneously. ( I added a serial
>> message within the oemidle code from platform directory )
>> Is anybody who can clear this question?
>
> It may be that the player does not need 100% of you CPU to perform
> playback and this leads to calls to OEMIdle.
> To keep your CPU from being idle you may use this code:
>
> for (;;);
>
> Even running at the lowest priority level this code will eat all the
> available CPU time, preventing the system from entering idle state.
>
>
> --
> Valter Minute
> www.fortechembeddedlabs.it
> Training, support and development for Windows CE
> (the reply address of this message is invalid)



Re: oemidle()... by Valter

Valter
Tue Apr 15 02:49:47 PDT 2008

"Chris Tacke, eMVP" <ctacke.at.opennetcf.dot.com> wrote in
news:#CZlgolnIHA.4536@TK2MSFTNGP06.phx.gbl:

> Providing the code was optimized of course. A Release build would
> likely strip that code.

You are right. I put a lot of infinite loops in my code by mistake,
but never tried to build that specific one :)


--
Valter Minute
www.fortechembeddedlabs.it
Training, support and development for Windows CE
(the reply address of this message is invalid)