Am Mittwoch, 16. Mai 2007 15:41 schrieb Jiri Kosina:
> On Wed, 16 May 2007, Pete Zaitcev wrote:
>
> > I did not verify the function in detail, but the patch looks sane at a
> > quick glance.
>
> Hi Pete and Oliver,
>
> (BTW, why was I again dropped from CC? :) )
>
> I consider the things below crucial before this could be merged:
>
> - what to do with devices with force feedback? (or PID devices in
> general). The effect could have up to infinite duration (until it is
> stopped by appropriate "stop" method), and we definitely absolutely must
> not autosuspend a device that is currently in a process of playback.
I suggest we don't suspend them. We'd increment the pm usage count on
open and decrement it on close, basically the same we need to do for
generic HID devices. I just need a way to identify PID devices.
> - think again about the output reports queuing. I am now inclined to think
> that we should simply wake the device up once the output report is to be
> delivered to it. There might be different situations other than just
> keyboard LEDs (there can be simply any kind of exotic HID device being
> controlled through hiddev and userspace could want to deliver the output
> report to it immediately, without any queuing)
If we do "busy style" autosuspend only for devices we positively identify,
is the point rendered moot? It seems to me waking needs modifications
to be safe, as freezing the tasks does not prevent io for HID devices.
> - (and of course coding style)
It shall be done, but not until the principal features are clear.
Regards
Oliver
-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
linux-usb-devel@lists.sourceforge.net
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
