On Fri, Sep 16, 2011 at 09:42:16AM +0300, Péter Ujfalusi wrote:
> On Thursday 15 September 2011 23:27:20 Dmitry Torokhov wrote:
> > This is not very helpful for the application trying to use the device.
> > Maybe we should zap the device when channels are routed through audio
> > instead of dripping requests?
> 
> Yes, I have thought about that as well, but it can cause other problems.
> I have modified the fftest tool from:
> http://www.koders.com/c/fidF60D5962FCA8B937A6D0947AA81AE95A8C58FB36.aspx
> for my needs.
> It opens the /dev/input/eventX, and it will keep it open as long as it is 
> running.
> It is OK for me, since the twl6040-vibra goes to low power mode when we do not 
> have effects running.
> I can imagine, that real world application will do the same, and that might 
> cause some problems, if I try to zap the device.
> Can we zap a device, while it is open?
> How application will handle that event?

The application will get an error on read or write, like -ENODEV, and will
hopefully "drop off".

-- 
Dmitry
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/