On Wed, Jun 15, 2011 at 3:55 PM, Hans Verkuil <hverkuil@xxxxxxxxx> wrote: > Why would that violate the spec? If the last filehandle is closed, then > you can safely poweroff the tuner. The only exception is when you have a radio > tuner whose audio output is hooked up to some line-in: there you can't power off > the tuner. The use case that some expect to work is: v4l2-ctl <set standard> v4l2-ctl <set frequency> cat /dev/video0 > out.mpg By powering off the tuner after v4l2-ctl closes the device node, the cat won't work as expected because the tuner will be powered down. >> We've been forced to choose between the purist perspective, which is >> properly preserving state, never powering down the tuner and sucking >> up 500ma on the USB port when not using the tuner, versus powering >> down the tuner when the last party closes the filehandle, which >> preserves power but breaks v4l2 conformance and in some cases is >> highly noticeable with tuners that require firmware to be reloaded >> when being powered back up. > > Seems fine to me. What isn't fine is that a driver like e.g. em28xx powers off > the tuner but doesn't power it on again on the next open. It won't do that > until the first S_FREQUENCY/S_TUNER/S_STD call. You don't want to power up the tuner unless you know the user intends to use it. For example, you don't want to power up the tuner if the user intends to capture on composite/s-video input (as this consumes considerably more power). Devin -- Devin J. Heitmueller - Kernel Labs http://www.kernellabs.com -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
