On 02/16/2015 11:50 AM, Mauro Carvalho Chehab wrote: > Em Mon, 16 Feb 2015 10:55:50 +0100 > Hans Verkuil <hverkuil@xxxxxxxxx> escreveu: > >> On 02/15/2015 11:27 AM, Mauro Carvalho Chehab wrote: > >>> In any case, for ALSA, we should do the right thing here: remove (actually >>> deprecate) whatever definition is there, and then re-add it only when we >>> actually have the patches inside the ALSA subsystem to support the media >>> controller, plus having the corresponding patches for the media-ctl in order >>> to support the devnode discovery using both udev and sysfs for their nodes. >> >> I actually thought about how alsa should be handled and it is doing the >> right thing. See my patch that I posted today, partially reverting your >> patch. > > Well, I can live with that patch for now, but I suspect that removing > major/minor from ALSA will make things very complex at the userspace side. > > Do you know how to convert from card/device/subdevice into a device node > patch using libudev and sysfs? You don't *want* a device. The user space alsa utilities all operate on card/device/subdevice level. I've never used an alsa device node directly, or seen anyone do that. That's all done by the alsa library. But regardless of whether or not it is right or wrong, as long as we do not actually implement this in a driver I would keep this and wait until we do have a working example. Replacing the alsa struct by a simple major/minor is certainly not enough, of that I am 100% certain. Let's keep it as is, and only touch it again when we actually get the first user. Regards, Hans -- 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
