Hi, 2014-10-27 0:30 GMT+01:00 Laurent Pinchart <laurent.pinchart@xxxxxxxxxxxxxxxx>: > Hi Jean-Michel, > > On Thursday 23 October 2014 07:51:50 Jean-Michel Hautbois wrote: >> 2014-10-23 1:53 GMT+02:00 Laurent Pinchart: >> > On Wednesday 22 October 2014 17:34:21 Jean-Michel Hautbois wrote: >> >> This patch adds support for DT parsing of ADV7604 as well as ADV7611. >> >> It needs to be improved in order to get ports parsing too. >> > >> > Let's improve it then :-) The DT bindings as proposed by this patch are >> > incomplete, that's just asking for trouble. >> > >> > How would you model the adv7604 ports ? >> >> I am opened to suggestions :). >> But it has to remain as simple as possible, ideally allowing for giving >> names to the ports. >> As done today, it works, ports are parsed but are all the same... > > The ADV7611 was easy, it had a single HDMI input only. The ADV7612 is easy as > well as it just has two separate HDMI inputs. > > The ADV7604 is a more complex beast. The HDMI inputs shouldn't be much of an > issue as they're independent and multiplexed internally. You can just create > one pad per HDMI input. > > The analog inputs, however, can't be modeled as easily. A naive approach would > be to create one pad for each of the 12 analog inputs, but the chip has three > separate ADCs and can combine 3 inputs in a single digital video stream. I > don't know how we should model support for that. Lars-Peter, Hans, would you > have a revolutionary idea to same the world today ? I get back to working on this specific part, but I don't know how these analog inputs should be modeled. On page 68 of ADV7604_HW_RevF there is Figure 11 showing typical configurations using AIN_SEL[2:0]. I can see 4 inputs muxed : this would suggest to have 4 pads for analog inputs. Not sure it makes sense though... Thanks, JM -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
