[13:19] <larsc> hm, there are some drivers which use the media_entity API that depend on CONFIG_MEDIA_CONTROLLER, some that don't depend on CONFIG_MEDIA_CONTROLLER but ifdef out the calls if it is disabled [13:19] <larsc> and some that don't depend on the config symbol but don't ifdef anything out [13:19] <larsc> what is the right approach? [13:53] <hverkuil> shuah: ping [13:53] <shuah> hverkuil: Hi Hans [13:53] <hverkuil> Happy New Year! [13:53] <shuah> Happy New Year to you too [13:54] <hverkuil> Quick question regarding au0828: is the v2 patch series the latest? [13:54] <hverkuil> (back from vacation, so I'm going through the posted patches :-) ) [13:55] <shuah> hverkuil: yes that is the latest [13:55] <hverkuil> OK, so that means I should review it, right? [13:55] <shuah> I made the changes you recommended on patch v1 and fixed a couple of problems I found [13:55] <pinchartl> larsc: it depends on whether your driver requires MC or if it can just expose it as an option [13:55] <shuah> please do :) [13:56] <hverkuil> darn :-) [13:56] <shuah> Welcome back :) [13:57] <pinchartl> the omap3isp driver can't be used without the MC API, while the uvcvideo driver uses it more as a proof of concept [13:58] <shuah> Patch v2 series has 3 patches - the last one is the timeout removal we discussed - that one did make the driver stable [13:59] <hverkuil> Never trusted all that timeout crap. [14:03] <larsc> pinchartl: thanks. It looks like this becomes mandatory though if the driver wants to use pad_ops [14:04] <pinchartl> larsc: yes, although we should decouple the pad ops from their exposure to userspace [14:40] <shuah> hverkuil: thanks for the review - looking at your comments now [21:52] *** diegovio1 changes his alias to diegoviola