On Wed, Dec 28, 2016 at 10:39:17AM +0200, Jani Nikula wrote:
On Tue, 27 Dec 2016, Daniel Vetter daniel@ffwll.ch wrote:
I'm not sure how this all should look like.
This we definitely agree on, and hence the RFC. :)
I'm pretty sure it should *not* look like it currently does.
Yup, agreed on that. Semi-random other idea: What about introducing a new ->probe helper callback on connectors which would essentially entai all the things you want to change? I.e. what's currently done with ->detect plus connector status forcing, and the ->get_modes vs. edid override stuff?
Step 1 would be to have a default implementation that does exactly the same thing. Plus reworking the hdp and poll helpers to prefer ->probe over just calling ->detect. Although for hpd I'm not sure whether using the same callback is a good idea really, at least in i915 we have a separate ->hpd callback for DP. And it needs it.
And then we can try to figure out how to re-split responsibilities between drivers and helpers (and maybe try to stuff parts of it into specialized helpers, e.g. for DP). -Daniel