Hi,

On Tue, Jul 27, 2021 at 11:20:54AM +0200, Daniel Vetter wrote:
> On Mon, Jul 26, 2021 at 05:16:57PM +0200, Maxime Ripard wrote:
> > Hi Daniel,
> > 
> > On Wed, Jul 21, 2021 at 02:05:01PM +0200, Daniel Vetter wrote:
> > > On Tue, Jul 20, 2021 at 03:45:19PM +0200, Maxime Ripard wrote:
> > > > Interactions between bridges, panels, MIPI-DSI host and the component
> > > > framework are not trivial and can lead to probing issues when
> > > > implementing a display driver. Let's document the various cases we need
> > > > too consider, and the solution to support all the cases.
> > > > 
> > > > Signed-off-by: Maxime Ripard <maxime@cerno.tech>
> > > 
> > > I still have this dream that eventually we resurrect a patch to add
> > > device_link to bridges/panels (ideally automatically), to help with some
> > > of the suspend/resume issues around here.
> > > 
> > > Will this make things worse?
> > > 
> > > I think it'd be really good to figure that out with some coding, since if
> > > we have incompatible solution to handle probe issues vs suspend/resume
> > > issues, we're screwed.
> > > 
> > > Atm the duct-tape is to carefully move things around between suspend and
> > > suspend_early hooks (and resume and resume_late) and hope it all works ...
> > 
> > My initial idea to fix this was indeed to use device links. I gave up
> > after a while since it doesn't look like there's a way to add a device
> > link before either the bridge or encoder probes.
> > 
> > Indeed the OF-Graph representation is device-specific, so it can't be
> > generic, and if you need to probe to add that link, well, it's already
> > too late for the probe ordering :)
> 
> But don't we still need the device_link for suspend/resume and module
> reload? All very annoying indeed anyway.

I guess we would still need it for proper suspend and resume ordering
(but I never really worked on that part, so I'm not sure), but it's a
bit orthogonal to the issue here since those can be added after probe

Maxime