On Tue, Jul 22, 2014 at 7:51 AM, Mark Brown <broonie at kernel.org> wrote: > On Mon, Jul 21, 2014 at 11:16:32PM +0200, Thierry Reding wrote: >> On Mon, Jul 21, 2014 at 09:35:27AM -0600, Stephen Warren wrote: > >> > > vdd-2v8-display needs to remain always-on however. Here we may hit a >> > > limitation of the simple-panel driver, where only one power supply can >> > > be provided. > >> > Can't we fix the simple-panel driver to allow a list of regulators in >> > the property? > >> I have no objection to allowing that. But I don't think there's a way to >> do that with the current regulator API. You can use the bulk API, but >> that requires separate properties, not multiple regulators in one >> property. > >> Perhaps one idea to solve this would be to make the regulator API return >> a regulator handle that in fact controls an array of regulators? Adding >> Mark. > > I'm really not comfortable with that idea, it seems like most of the > users would be abusing it - one of the biggest issues is always getting > people to understand that their driver may be used in other systems with > the supplies mapped differently. If you were going to do something > along those lines you'd need to do something that enumerated all the > supply properties on the device.
I also don't think that would do it - for many displays the power-on sequence is not as simple as "switch all the regulators on". Maybe what we would want is to have panel-simple allow panels to provide a "probe" callback so they can request and manage any extra resource they need. This would of course imply that custom enable/disable and suspend/resume callbacks. Then the driver might not be "simple" anymore, but IMHO it would not be that bad.
