Hi Laurent.
> > > +static int td043mtea1_disable(struct drm_panel *panel)
> > > +{
> > > + struct td043mtea1_device *lcd = to_td043mtea1_device(panel);
> > > +
> > > + if (!lcd->spi_suspended)
> > > + td043mtea1_power_off(lcd);
> > > +
> > > + return 0;
> > > +}
> > > +
> > > +static int td043mtea1_enable(struct drm_panel *panel)
> > > +{
> > > + struct td043mtea1_device *lcd = to_td043mtea1_device(panel);
> > > + int ret;
> > > +
> > > + /*
> > > + * If we are resuming from system suspend, SPI might not be enabled
> > > + * yet, so we'll program the LCD from SPI PM resume callback.
> > > + */
> > > + if (lcd->spi_suspended)
> > > + return 0;
> >
> > I do not recall this is needed in other panel drivers, so look at what
> > other spi based panels do here.
> > I think this is something that today is not required.
>
> The problem here is that the display controller may be resumed before
> the SPI bus. Has that been solved somewhere in core code ?
I dunno. So the conclusion is to keep it as is, and any change
will wait until someone with HW can step up.
As for all your other feedback to this and the other panel drivers
they did not trigger any repsonse from me.
Looks forward for next iteration of this nice set of patches.
Can we maybe get the panel drivers in before some of the infrastructure
work?
I know the users then may come a bit later, but I think thats OK.
Sam
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
