On Tue, Jul 4, 2023 at 5:13 PM Frank Binns <Frank.Binns@xxxxxxxxxx> wrote: > > > Signed-off-by: Sarah Walker <sarah.walker@xxxxxxxxxx> > > > +properties: > > > + compatible: > > > + oneOf: > > > + - items: > > > + - enum: > > > + - ti,am62-gpu > > > + - const: img,powervr-seriesaxe > > > > I don't see why you need to restrict the bindings to just the stuff > > you happen to > > be writing Linux drivers for right now. > > > > Add all of them! > > The main thinking here was to start off with a single simple case (TI AM62) to > support the initial driver upstreaming rather than trying to cover too many > things at once. This can then be built upon for other GPUs. So, for example, we > can extend the bindings to add a second power domain for those that need it, > such as the GPU in the Mediatek MT8173. The benefit of this approach is that the > bindings, dts and code changes can all be reviewed together so that all the > context is present. I understand that you want to scale down the problem so it becomes easier. However the reverse can also be said: by removing the other platforms you may end up with one way street bindings that are really hard to augment to accomodate for the other family members later on. But as long as H. Nikolaus is involved in the review I suppose he will be able to point out obvious cases, and I expect that you read through the old thread: https://lore.kernel.org/all/cover.1587760454.git.hns@xxxxxxxxxxxxx/ Yours, Linus Walleij
