As discussed when I met at the Plasma sprint - changing the viewport is semantically exactly what we want.

Yes it does break many things, but I have also (hopefullly!) fixed those many things.
Can you reconsider this and D3500 in light of that.

Then I think then this branch is all done \o/

Just a heads up: Atomic mode setting allows to scale buffers / content of a plane directly instead of using GL. But I assume this would be possible to add later.

Something else: How does this work, when Gl / compositing is skipped (i.e. in my case for the direct scanout of a wayland buffer)? Is it then not scaling at all? We need to make sure that in this case the dimensions are still fine. Otherwise we need to restrict direct scanout to AMS only.

In D3504#97513, @subdiff wrote:

Just a heads up: Atomic mode setting allows to scale buffers / content of a plane directly instead of using GL. But I assume this would be possible to add later.

Although it sort of looks like it, we're not scaling the buffer, in fact this line does the exact opposite setting our viewport to match the size used in the gbm_surface_create.
It's the projection that's scaled, not the buffer.

Something else: How does this work, when Gl / compositing is skipped (i.e. in my case for the direct scanout of a wayland buffer)? Is it then not scaling at all? We need to make sure that in this case the dimensions are still fine. Otherwise we need to restrict direct scanout to AMS only.

From the wayland docs:

A method of "scale" or "driver" implies a scaling operation of the surface, either via a direct scaling operation or a change of the output mode. This will override any kind of output scaling, so that mapping a surface with a buffer size equal to the mode can fill the screen independent of buffer_scale.
A method of "fill" means we don't scale up the buffer, however any output scale is applied.

So given we currently don't do fullscreen_method==fill, we don't need to do any changes for scaling and everything will just work \o/