Hi,
I had a look into this and how other drivers are doing it. msm[1] and etnaviv[2] seem to use the same dual-use variable.
Ok, lets do it the same way then. What is the status of the userspace side of this?
The patch is hosted here[1], but is as of yet unmerged.
I know the drm subsystem requires a userspace implementation of all features, but does it have to be upstreamed first?
Which part is merged first doesn't matter much I think, but both kernel and userspace side patches should be reviewed, especially the API bits, so we don't end up with a bad API.
cheers, Gerd