IOMMU/DT folks, any thoughts about this approach? The patches that are of interest outside of Host1x/TegraDRM specifics are patches 1, 2, 4, and 5.
FWIW it looks fairly innocuous to me. I don't understand host1x - neither hardware nor driver abstractions - well enough to meaningfully review it all (e.g. maybe it's deliberate that the bus .dma_configure method isn't used?), but the SMMU patch seems fine given the Kconfig solution to avoid module linkage problems.