Hi Houlong, On Mon, Aug 16, 2021 at 1:00 PM houlong wei <houlong.wei@xxxxxxxxxxxx> wrote: > Hi Eizan, > > "mdp-rdma0" may be not the only one master device node. In fact, there > are 2 "mdp-rdma" in mt8173. You can see "mdp_rdma1" via below link. > > https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm64/boot/dts/mediatek/mt8173.dtsi?h=v5.13.11#n1016 > If we add "mediatek,mt8173-mdp" to "mdp_rdma1" like below, we will have > one more V4L2 video devie node. > > mdp_rdma1: rdma@14002000 { > compatible = "mediatek,mt8173-mdp-rdma", > "mediatek,mt8173-mdp"; > ... > } > > We should consider the case that there are more than one "MDP_RDMA" > components. Would it be okay with you if we added support for multiple MDP master device nodes in follow-up changes? My rationale is this: - As far as I can tell, the mediatek integration with V4L2 currently only handles a single MDP master device node. It's not clear to me the scope of changes that will be needed to make things work properly with multiple nodes. - The patch series makes video decode work (admittedly, in light of your comments not optimally) upstream, which is better than not landing these changes at all. I'd like to say that I'm very open to (and excited about) discussing further work to support multiple MDP master nodes, perhaps we can work together on this so I can understand what needs to be done. Please let me know your thoughts, Eizan
