On Sun, 22 Sept 2024 at 22:51, Zhi Wang <zhiw@xxxxxxxxxx> wrote: > > All the resources that required by a new vGPU has been set up. It is time > to activate it. > > Send the NV2080_CTRL_CMD_VGPU_MGR_INTERNAL_BOOTLOAD_GSP_VGPU_PLUGIN_TASK > GSP RPC to activate the new vGPU. This patch is probably the best example of how this can't work. The GSP firmware interfaces are not guaranteed stable. Exposing these interfaces outside the nvkm core is unacceptable, as otherwise we would have to adapt the whole kernel depending on the loaded firmware. You cannot use any nvidia sdk headers, these all have to be abstracted behind things that have no bearing on the API. If a new NVIDIA driver release was to add a parameter inside NV2080_CTRL_VGPU_MGR_INTERNAL_BOOTLOAD_GSP_VGPU_PLUGIN_TASK_PARAMS how would you handle it? Outside of the other discussion, this is the fundamental problem with working on the GSP firmware. We cannot trust that any API exposed won't change, and NVIDIA aren't in a position to guarantee it. Dave.
