I tried to call mesh.SetIndexBufferParams to declare a mesh’s index buffer’s length, but the manual says that the index buffer will be uninitialized and the SetIndexBufferData should be used.
But I don’t want this step’s communication between C/GPU and prefer to send a more simplified job config array to the compute shader and let it gen the mesh’s graphic buffers.
So I wonder is there a CPU method to only allocate a piece of memory in GPU for the mesh or I can change the mesh’s index buffer id to another graphic buffer’s id defined by myself?