Have a look at this thread for information on the DPAA2 Network passthrough:
DPAA2 has a ‘container’ concept, so each container (DPRC) will have a network interface and some supporting resources (like buffers and data portals) in it.
The virtualization passthrough (like SR-IOV) works by creating a child DPRC and having the resources you want in the VM inside it, with some resources linked to the parent container.
Just to clarify, some NXP documentation refers to SR-IOV being supported on this CPU. It’s not the same, they are referring to when the CPU is being used as an “accelerator” card inside another server
From this presentation: https://www.nxp.com/docs/en/supporting-information/FTF-NET-N1881-PDF.pdf