22 pointsby bgpdude5 hours ago2 comments
  • spaceywilly2 hours ago
    That’s very interesting, thank you for sharing. Looks like it could be a very useful tool for testing high performance networking.

    I wonder if something similar could be done using TC BPF instead of AF_XDP? My only reservation about AF XDP is that it requires a special NIC to support it, so it may not be useful for a “regular Joe” user. I wonder if TC BPF would also work since it similarly bypasses the Kernel networking stack, I believe you can put packets directly into the NIC TX queue for transmission

    • tptacek2 hours ago
      You can, but the interesting thing about AF_XDP is that you've got a userland path to writing directly to the card's DMA buffers; TC BPF still allocates an skbuff for every packet you send.
    • Palomides41 minutes ago
      it seems like every NIC on the market that can do 100Gb has support in its linux kernel driver, so probably not a big deal in practice
    • bgpdude2 hours ago
      you can use generic af_xdp which sits at the TC layer. Just get a bit less performance.
      • bgpdude2 hours ago
        That's what the veth examples are using