> We are in the prototyping phase of porting BPF to Solaris kernel, the > purpose is to improve the performance of those applications that use > libpcap. > > To provide multiple sets of API to use BPF functionality would not be > difficult IMO, but for PF_PACKET's case, as Darren pointed out, > PF_PACKET applications don't use BPF language, so I don't know if this > makes much sense.
So with libpcap today, it's not possible to use PF_PACKET and a kernel-based BPF? What environments can support a kernel-based BPF? > BTW: Volo will provide socket filter mechanism after the first phase > (already sent out for design review). Hopefully there will be a > follow-on putback. Understood. I was trying to understand if we have a hard requirement on that part of Volo in order to have a good architecture for BPF with PF_PACKET. But it sounds like that combination doesn't matter. -- meem _______________________________________________ networking-discuss mailing list [email protected]
