Re: [Patch bpf-next v3 4/5] skmsg: use skb ext instead of TCP_SKB_CB

[John Fastabend <john.fastabend@gmail.com>]

Lorenz Bauer wrote:
> On Mon, 15 Feb 2021 at 19:20, John Fastabend <john.fastabend@gmail.com> wrote:
> >
> > Cong Wang wrote:
> > > From: Cong Wang <cong.wang@bytedance.com>
> > >
> > > Currently TCP_SKB_CB() is hard-coded in skmsg code, it certainly
> > > does not work for any other non-TCP protocols. We can move them to
> > > skb ext instead of playing with skb cb, which is harder to make
> > > correct.
> > >
> > > Cc: John Fastabend <john.fastabend@gmail.com>
> > > Cc: Daniel Borkmann <daniel@iogearbox.net>
> > > Cc: Jakub Sitnicki <jakub@cloudflare.com>
> > > Reviewed-by: Lorenz Bauer <lmb@cloudflare.com>
> > > Signed-off-by: Cong Wang <cong.wang@bytedance.com>
> > > ---
> >
> > I'm not seeing the advantage of doing this at the moment. We can
> > continue to use cb[] here, which is simpler IMO and use the ext
> > if needed for the other use cases. This is adding a per packet
> > alloc cost that we don't have at the moment as I understand it.
> 
> John, do you have a benchmark we can look at? Right now we're arguing
> in the abstract.

Sure, but looks like Cong found some spare fields in sk_buff so
that looks much nicer.

I'll mess aound a bit with our benchmarks and see where we can
publish them. It would be good to have some repeatable tests
here folks can use.

Thanks,
John