[RFC,bpf-next,0/6] Add bpf_xdp_get_xfrm_state() kfunc

Message

Daniel Xu Oct. 27, 2023, 6:46 p.m. UTC

This patchset adds a kfunc helper, bpf_xdp_get_xfrm_state(), that wraps
xfrm_state_lookup(). The intent is to support software RSS (via XDP) for
the ongoing/upcoming ipsec pcpu work [0]. Recent experiments performed
on (hopefully) reproducible AWS testbeds indicate that single tunnel
pcpu ipsec can reach line rate on 100G ENA nics.

More details about that will be presented at netdev next week [1].

Antony did the initial stable bpf helper - I later ported it to unstable
kfuncs. So for the series, please apply a Co-developed-by for Antony,
provided he acks and signs off on this.

[0]: https://datatracker.ietf.org/doc/html/draft-ietf-ipsecme-multi-sa-performance-02
[1]: https://netdevconf.info/0x17/sessions/workshop/security-workshop.html

Daniel Xu (6):
  bpf: xfrm: Add bpf_xdp_get_xfrm_state() kfunc
  bpf: selftests: test_tunnel: Use ping -6 over ping6
  bpf: selftests: test_tunnel: Mount bpffs if necessary
  bpf: selftests: test_tunnel: Use vmlinux.h declarations
  bpf: selftests: test_tunnel: Disable CO-RE relocations
  bpf: xfrm: Add selftest for bpf_xdp_get_xfrm_state()

 include/net/xfrm.h                            |   9 ++
 net/xfrm/Makefile                             |   1 +
 net/xfrm/xfrm_policy.c                        |   2 +
 net/xfrm/xfrm_state_bpf.c                     | 105 ++++++++++++++++++
 .../selftests/bpf/progs/bpf_tracing_net.h     |   1 +
 .../selftests/bpf/progs/test_tunnel_kern.c    |  95 +++++++++-------
 tools/testing/selftests/bpf/test_tunnel.sh    |  43 ++++---
 7 files changed, 202 insertions(+), 54 deletions(-)
 create mode 100644 net/xfrm/xfrm_state_bpf.c

Comments

Antony Antony Oct. 29, 2023, 2:13 a.m. UTC | #1

On Fri, Oct 27, 2023 at 12:46:16 -0600, Daniel Xu wrote:
> This patchset adds a kfunc helper, bpf_xdp_get_xfrm_state(), that wraps
> xfrm_state_lookup(). The intent is to support software RSS (via XDP) for
> the ongoing/upcoming ipsec pcpu work [0]. Recent experiments performed
> on (hopefully) reproducible AWS testbeds indicate that single tunnel
> pcpu ipsec can reach line rate on 100G ENA nics.
> 
> More details about that will be presented at netdev next week [1].
> 
> Antony did the initial stable bpf helper - I later ported it to unstable
> kfuncs. So for the series, please apply a Co-developed-by for Antony,
> provided he acks and signs off on this.

Thanks Daniel for working on this and bringing it upstreadm.

Co-developed-by: Antony Antony <[e-mail]>
Signed-off-by: Antony Antony <[e-mail]>

> 
> [0]: https://datatracker.ietf.org/doc/html/draft-ietf-ipsecme-multi-sa-performance-02
> [1]: https://netdevconf.info/0x17/sessions/workshop/security-workshop.html
> 
> Daniel Xu (6):
>   bpf: xfrm: Add bpf_xdp_get_xfrm_state() kfunc
>   bpf: selftests: test_tunnel: Use ping -6 over ping6
>   bpf: selftests: test_tunnel: Mount bpffs if necessary
>   bpf: selftests: test_tunnel: Use vmlinux.h declarations
>   bpf: selftests: test_tunnel: Disable CO-RE relocations
>   bpf: xfrm: Add selftest for bpf_xdp_get_xfrm_state()
> 
>  include/net/xfrm.h                            |   9 ++
>  net/xfrm/Makefile                             |   1 +
>  net/xfrm/xfrm_policy.c                        |   2 +
>  net/xfrm/xfrm_state_bpf.c                     | 105 ++++++++++++++++++
>  .../selftests/bpf/progs/bpf_tracing_net.h     |   1 +
>  .../selftests/bpf/progs/test_tunnel_kern.c    |  95 +++++++++-------
>  tools/testing/selftests/bpf/test_tunnel.sh    |  43 ++++---
>  7 files changed, 202 insertions(+), 54 deletions(-)
>  create mode 100644 net/xfrm/xfrm_state_bpf.c
> 
> -- 
> 2.42.0
>