[RFC PATCH bpf-next 0/2] bpf: Implement shared persistent fast(er) sk_storoage mode

[Hans Montero <hansmontero99@gmail.com>]

From: Hans Montero <hjm2133@columbia.edu>

This patch set adds a BPF local storage optimization. The first patch adds the
feature, and the second patch extends the bpf selftests so that the feature is
tested.

We are running BPF programs for each egress packet and noticed that
bpf_sk_storage_get incurs a significant amount of cpu time. By inlining the
storage into struct sock and accessing that instead of performing a map lookup,
we expect to reduce overhead for our specific use-case. This also has a
side-effect of persisting the socket storage, which can be beneficial.

This optimization is disabled by default and can be enabled by setting
CONFIG_BPF_SHARED_LOCAL_STORAGE_SIZE, the byte length of the inline buffer, to
a non-zero number.

Hans Montero (2):
  bpf: Implement shared sk_storage optimization
  selftests/bpf: Extend tests for shared sk_storage

 include/net/sock.h                            |  3 ++
 include/uapi/linux/bpf.h                      |  6 +++
 kernel/bpf/Kconfig                            | 11 +++++
 kernel/bpf/bpf_local_storage.c                |  3 +-
 net/core/bpf_sk_storage.c                     | 47 ++++++++++++++++++-
 tools/testing/selftests/bpf/config            |  1 +
 .../selftests/bpf/prog_tests/bpf_iter.c       | 31 +++++++++++-
 .../bpf/prog_tests/test_local_storage.c       |  3 ++
 .../progs/bpf_iter_bpf_sk_storage_helpers.c   | 27 ++++++++++-
 .../selftests/bpf/progs/local_storage.c       | 30 ++++++++++++
 10 files changed, 156 insertions(+), 6 deletions(-)

-- 
2.30.2