Subject: xsk: i40e: Tx performance improvements
This patch set improves the performance of mainly the Tx processing of
AF_XDP sockets. Though, patch 4 also improves the Rx path. All in all,
this patch set improves the throughput of the l2fwd xdpsock
application by around 14%. If we just take a look at Tx processing part,
it is improved by nearly 50%.
Hopefully the new batched Tx interfaces should be of value to other
drivers implementing AF_XDP zero-copy support. But patch #4 is generic
and will improve performance of all drivers when using AF_XDP sockets
(under the premises explained in that patch).
@Daniel. In patch 4, I apply all the padding required to hinder the
adjacency prefetcher to prefetch the wrong things. After this patch
set, I will submit another patch set that introduces
____cacheline_padding_in_smp in include/linux/cache.h according to your
suggestions. The last patch in that patch set will then convert the
explicit paddings that we have now to ____cacheline_padding_in_smp.
This patch has been applied against commit d0b3d2d7e50d ("Merge branch 'selftests/bpf: Migrate test_tcpbpf_user to be a part of test_progs'")
Structure of the patch set:
Patch 1: Introduce lazy Tx completions in the i40e driver.
Patch 2: For the xdpsock sample, increment Tx stats at sending instead
of at completion.
Patch 3: Remove an unnecessary sw ring access from the Tx path in i40e.
Patch 4: Introduce padding between all pointers and fields in the ring.
Patch 5: Introduce batched Tx descriptor interfaces.
Patch 6: Use the new batched interfaces in the i40e driver to get higher
throughput.
Thanks: Magnus
Magnus Karlsson (6):
i40e: introduce lazy Tx completions for AF_XDP zero-copy
samples/bpf: increment Tx stats at sending
i40e: remove unnecessary sw_ring access from xsk Tx
xsk: introduce padding between more ring pointers
xsk: introduce batched Tx descriptor interfaces
i40e: use batched xsk Tx interfaces to increase performance
drivers/net/ethernet/intel/i40e/i40e_ethtool.c | 2 +-
drivers/net/ethernet/intel/i40e/i40e_main.c | 4 +-
drivers/net/ethernet/intel/i40e/i40e_txrx.c | 14 ++-
drivers/net/ethernet/intel/i40e/i40e_txrx.h | 3 +-
drivers/net/ethernet/intel/i40e/i40e_xsk.c | 140 +++++++++++++++++--------
include/net/xdp_sock_drv.h | 7 ++
net/xdp/xsk.c | 43 ++++++++
net/xdp/xsk_queue.h | 93 +++++++++++++---
samples/bpf/xdpsock_user.c | 6 +-
9 files changed, 249 insertions(+), 63 deletions(-)
--
2.7.4
