Two issues introduced by my recent asynchronous filter handling changes:
1. The old filter_rfs_insert would replace a matching filter of equal
   priority; we need to pass the appropriate argument to filter_insert to
   make it do the same.
2. It's possible to cause the kernel to hammer ndo_rx_flow_steer very
   hard, so make sure we don't build up too huge a backlog of workitems.

Possibly it would be better to fix #2 on the kernel side; I think the way
 to do that would be to maintain a forward (as well as reverse) queue-to-
 cpu map and replace the set_rps_cpu() check
    if (rxq_index == skb_get_rx_queue(skb))
 with something like (pseudocode)
    if (irqaffinity of queue[skb_get_rx_queue(skb)] includes next_cpu)
 but I'm not sure whether it's right or even necessary, and in any case
 it's not a regression in 4.17 so isn't 'net' material.
(There's also the issue that we come up in the bad configuration by
 default, but that too is a problem for another time.)

Edward Cree (2):
  sfc: insert ARFS filters with replace_equal=true
  sfc: limit ARFS workitems in flight per channel

 drivers/net/ethernet/sfc/net_driver.h | 25 +++++++++++++++
 drivers/net/ethernet/sfc/rx.c         | 60 ++++++++++++++++++-----------------
 2 files changed, 56 insertions(+), 29 deletions(-)

Reply via email to