Seth reported in [0] that he couldn't get traffic flowing again after a round of down/up of interface that i40e driver manages.
While looking into fixing Tx disable timeout issue I also noticed that there is a doubled function call on Rx side which is fixed in patch 1. Thanks, Maciej [0]: https://lore.kernel.org/netdev/ZbkE7Ep1N1Ou17sA@do-x1extreme/ Maciej Fijalkowski (2): i40e: avoid double calling i40e_pf_rxq_wait() i40e: take into account XDP Tx queues when stopping rings drivers/net/ethernet/intel/i40e/i40e_main.c | 21 ++++++++------------- 1 file changed, 8 insertions(+), 13 deletions(-) -- 2.34.1
