On Fri, Dec 17, 2021 at 12:38 AM Maciej Fijalkowski
<maciej.fijalkowski@xxxxxxxxx> wrote:
>
> Commit 9610bd988df9 ("ice: optimize XDP_TX workloads") introduced
> @next_dd and @next_rs to ice_tx_ring struct. Currently, their state is
> not restored in ice_clean_tx_ring(), which was not causing any troubles
> as the XDP rings are gone after we're done with XDP prog on interface.
>
> For upcoming usage of mentioned fields in AF_XDP, this might expose us
> to a potential dead Tx side. Scenario would look like following (based
> on xdpsock):
>
> - two xdpsock instances are spawned in Tx mode
> - one of them is killed
> - XDP prog is kept on interface due to the other xdpsock still running
> * this means that XDP rings stayed in place
> - xdpsock is launched again on same queue id that was terminated on
> - @next_dd and @next_rs setting is bogus, therefore transmit side is
> broken
>
> To protect us from the above, restore the default @next_rs and @next_dd
> values when cleaning the Tx ring.
Thank you Maciej.
Acked-by: Magnus Karlsson <magnus.karlsson@xxxxxxxxx>
> Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@xxxxxxxxx>
> ---
> drivers/net/ethernet/intel/ice/ice_txrx.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c
> index bc3ba19dc88f..0f3f92ce8a95 100644
> --- a/drivers/net/ethernet/intel/ice/ice_txrx.c
> +++ b/drivers/net/ethernet/intel/ice/ice_txrx.c
> @@ -172,6 +172,8 @@ void ice_clean_tx_ring(struct ice_tx_ring *tx_ring)
>
> tx_ring->next_to_use = 0;
> tx_ring->next_to_clean = 0;
> + tx_ring->next_dd = ICE_TX_THRESH - 1;
> + tx_ring->next_rs = ICE_TX_THRESH - 1;
>
> if (!tx_ring->netdev)
> return;
> --
> 2.33.1
>
