i40e: take into account XDP Tx queues when stopping rings
authorMaciej Fijalkowski <maciej.fijalkowski@intel.com>
Tue, 6 Feb 2024 12:41:32 +0000 (13:41 +0100)
committerTony Nguyen <anthony.l.nguyen@intel.com>
Tue, 13 Feb 2024 00:58:08 +0000 (16:58 -0800)
Seth reported that on his side XDP traffic can not survive a round of
down/up against i40e interface. Dmesg output was telling us that we were
not able to disable the very first XDP ring. That was due to the fact
that in i40e_vsi_stop_rings() in a pre-work that is done before calling
i40e_vsi_wait_queues_disabled(), XDP Tx queues were not taken into the
account.

To fix this, let us distinguish between Rx and Tx queue boundaries and
take into the account XDP queues for Tx side.

Reported-by: Seth Forshee <sforshee@kernel.org>
Closes: https://lore.kernel.org/netdev/ZbkE7Ep1N1Ou17sA@do-x1extreme/
Fixes: 65662a8dcdd0 ("i40e: Fix logic of disabling queues")
Tested-by: Seth Forshee <sforshee@kernel.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Reviewed-by: Ivan Vecera <ivecera@redhat.com>
Tested-by: Chandan Kumar Rout <chandanx.rout@intel.com> (A Contingent Worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
drivers/net/ethernet/intel/i40e/i40e_main.c

index 06078c4d54e82fd98f1d5b8ffc0928c8536695e3..54eb55464e3151f00665f9b780c65a1861f8ca07 100644 (file)
@@ -4926,21 +4926,23 @@ int i40e_vsi_start_rings(struct i40e_vsi *vsi)
 void i40e_vsi_stop_rings(struct i40e_vsi *vsi)
 {
        struct i40e_pf *pf = vsi->back;
-       int pf_q, q_end;
+       u32 pf_q, tx_q_end, rx_q_end;
 
        /* When port TX is suspended, don't wait */
        if (test_bit(__I40E_PORT_SUSPENDED, vsi->back->state))
                return i40e_vsi_stop_rings_no_wait(vsi);
 
-       q_end = vsi->base_queue + vsi->num_queue_pairs;
-       for (pf_q = vsi->base_queue; pf_q < q_end; pf_q++)
-               i40e_pre_tx_queue_cfg(&pf->hw, (u32)pf_q, false);
+       tx_q_end = vsi->base_queue +
+               vsi->alloc_queue_pairs * (i40e_enabled_xdp_vsi(vsi) ? 2 : 1);
+       for (pf_q = vsi->base_queue; pf_q < tx_q_end; pf_q++)
+               i40e_pre_tx_queue_cfg(&pf->hw, pf_q, false);
 
-       for (pf_q = vsi->base_queue; pf_q < q_end; pf_q++)
+       rx_q_end = vsi->base_queue + vsi->num_queue_pairs;
+       for (pf_q = vsi->base_queue; pf_q < rx_q_end; pf_q++)
                i40e_control_rx_q(pf, pf_q, false);
 
        msleep(I40E_DISABLE_TX_GAP_MSEC);
-       for (pf_q = vsi->base_queue; pf_q < q_end; pf_q++)
+       for (pf_q = vsi->base_queue; pf_q < tx_q_end; pf_q++)
                wr32(&pf->hw, I40E_QTX_ENA(pf_q), 0);
 
        i40e_vsi_wait_queues_disabled(vsi);