netdev-dpdk: Check pending reset when adding device.

When a device reset interrupt event (RTE_ETH_EVENT_INTR_RESET)
is detected for a DPDK device added to OVS, a device reset is
performed.

If a device reset interrupt event is detected for a device before
it is added to OVS, device reset is not called.

If that device is later attempted to be added to OVS, it may fail
while being configured as it is still pending a reset. This was
observed with the DPDK iavf driver, by forcing a reset event from
the ice driver and then adding the iavf device to OVS.

$ echo 2 > /sys/class/net/ens3f0/device/sriov_numvfs
$ ./devbind.py -b vfio-pci 0000:d8:01.1
$ ip link set ens3f0 vf 1 mac 26:ab:e6:6f:79:4d
$ ovs-vsctl add-port br0 dpdk0 -- set Interface dpdk0 type=dpdk \
      options:dpdk-devargs=0000:d8:01.1

|dpdk|ERR|Port1 dev_configure = -1
|netdev_dpdk|WARN|Interface dpdk0 eth_dev setup error
    Operation not permitted
|netdev_dpdk|ERR|Interface dpdk0(rxq:1 txq:5 lsc interrupt mode:false)
    configure error: Operation not permitted
|dpif_netdev|ERR|Failed to set interface dpdk0 new configuration

Add a check if there was any previous device reset interrupt events
when a device is added to OVS. If there was, perform the reset
before continuing with the rest of the configuration.

netdev_dpdk_pending_reset[] already tracks device reset interrupt
events for all devices, so it can be used to check if there is a
reset needed during configuration of newly added devices. By extending
it's usage dev->reset_needed is no longer needed.

Signed-off-by: Kevin Traynor <[email protected]>

Author: kevintraynor

lib/netdev-dpdk.c

@@ -464,7 +464,7 @@ struct netdev_dpdk {
         bool attached;
         /* If true, rte_eth_dev_start() was successfully called */
         bool started;
-        bool reset_needed;
+        /* todo: check padding */
         /* 1 pad byte here. */
         struct eth_addr hwaddr;
         int mtu;
@@ -1531,7 +1531,6 @@ common_construct(struct netdev *netdev, dpdk_port_t port_no,
     dev->virtio_features_state = OVS_VIRTIO_F_CLEAN;
     dev->attached = false;
     dev->started = false;
-    dev->reset_needed = false;
 
     ovsrcu_init(&dev->qos_conf, NULL);
 
@@ -2154,13 +2153,11 @@ netdev_dpdk_run(const struct netdev_class *netdev_class OVS_UNUSED)
             if (!pending_reset) {
                 continue;
             }
-            atomic_store_relaxed(&netdev_dpdk_pending_reset[port_id], false);
 
             ovs_mutex_lock(&dpdk_mutex);
             dev = netdev_dpdk_lookup_by_port_id(port_id);
             if (dev) {
                 ovs_mutex_lock(&dev->mutex);
-                dev->reset_needed = true;
                 netdev_request_reconfigure(&dev->up);
                 VLOG_DBG_RL(&rl, "%s: Device reset requested.",
                             netdev_get_name(&dev->up));
@@ -6074,6 +6071,7 @@ netdev_dpdk_reconfigure(struct netdev *netdev)
     struct netdev_dpdk *dev = netdev_dpdk_cast(netdev);
     bool try_rx_steer;
     int err = 0;
+    bool pending_reset = false;
 
     ovs_mutex_lock(&dev->mutex);
 
@@ -6083,6 +6081,9 @@ netdev_dpdk_reconfigure(struct netdev *netdev)
         dev->requested_n_rxq += 1;
     }
 
+    atomic_read_relaxed(&netdev_dpdk_pending_reset[dev->port_id],
+                        &pending_reset);
+
     if (netdev->n_txq == dev->requested_n_txq
         && netdev->n_rxq == dev->requested_n_rxq
         && dev->rx_steer_flags == dev->requested_rx_steer_flags
@@ -6092,7 +6093,7 @@ netdev_dpdk_reconfigure(struct netdev *netdev)
         && dev->txq_size == dev->requested_txq_size
         && eth_addr_equals(dev->hwaddr, dev->requested_hwaddr)
         && dev->socket_id == dev->requested_socket_id
-        && dev->started && !dev->reset_needed) {
+        && dev->started && !pending_reset) {
         /* Reconfiguration is unnecessary */
 
         goto out;
@@ -6101,10 +6102,10 @@ netdev_dpdk_reconfigure(struct netdev *netdev)
 retry:
     dpdk_rx_steer_unconfigure(dev);
 
-    if (dev->reset_needed) {
+    if (pending_reset) {
+        atomic_store_relaxed(&netdev_dpdk_pending_reset[dev->port_id], false);
         rte_eth_dev_reset(dev->port_id);
         if_notifier_manual_report();
-        dev->reset_needed = false;
     } else {
         rte_eth_dev_stop(dev->port_id);
     }