RPS: Ensure that an expired hardware filter can be re-added later

Amir Vadai wrote:
> When a stream is paused, and its rule is expired while it is paused,
> no new rule will be configured to the HW when traffic resume.
[...]
> - When stream was resumed, traffic was steered again by RSS, and
> because current-cpu was equal to desired-cpu,  ndo_rx_flow_steer
> wasn't called and no rule was configured to the HW.

Fix this by setting the flow's current CPU only in the table for the
newly selected RX queue.

Reported-and-tested-by: Amir Vadai <amirv@dev.mellanox.co.il>
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/core/dev.c

@@ -2670,10 +2670,7 @@ static struct rps_dev_flow *
 set_rps_cpu(struct net_device *dev, struct sk_buff *skb,
 	    struct rps_dev_flow *rflow, u16 next_cpu)
 {
-	u16 tcpu;
-
-	tcpu = rflow->cpu = next_cpu;
-	if (tcpu != RPS_NO_CPU) {
+	if (next_cpu != RPS_NO_CPU) {
 #ifdef CONFIG_RFS_ACCEL
 		struct netdev_rx_queue *rxqueue;
 		struct rps_dev_flow_table *flow_table;
@@ -2701,16 +2698,16 @@ set_rps_cpu(struct net_device *dev, struct sk_buff *skb,
 			goto out;
 		old_rflow = rflow;
 		rflow = &flow_table->flows[flow_id];
-		rflow->cpu = next_cpu;
 		rflow->filter = rc;
 		if (old_rflow->filter == rflow->filter)
 			old_rflow->filter = RPS_NO_FILTER;
 	out:
 #endif
 		rflow->last_qtail =
-			per_cpu(softnet_data, tcpu).input_queue_head;
+			per_cpu(softnet_data, next_cpu).input_queue_head;
 	}
 
+	rflow->cpu = next_cpu;
 	return rflow;
 }