net: Refactor XPS for CPUs and Rx queues

Refactor XPS code to support Tx queue selection based on
CPU(s) map or Rx queue(s) map.

Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/linux/cpumask.h

@@ -115,12 +115,17 @@ extern struct cpumask __cpu_active_mask;
 #define cpu_active(cpu)		((cpu) == 0)
 #endif
 
+static inline void cpu_max_bits_warn(unsigned int cpu, unsigned int bits)
+{
+#ifdef CONFIG_DEBUG_PER_CPU_MAPS
+	WARN_ON_ONCE(cpu >= bits);
+#endif /* CONFIG_DEBUG_PER_CPU_MAPS */
+}
+
 /* verify cpu argument to cpumask_* operators */
 static inline unsigned int cpumask_check(unsigned int cpu)
 {
-#ifdef CONFIG_DEBUG_PER_CPU_MAPS
+	cpu_max_bits_warn(cpu, nr_cpumask_bits);
-	WARN_ON_ONCE(cpu >= nr_cpumask_bits);
-#endif /* CONFIG_DEBUG_PER_CPU_MAPS */
 	return cpu;
 }
 

include/linux/netdevice.h

@@ -731,10 +731,15 @@ struct xps_map {
  */
 struct xps_dev_maps {
 	struct rcu_head rcu;
-	struct xps_map __rcu *cpu_map[0];
+	struct xps_map __rcu *attr_map[0]; /* Either CPUs map or RXQs map */
 };
-#define XPS_DEV_MAPS_SIZE(_tcs) (sizeof(struct xps_dev_maps) +		\
+
+#define XPS_CPU_DEV_MAPS_SIZE(_tcs) (sizeof(struct xps_dev_maps) +	\
 	(nr_cpu_ids * (_tcs) * sizeof(struct xps_map *)))
+
+#define XPS_RXQ_DEV_MAPS_SIZE(_tcs, _rxqs) (sizeof(struct xps_dev_maps) +\
+	(_rxqs * (_tcs) * sizeof(struct xps_map *)))
+
 #endif /* CONFIG_XPS */
 
 #define TC_MAX_QUEUE	16
@@ -1910,7 +1915,8 @@ struct net_device {
 	int			watchdog_timeo;
 
 #ifdef CONFIG_XPS
-	struct xps_dev_maps __rcu *xps_maps;
+	struct xps_dev_maps __rcu *xps_cpus_map;
+	struct xps_dev_maps __rcu *xps_rxqs_map;
 #endif
 #ifdef CONFIG_NET_CLS_ACT
 	struct mini_Qdisc __rcu	*miniq_egress;
@@ -3259,6 +3265,92 @@ static inline void netif_wake_subqueue(struct net_device *dev, u16 queue_index)
 #ifdef CONFIG_XPS
 int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
 			u16 index);
+int __netif_set_xps_queue(struct net_device *dev, const unsigned long *mask,
+			  u16 index, bool is_rxqs_map);
+
+/**
+ *	netif_attr_test_mask - Test a CPU or Rx queue set in a mask
+ *	@j: CPU/Rx queue index
+ *	@mask: bitmask of all cpus/rx queues
+ *	@nr_bits: number of bits in the bitmask
+ *
+ * Test if a CPU or Rx queue index is set in a mask of all CPU/Rx queues.
+ */
+static inline bool netif_attr_test_mask(unsigned long j,
+					const unsigned long *mask,
+					unsigned int nr_bits)
+{
+	cpu_max_bits_warn(j, nr_bits);
+	return test_bit(j, mask);
+}
+
+/**
+ *	netif_attr_test_online - Test for online CPU/Rx queue
+ *	@j: CPU/Rx queue index
+ *	@online_mask: bitmask for CPUs/Rx queues that are online
+ *	@nr_bits: number of bits in the bitmask
+ *
+ * Returns true if a CPU/Rx queue is online.
+ */
+static inline bool netif_attr_test_online(unsigned long j,
+					  const unsigned long *online_mask,
+					  unsigned int nr_bits)
+{
+	cpu_max_bits_warn(j, nr_bits);
+
+	if (online_mask)
+		return test_bit(j, online_mask);
+
+	return (j < nr_bits);
+}
+
+/**
+ *	netif_attrmask_next - get the next CPU/Rx queue in a cpu/Rx queues mask
+ *	@n: CPU/Rx queue index
+ *	@srcp: the cpumask/Rx queue mask pointer
+ *	@nr_bits: number of bits in the bitmask
+ *
+ * Returns >= nr_bits if no further CPUs/Rx queues set.
+ */
+static inline unsigned int netif_attrmask_next(int n, const unsigned long *srcp,
+					       unsigned int nr_bits)
+{
+	/* -1 is a legal arg here. */
+	if (n != -1)
+		cpu_max_bits_warn(n, nr_bits);
+
+	if (srcp)
+		return find_next_bit(srcp, nr_bits, n + 1);
+
+	return n + 1;
+}
+
+/**
+ *	netif_attrmask_next_and - get the next CPU/Rx queue in *src1p & *src2p
+ *	@n: CPU/Rx queue index
+ *	@src1p: the first CPUs/Rx queues mask pointer
+ *	@src2p: the second CPUs/Rx queues mask pointer
+ *	@nr_bits: number of bits in the bitmask
+ *
+ * Returns >= nr_bits if no further CPUs/Rx queues set in both.
+ */
+static inline int netif_attrmask_next_and(int n, const unsigned long *src1p,
+					  const unsigned long *src2p,
+					  unsigned int nr_bits)
+{
+	/* -1 is a legal arg here. */
+	if (n != -1)
+		cpu_max_bits_warn(n, nr_bits);
+
+	if (src1p && src2p)
+		return find_next_and_bit(src1p, src2p, nr_bits, n + 1);
+	else if (src1p)
+		return find_next_bit(src1p, nr_bits, n + 1);
+	else if (src2p)
+		return find_next_bit(src2p, nr_bits, n + 1);
+
+	return n + 1;
+}
 #else
 static inline int netif_set_xps_queue(struct net_device *dev,
 				      const struct cpumask *mask,

net/core/dev.c

@@ -2092,7 +2092,7 @@ static bool remove_xps_queue(struct xps_dev_maps *dev_maps,
 	int pos;
 
 	if (dev_maps)
-		map = xmap_dereference(dev_maps->cpu_map[tci]);
+		map = xmap_dereference(dev_maps->attr_map[tci]);
 	if (!map)
 		return false;
 
@@ -2105,7 +2105,7 @@ static bool remove_xps_queue(struct xps_dev_maps *dev_maps,
 			break;
 		}
 
-		RCU_INIT_POINTER(dev_maps->cpu_map[tci], NULL);
+		RCU_INIT_POINTER(dev_maps->attr_map[tci], NULL);
 		kfree_rcu(map, rcu);
 		return false;
 	}
@@ -2135,31 +2135,58 @@ static bool remove_xps_queue_cpu(struct net_device *dev,
 	return active;
 }
 
+static void clean_xps_maps(struct net_device *dev, const unsigned long *mask,
+			   struct xps_dev_maps *dev_maps, unsigned int nr_ids,
+			   u16 offset, u16 count, bool is_rxqs_map)
+{
+	bool active = false;
+	int i, j;
+
+	for (j = -1; j = netif_attrmask_next(j, mask, nr_ids),
+	     j < nr_ids;)
+		active |= remove_xps_queue_cpu(dev, dev_maps, j, offset,
+					       count);
+	if (!active) {
+		if (is_rxqs_map) {
+			RCU_INIT_POINTER(dev->xps_rxqs_map, NULL);
+		} else {
+			RCU_INIT_POINTER(dev->xps_cpus_map, NULL);
+
+			for (i = offset + (count - 1); count--; i--)
+				netdev_queue_numa_node_write(
+					netdev_get_tx_queue(dev, i),
+							NUMA_NO_NODE);
+		}
+		kfree_rcu(dev_maps, rcu);
+	}
+}
+
 static void netif_reset_xps_queues(struct net_device *dev, u16 offset,
 				   u16 count)
 {
+	const unsigned long *possible_mask = NULL;
 	struct xps_dev_maps *dev_maps;
-	int cpu, i;
+	unsigned int nr_ids;
-	bool active = false;
 
 	mutex_lock(&xps_map_mutex);
-	dev_maps = xmap_dereference(dev->xps_maps);
 
+	dev_maps = xmap_dereference(dev->xps_rxqs_map);
+	if (dev_maps) {
+		nr_ids = dev->num_rx_queues;
+		clean_xps_maps(dev, possible_mask, dev_maps, nr_ids, offset,
+			       count, true);
+
+	}
+
+	dev_maps = xmap_dereference(dev->xps_cpus_map);
 	if (!dev_maps)
 		goto out_no_maps;
 
-	for_each_possible_cpu(cpu)
+	if (num_possible_cpus() > 1)
-		active |= remove_xps_queue_cpu(dev, dev_maps, cpu,
+		possible_mask = cpumask_bits(cpu_possible_mask);
-					       offset, count);
+	nr_ids = nr_cpu_ids;
-
+	clean_xps_maps(dev, possible_mask, dev_maps, nr_ids, offset, count,
-	if (!active) {
+		       false);
-		RCU_INIT_POINTER(dev->xps_maps, NULL);
-		kfree_rcu(dev_maps, rcu);
-	}
-
-	for (i = offset + (count - 1); count--; i--)
-		netdev_queue_numa_node_write(netdev_get_tx_queue(dev, i),
-					     NUMA_NO_NODE);
 
 out_no_maps:
 	mutex_unlock(&xps_map_mutex);
@@ -2170,8 +2197,8 @@ static void netif_reset_xps_queues_gt(struct net_device *dev, u16 index)
 	netif_reset_xps_queues(dev, index, dev->num_tx_queues - index);
 }
 
-static struct xps_map *expand_xps_map(struct xps_map *map,
+static struct xps_map *expand_xps_map(struct xps_map *map, int attr_index,
-				      int cpu, u16 index)
+				      u16 index, bool is_rxqs_map)
 {
 	struct xps_map *new_map;
 	int alloc_len = XPS_MIN_MAP_ALLOC;
@@ -2183,7 +2210,7 @@ static struct xps_map *expand_xps_map(struct xps_map *map,
 		return map;
 	}
 
-	/* Need to add queue to this CPU's existing map */
+	/* Need to add tx-queue to this CPU's/rx-queue's existing map */
 	if (map) {
 		if (pos < map->alloc_len)
 			return map;
@@ -2191,9 +2218,14 @@ static struct xps_map *expand_xps_map(struct xps_map *map,
 		alloc_len = map->alloc_len * 2;
 	}
 
-	/* Need to allocate new map to store queue on this CPU's map */
+	/* Need to allocate new map to store tx-queue on this CPU's/rx-queue's
-	new_map = kzalloc_node(XPS_MAP_SIZE(alloc_len), GFP_KERNEL,
+	 *  map
-			       cpu_to_node(cpu));
+	 */
+	if (is_rxqs_map)
+		new_map = kzalloc(XPS_MAP_SIZE(alloc_len), GFP_KERNEL);
+	else
+		new_map = kzalloc_node(XPS_MAP_SIZE(alloc_len), GFP_KERNEL,
+				       cpu_to_node(attr_index));
 	if (!new_map)
 		return NULL;
 
@@ -2205,14 +2237,16 @@ static struct xps_map *expand_xps_map(struct xps_map *map,
 	return new_map;
 }
 
-int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
+int __netif_set_xps_queue(struct net_device *dev, const unsigned long *mask,
-			u16 index)
+			  u16 index, bool is_rxqs_map)
 {
+	const unsigned long *online_mask = NULL, *possible_mask = NULL;
 	struct xps_dev_maps *dev_maps, *new_dev_maps = NULL;
-	int i, cpu, tci, numa_node_id = -2;
+	int i, j, tci, numa_node_id = -2;
 	int maps_sz, num_tc = 1, tc = 0;
 	struct xps_map *map, *new_map;
 	bool active = false;
+	unsigned int nr_ids;
 
 	if (dev->num_tc) {
 		num_tc = dev->num_tc;
@@ -2221,16 +2255,27 @@ int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
 			return -EINVAL;
 	}
 
-	maps_sz = XPS_DEV_MAPS_SIZE(num_tc);
+	mutex_lock(&xps_map_mutex);
+	if (is_rxqs_map) {
+		maps_sz = XPS_RXQ_DEV_MAPS_SIZE(num_tc, dev->num_rx_queues);
+		dev_maps = xmap_dereference(dev->xps_rxqs_map);
+		nr_ids = dev->num_rx_queues;
+	} else {
+		maps_sz = XPS_CPU_DEV_MAPS_SIZE(num_tc);
+		if (num_possible_cpus() > 1) {
+			online_mask = cpumask_bits(cpu_online_mask);
+			possible_mask = cpumask_bits(cpu_possible_mask);
+		}
+		dev_maps = xmap_dereference(dev->xps_cpus_map);
+		nr_ids = nr_cpu_ids;
+	}
+
 	if (maps_sz < L1_CACHE_BYTES)
 		maps_sz = L1_CACHE_BYTES;
 
-	mutex_lock(&xps_map_mutex);
-
-	dev_maps = xmap_dereference(dev->xps_maps);
-
 	/* allocate memory for queue storage */
-	for_each_cpu_and(cpu, cpu_online_mask, mask) {
+	for (j = -1; j = netif_attrmask_next_and(j, online_mask, mask, nr_ids),
+	     j < nr_ids;) {
 		if (!new_dev_maps)
 			new_dev_maps = kzalloc(maps_sz, GFP_KERNEL);
 		if (!new_dev_maps) {
@@ -2238,73 +2283,81 @@ int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
 			return -ENOMEM;
 		}
 
-		tci = cpu * num_tc + tc;
+		tci = j * num_tc + tc;
-		map = dev_maps ? xmap_dereference(dev_maps->cpu_map[tci]) :
+		map = dev_maps ? xmap_dereference(dev_maps->attr_map[tci]) :
 				 NULL;
 
-		map = expand_xps_map(map, cpu, index);
+		map = expand_xps_map(map, j, index, is_rxqs_map);
 		if (!map)
 			goto error;
 
-		RCU_INIT_POINTER(new_dev_maps->cpu_map[tci], map);
+		RCU_INIT_POINTER(new_dev_maps->attr_map[tci], map);
 	}
 
 	if (!new_dev_maps)
 		goto out_no_new_maps;
 
-	for_each_possible_cpu(cpu) {
+	for (j = -1; j = netif_attrmask_next(j, possible_mask, nr_ids),
+	     j < nr_ids;) {
 		/* copy maps belonging to foreign traffic classes */
-		for (i = tc, tci = cpu * num_tc; dev_maps && i--; tci++) {
+		for (i = tc, tci = j * num_tc; dev_maps && i--; tci++) {
 			/* fill in the new device map from the old device map */
-			map = xmap_dereference(dev_maps->cpu_map[tci]);
+			map = xmap_dereference(dev_maps->attr_map[tci]);
-			RCU_INIT_POINTER(new_dev_maps->cpu_map[tci], map);
+			RCU_INIT_POINTER(new_dev_maps->attr_map[tci], map);
 		}
 
 		/* We need to explicitly update tci as prevous loop
 		 * could break out early if dev_maps is NULL.
 		 */
-		tci = cpu * num_tc + tc;
+		tci = j * num_tc + tc;
 
-		if (cpumask_test_cpu(cpu, mask) && cpu_online(cpu)) {
+		if (netif_attr_test_mask(j, mask, nr_ids) &&
-			/* add queue to CPU maps */
+		    netif_attr_test_online(j, online_mask, nr_ids)) {
+			/* add tx-queue to CPU/rx-queue maps */
 			int pos = 0;
 
-			map = xmap_dereference(new_dev_maps->cpu_map[tci]);
+			map = xmap_dereference(new_dev_maps->attr_map[tci]);
 			while ((pos < map->len) && (map->queues[pos] != index))
 				pos++;
 
 			if (pos == map->len)
 				map->queues[map->len++] = index;
 #ifdef CONFIG_NUMA
-			if (numa_node_id == -2)
+			if (!is_rxqs_map) {
-				numa_node_id = cpu_to_node(cpu);
+				if (numa_node_id == -2)
-			else if (numa_node_id != cpu_to_node(cpu))
+					numa_node_id = cpu_to_node(j);
-				numa_node_id = -1;
+				else if (numa_node_id != cpu_to_node(j))
+					numa_node_id = -1;
+			}
 #endif
 		} else if (dev_maps) {
 			/* fill in the new device map from the old device map */
-			map = xmap_dereference(dev_maps->cpu_map[tci]);
+			map = xmap_dereference(dev_maps->attr_map[tci]);
-			RCU_INIT_POINTER(new_dev_maps->cpu_map[tci], map);
+			RCU_INIT_POINTER(new_dev_maps->attr_map[tci], map);
 		}
 
 		/* copy maps belonging to foreign traffic classes */
 		for (i = num_tc - tc, tci++; dev_maps && --i; tci++) {
 			/* fill in the new device map from the old device map */
-			map = xmap_dereference(dev_maps->cpu_map[tci]);
+			map = xmap_dereference(dev_maps->attr_map[tci]);
-			RCU_INIT_POINTER(new_dev_maps->cpu_map[tci], map);
+			RCU_INIT_POINTER(new_dev_maps->attr_map[tci], map);
 		}
 	}
 
-	rcu_assign_pointer(dev->xps_maps, new_dev_maps);
+	if (is_rxqs_map)
+		rcu_assign_pointer(dev->xps_rxqs_map, new_dev_maps);
+	else
+		rcu_assign_pointer(dev->xps_cpus_map, new_dev_maps);
 
 	/* Cleanup old maps */
 	if (!dev_maps)
 		goto out_no_old_maps;
 
-	for_each_possible_cpu(cpu) {
+	for (j = -1; j = netif_attrmask_next(j, possible_mask, nr_ids),
-		for (i = num_tc, tci = cpu * num_tc; i--; tci++) {
+	     j < nr_ids;) {
-			new_map = xmap_dereference(new_dev_maps->cpu_map[tci]);
+		for (i = num_tc, tci = j * num_tc; i--; tci++) {
-			map = xmap_dereference(dev_maps->cpu_map[tci]);
+			new_map = xmap_dereference(new_dev_maps->attr_map[tci]);
+			map = xmap_dereference(dev_maps->attr_map[tci]);
 			if (map && map != new_map)
 				kfree_rcu(map, rcu);
 		}
@@ -2317,19 +2370,23 @@ out_no_old_maps:
 	active = true;
 
 out_no_new_maps:
-	/* update Tx queue numa node */
+	if (!is_rxqs_map) {
-	netdev_queue_numa_node_write(netdev_get_tx_queue(dev, index),
+		/* update Tx queue numa node */
-				     (numa_node_id >= 0) ? numa_node_id :
+		netdev_queue_numa_node_write(netdev_get_tx_queue(dev, index),
-				     NUMA_NO_NODE);
+					     (numa_node_id >= 0) ?
+					     numa_node_id : NUMA_NO_NODE);
+	}
 
 	if (!dev_maps)
 		goto out_no_maps;
 
-	/* removes queue from unused CPUs */
+	/* removes tx-queue from unused CPUs/rx-queues */
-	for_each_possible_cpu(cpu) {
+	for (j = -1; j = netif_attrmask_next(j, possible_mask, nr_ids),
-		for (i = tc, tci = cpu * num_tc; i--; tci++)
+	     j < nr_ids;) {
+		for (i = tc, tci = j * num_tc; i--; tci++)
 			active |= remove_xps_queue(dev_maps, tci, index);
-		if (!cpumask_test_cpu(cpu, mask) || !cpu_online(cpu))
+		if (!netif_attr_test_mask(j, mask, nr_ids) ||
+		    !netif_attr_test_online(j, online_mask, nr_ids))
 			active |= remove_xps_queue(dev_maps, tci, index);
 		for (i = num_tc - tc, tci++; --i; tci++)
 			active |= remove_xps_queue(dev_maps, tci, index);
@@ -2337,7 +2394,10 @@ out_no_new_maps:
 
 	/* free map if not active */
 	if (!active) {
-		RCU_INIT_POINTER(dev->xps_maps, NULL);
+		if (is_rxqs_map)
+			RCU_INIT_POINTER(dev->xps_rxqs_map, NULL);
+		else
+			RCU_INIT_POINTER(dev->xps_cpus_map, NULL);
 		kfree_rcu(dev_maps, rcu);
 	}
 
@@ -2347,11 +2407,12 @@ out_no_maps:
 	return 0;
 error:
 	/* remove any maps that we added */
-	for_each_possible_cpu(cpu) {
+	for (j = -1; j = netif_attrmask_next(j, possible_mask, nr_ids),
-		for (i = num_tc, tci = cpu * num_tc; i--; tci++) {
+	     j < nr_ids;) {
-			new_map = xmap_dereference(new_dev_maps->cpu_map[tci]);
+		for (i = num_tc, tci = j * num_tc; i--; tci++) {
+			new_map = xmap_dereference(new_dev_maps->attr_map[tci]);
 			map = dev_maps ?
-			      xmap_dereference(dev_maps->cpu_map[tci]) :
+			      xmap_dereference(dev_maps->attr_map[tci]) :
 			      NULL;
 			if (new_map && new_map != map)
 				kfree(new_map);
@@ -2363,6 +2424,12 @@ error:
 	kfree(new_dev_maps);
 	return -ENOMEM;
 }
+
+int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
+			u16 index)
+{
+	return __netif_set_xps_queue(dev, cpumask_bits(mask), index, false);
+}
 EXPORT_SYMBOL(netif_set_xps_queue);
 
 #endif
@@ -3384,7 +3451,7 @@ static inline int get_xps_queue(struct net_device *dev, struct sk_buff *skb)
 	int queue_index = -1;
 
 	rcu_read_lock();
-	dev_maps = rcu_dereference(dev->xps_maps);
+	dev_maps = rcu_dereference(dev->xps_cpus_map);
 	if (dev_maps) {
 		unsigned int tci = skb->sender_cpu - 1;
 
@@ -3393,7 +3460,7 @@ static inline int get_xps_queue(struct net_device *dev, struct sk_buff *skb)
 			tci += netdev_get_prio_tc_map(dev, skb->priority);
 		}
 
-		map = rcu_dereference(dev_maps->cpu_map[tci]);
+		map = rcu_dereference(dev_maps->attr_map[tci]);
 		if (map) {
 			if (map->len == 1)
 				queue_index = map->queues[0];

net/core/net-sysfs.c

@@ -1227,13 +1227,13 @@ static ssize_t xps_cpus_show(struct netdev_queue *queue,
 		return -ENOMEM;
 
 	rcu_read_lock();
-	dev_maps = rcu_dereference(dev->xps_maps);
+	dev_maps = rcu_dereference(dev->xps_cpus_map);
 	if (dev_maps) {
 		for_each_possible_cpu(cpu) {
 			int i, tci = cpu * num_tc + tc;
 			struct xps_map *map;
 
-			map = rcu_dereference(dev_maps->cpu_map[tci]);
+			map = rcu_dereference(dev_maps->attr_map[tci]);
 			if (!map)
 				continue;