cpumask_set_cpu_local_first => cpumask_local_spread, lament

da91309e0a7e (cpumask: Utility function to set n'th cpu...) created a
genuinely weird function.  I never saw it before, it went through DaveM.
(He only does this to make us other maintainers feel better about our own
mistakes.)

cpumask_set_cpu_local_first's purpose is say "I need to spread things
across N online cpus, choose the ones on this numa node first"; you call
it in a loop.

It can fail.  One of the two callers ignores this, the other aborts and
fails the device open.

It can fail in two ways: allocating the off-stack cpumask, or through a
convoluted codepath which AFAICT can only occur if cpu_online_mask
changes.  Which shouldn't happen, because if cpu_online_mask can change
while you call this, it could return a now-offline cpu anyway.

It contains a nonsensical test "!cpumask_of_node(numa_node)".  This was
drawn to my attention by Geert, who said this causes a warning on Sparc.
It sets a single bit in a cpumask instead of returning a cpu number,
because that's what the callers want.

It could be made more efficient by passing the previous cpu rather than
an index, but that would be more invasive to the callers.

Fixes: da91309e0a7e8966d916a74cce42ed170fde06bf
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> (then rebased)
Tested-by: Amir Vadai <amirv@mellanox.com>
Acked-by: Amir Vadai <amirv@mellanox.com>
Acked-by: David S. Miller <davem@davemloft.net>

5 files changed, 37 insertions, 65 deletions

diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index a6dcbf850c1f..6f9ffb9026cd 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -2358,11 +2358,11 @@ static int be_evt_queues_create(struct be_adapter *adapter)
 				    adapter->cfg_num_qs);
 
 	for_all_evt_queues(adapter, eqo, i) {
+		int numa_node = dev_to_node(&adapter->pdev->dev);
 		if (!zalloc_cpumask_var(&eqo->affinity_mask, GFP_KERNEL))
 			return -ENOMEM;
-		cpumask_set_cpu_local_first(i, dev_to_node(&adapter->pdev->dev),
-					    eqo->affinity_mask);
-
+		cpumask_set_cpu(cpumask_local_spread(i, numa_node),
+				eqo->affinity_mask);
 		netif_napi_add(adapter->netdev, &eqo->napi, be_poll,
 			       BE_NAPI_WEIGHT);
 		napi_hash_add(&eqo->napi);
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
index 32f5ec737472..cf467a9f6cc7 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
@@ -1501,17 +1501,13 @@ static int mlx4_en_init_affinity_hint(struct mlx4_en_priv *priv, int ring_idx)
 {
 	struct mlx4_en_rx_ring *ring = priv->rx_ring[ring_idx];
 	int numa_node = priv->mdev->dev->numa_node;
-	int ret = 0;
 
 	if (!zalloc_cpumask_var(&ring->affinity_mask, GFP_KERNEL))
 		return -ENOMEM;
 
-	ret = cpumask_set_cpu_local_first(ring_idx, numa_node,
-					  ring->affinity_mask);
-	if (ret)
-		free_cpumask_var(ring->affinity_mask);
-
-	return ret;
+	cpumask_set_cpu(cpumask_local_spread(ring_idx, numa_node),
+			ring->affinity_mask);
+	return 0;
 }
 
 static void mlx4_en_free_affinity_hint(struct mlx4_en_priv *priv, int ring_idx)
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_tx.c b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
index f7bf312fb443..7bed3a88579f 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
@@ -144,9 +144,9 @@ int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv,
 	ring->queue_index = queue_index;
 
 	if (queue_index < priv->num_tx_rings_p_up)
-		cpumask_set_cpu_local_first(queue_index,
-					    priv->mdev->dev->numa_node,
-					    &ring->affinity_mask);
+		cpumask_set_cpu(cpumask_local_spread(queue_index,
+						     priv->mdev->dev->numa_node),
+				&ring->affinity_mask);
 
 	*pring = ring;
 	return 0;
diff --git a/include/linux/cpumask.h b/include/linux/cpumask.h
index 27e285b92b5f..59915ea5373c 100644
--- a/include/linux/cpumask.h
+++ b/include/linux/cpumask.h
@@ -151,10 +151,8 @@ static inline unsigned int cpumask_any_but(const struct cpumask *mask,
 	return 1;
 }
 
-static inline int cpumask_set_cpu_local_first(int i, int numa_node, cpumask_t *dstp)
+static inline unsigned int cpumask_local_spread(unsigned int i, int node)
 {
-	set_bit(0, cpumask_bits(dstp));
-
 	return 0;
 }
 
@@ -208,7 +206,7 @@ static inline unsigned int cpumask_next_zero(int n, const struct cpumask *srcp)
 
 int cpumask_next_and(int n, const struct cpumask *, const struct cpumask *);
 int cpumask_any_but(const struct cpumask *mask, unsigned int cpu);
-int cpumask_set_cpu_local_first(int i, int numa_node, cpumask_t *dstp);
+unsigned int cpumask_local_spread(unsigned int i, int node);
 
 /**
  * for_each_cpu - iterate over every cpu in a mask
diff --git a/lib/cpumask.c b/lib/cpumask.c
index 830dd5dec40f..5f627084f2e9 100644
--- a/lib/cpumask.c
+++ b/lib/cpumask.c
@@ -139,64 +139,42 @@ void __init free_bootmem_cpumask_var(cpumask_var_t mask)
 #endif
 
 /**
- * cpumask_set_cpu_local_first - set i'th cpu with local numa cpu's first
- *
+ * cpumask_local_spread - select the i'th cpu with local numa cpu's first
  * @i: index number
- * @numa_node: local numa_node
- * @dstp: cpumask with the relevant cpu bit set according to the policy
+ * @node: local numa_node
  *
- * This function sets the cpumask according to a numa aware policy.
- * cpumask could be used as an affinity hint for the IRQ related to a
- * queue. When the policy is to spread queues across cores - local cores
- * first.
+ * This function selects an online CPU according to a numa aware policy;
+ * local cpus are returned first, followed by non-local ones, then it
+ * wraps around.
  *
- * Returns 0 on success, -ENOMEM for no memory, and -EAGAIN when failed to set
- * the cpu bit and need to re-call the function.
+ * It's not very efficient, but useful for setup.
  */
-int cpumask_set_cpu_local_first(int i, int numa_node, cpumask_t *dstp)
+unsigned int cpumask_local_spread(unsigned int i, int node)
 {
-	cpumask_var_t mask;
 	int cpu;
-	int ret = 0;
-
-	if (!zalloc_cpumask_var(&mask, GFP_KERNEL))
-		return -ENOMEM;
 
+	/* Wrap: we always want a cpu. */
 	i %= num_online_cpus();
 
-	if (numa_node == -1 || !cpumask_of_node(numa_node)) {
-		/* Use all online cpu's for non numa aware system */
-		cpumask_copy(mask, cpu_online_mask);
+	if (node == -1) {
+		for_each_cpu(cpu, cpu_online_mask)
+			if (i-- == 0)
+				return cpu;
 	} else {
-		int n;
-
-		cpumask_and(mask,
-			    cpumask_of_node(numa_node), cpu_online_mask);
-
-		n = cpumask_weight(mask);
-		if (i >= n) {
-			i -= n;
-
-			/* If index > number of local cpu's, mask out local
-			 * cpu's
-			 */
-			cpumask_andnot(mask, cpu_online_mask, mask);
+		/* NUMA first. */
+		for_each_cpu_and(cpu, cpumask_of_node(node), cpu_online_mask)
+			if (i-- == 0)
+				return cpu;
+
+		for_each_cpu(cpu, cpu_online_mask) {
+			/* Skip NUMA nodes, done above. */
+			if (cpumask_test_cpu(cpu, cpumask_of_node(node)))
+				continue;
+
+			if (i-- == 0)
+				return cpu;
 		}
 	}
-
-	for_each_cpu(cpu, mask) {
-		if (--i < 0)
-			goto out;
-	}
-
-	ret = -EAGAIN;
-
-out:
-	free_cpumask_var(mask);
-
-	if (!ret)
-		cpumask_set_cpu(cpu, dstp);
-
-	return ret;
+	BUG();
 }
-EXPORT_SYMBOL(cpumask_set_cpu_local_first);
+EXPORT_SYMBOL(cpumask_local_spread);