xsk: Make xskmap flush_list common for all map instances

The xskmap flush list is used to track entries that need to flushed
from via the xdp_do_flush_map() function. This list used to be
per-map, but there is really no reason for that. Instead make the
flush list global for all xskmaps, which simplifies __xsk_map_flush()
and xsk_map_alloc().

Signed-off-by: Björn Töpel <bjorn.topel@intel.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://lore.kernel.org/bpf/20191219061006.21980-5-bjorn.topel@gmail.com

include/net/xdp_sock.h

@@ -72,7 +72,6 @@ struct xdp_umem {
 
 struct xsk_map {
 	struct bpf_map map;
-	struct list_head __percpu *flush_list;
 	spinlock_t lock; /* Synchronize map updates */
 	struct xdp_sock *xsk_map[];
 };
@@ -139,9 +138,8 @@ void xsk_map_try_sock_delete(struct xsk_map *map, struct xdp_sock *xs,
 			     struct xdp_sock **map_entry);
 int xsk_map_inc(struct xsk_map *map);
 void xsk_map_put(struct xsk_map *map);
-int __xsk_map_redirect(struct bpf_map *map, struct xdp_buff *xdp,
-		       struct xdp_sock *xs);
-void __xsk_map_flush(struct bpf_map *map);
+int __xsk_map_redirect(struct xdp_sock *xs, struct xdp_buff *xdp);
+void __xsk_map_flush(void);
 
 static inline struct xdp_sock *__xsk_map_lookup_elem(struct bpf_map *map,
 						     u32 key)
@@ -369,13 +367,12 @@ static inline u64 xsk_umem_adjust_offset(struct xdp_umem *umem, u64 handle,
 	return 0;
 }
 
-static inline int __xsk_map_redirect(struct bpf_map *map, struct xdp_buff *xdp,
-				     struct xdp_sock *xs)
+static inline int __xsk_map_redirect(struct xdp_sock *xs, struct xdp_buff *xdp)
 {
 	return -EOPNOTSUPP;
 }
 
-static inline void __xsk_map_flush(struct bpf_map *map)
+static inline void __xsk_map_flush(void)
 {
 }
 

kernel/bpf/xskmap.c

@@ -72,9 +72,9 @@ static void xsk_map_sock_delete(struct xdp_sock *xs,
 static struct bpf_map *xsk_map_alloc(union bpf_attr *attr)
 {
 	struct bpf_map_memory mem;
-	int cpu, err, numa_node;
+	int err, numa_node;
 	struct xsk_map *m;
-	u64 cost, size;
+	u64 size;
 
 	if (!capable(CAP_NET_ADMIN))
 		return ERR_PTR(-EPERM);
@@ -86,9 +86,8 @@ static struct bpf_map *xsk_map_alloc(union bpf_attr *attr)
 
 	numa_node = bpf_map_attr_numa_node(attr);
 	size = struct_size(m, xsk_map, attr->max_entries);
-	cost = size + array_size(sizeof(*m->flush_list), num_possible_cpus());
 
-	err = bpf_map_charge_init(&mem, cost);
+	err = bpf_map_charge_init(&mem, size);
 	if (err < 0)
 		return ERR_PTR(err);
 
@@ -102,16 +101,6 @@ static struct bpf_map *xsk_map_alloc(union bpf_attr *attr)
 	bpf_map_charge_move(&m->map.memory, &mem);
 	spin_lock_init(&m->lock);
 
-	m->flush_list = alloc_percpu(struct list_head);
-	if (!m->flush_list) {
-		bpf_map_charge_finish(&m->map.memory);
-		bpf_map_area_free(m);
-		return ERR_PTR(-ENOMEM);
-	}
-
-	for_each_possible_cpu(cpu)
-		INIT_LIST_HEAD(per_cpu_ptr(m->flush_list, cpu));
-
 	return &m->map;
 }
 
@@ -121,7 +110,6 @@ static void xsk_map_free(struct bpf_map *map)
 
 	bpf_clear_redirect_map(map);
 	synchronize_net();
-	free_percpu(m->flush_list);
 	bpf_map_area_free(m);
 }
 

net/core/filter.c

@@ -3511,8 +3511,7 @@ err:
 
 static int __bpf_tx_xdp_map(struct net_device *dev_rx, void *fwd,
 			    struct bpf_map *map,
-			    struct xdp_buff *xdp,
-			    u32 index)
+			    struct xdp_buff *xdp)
 {
 	int err;
 
@@ -3537,7 +3536,7 @@ static int __bpf_tx_xdp_map(struct net_device *dev_rx, void *fwd,
 	case BPF_MAP_TYPE_XSKMAP: {
 		struct xdp_sock *xs = fwd;
 
-		err = __xsk_map_redirect(map, xdp, xs);
+		err = __xsk_map_redirect(xs, xdp);
 		return err;
 	}
 	default:
@@ -3562,7 +3561,7 @@ void xdp_do_flush_map(void)
 			__cpu_map_flush(map);
 			break;
 		case BPF_MAP_TYPE_XSKMAP:
-			__xsk_map_flush(map);
+			__xsk_map_flush();
 			break;
 		default:
 			break;
@@ -3619,7 +3618,7 @@ static int xdp_do_redirect_map(struct net_device *dev, struct xdp_buff *xdp,
 	if (ri->map_to_flush && unlikely(ri->map_to_flush != map))
 		xdp_do_flush_map();
 
-	err = __bpf_tx_xdp_map(dev, fwd, map, xdp, index);
+	err = __bpf_tx_xdp_map(dev, fwd, map, xdp);
 	if (unlikely(err))
 		goto err;
 

net/xdp/xsk.c

@@ -31,6 +31,8 @@
 
 #define TX_BATCH_SIZE 16
 
+static DEFINE_PER_CPU(struct list_head, xskmap_flush_list);
+
 bool xsk_is_setup_for_bpf_map(struct xdp_sock *xs)
 {
 	return READ_ONCE(xs->rx) &&  READ_ONCE(xs->umem) &&
@@ -264,11 +266,9 @@ out_unlock:
 	return err;
 }
 
-int __xsk_map_redirect(struct bpf_map *map, struct xdp_buff *xdp,
-		       struct xdp_sock *xs)
+int __xsk_map_redirect(struct xdp_sock *xs, struct xdp_buff *xdp)
 {
-	struct xsk_map *m = container_of(map, struct xsk_map, map);
-	struct list_head *flush_list = this_cpu_ptr(m->flush_list);
+	struct list_head *flush_list = this_cpu_ptr(&xskmap_flush_list);
 	int err;
 
 	err = xsk_rcv(xs, xdp);
@@ -281,10 +281,9 @@ int __xsk_map_redirect(struct bpf_map *map, struct xdp_buff *xdp,
 	return 0;
 }
 
-void __xsk_map_flush(struct bpf_map *map)
+void __xsk_map_flush(void)
 {
-	struct xsk_map *m = container_of(map, struct xsk_map, map);
-	struct list_head *flush_list = this_cpu_ptr(m->flush_list);
+	struct list_head *flush_list = this_cpu_ptr(&xskmap_flush_list);
 	struct xdp_sock *xs, *tmp;
 
 	list_for_each_entry_safe(xs, tmp, flush_list, flush_node) {
@@ -1177,7 +1176,7 @@ static struct pernet_operations xsk_net_ops = {
 
 static int __init xsk_init(void)
 {
-	int err;
+	int err, cpu;
 
 	err = proto_register(&xsk_proto, 0 /* no slab */);
 	if (err)
@@ -1195,6 +1194,8 @@ static int __init xsk_init(void)
 	if (err)
 		goto out_pernet;
 
+	for_each_possible_cpu(cpu)
+		INIT_LIST_HEAD(&per_cpu(xskmap_flush_list, cpu));
 	return 0;
 
 out_pernet: