Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net

Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 57 insertions, 20 deletions

diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c
index 1469365bac7e..1d519b0e51a5 100644
--- a/net/netfilter/nf_conntrack_netlink.c
+++ b/net/netfilter/nf_conntrack_netlink.c
@@ -2962,6 +2962,7 @@ static int ctnetlink_exp_dump_mask(struct sk_buff *skb,
 	memset(&m, 0xFF, sizeof(m));
 	memcpy(&m.src.u3, &mask->src.u3, sizeof(m.src.u3));
 	m.src.u.all = mask->src.u.all;
+	m.src.l3num = tuple->src.l3num;
 	m.dst.protonum = tuple->dst.protonum;
 
 	nest_parms = nla_nest_start(skb, CTA_EXPECT_MASK);
diff --git a/net/netfilter/nf_conntrack_proto_gre.c b/net/netfilter/nf_conntrack_proto_gre.c
index 5b05487a60d2..db11e403d818 100644
--- a/net/netfilter/nf_conntrack_proto_gre.c
+++ b/net/netfilter/nf_conntrack_proto_gre.c
@@ -218,9 +218,6 @@ int nf_conntrack_gre_packet(struct nf_conn *ct,
 			    enum ip_conntrack_info ctinfo,
 			    const struct nf_hook_state *state)
 {
-	if (state->pf != NFPROTO_IPV4)
-		return -NF_ACCEPT;
-
 	if (!nf_ct_is_confirmed(ct)) {
 		unsigned int *timeouts = nf_ct_timeout_lookup(ct);
 
diff --git a/net/netfilter/nf_flow_table_core.c b/net/netfilter/nf_flow_table_core.c
index 8fa7bf9d5f3f..1bce1d2805c4 100644
--- a/net/netfilter/nf_flow_table_core.c
+++ b/net/netfilter/nf_flow_table_core.c
@@ -512,7 +512,7 @@ int nf_flow_table_init(struct nf_flowtable *flowtable)
 {
 	int err;
 
-	INIT_DEFERRABLE_WORK(&flowtable->gc_work, nf_flow_offload_work_gc);
+	INIT_DELAYED_WORK(&flowtable->gc_work, nf_flow_offload_work_gc);
 	flow_block_init(&flowtable->flow_block);
 	init_rwsem(&flowtable->flow_block_lock);
 
diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
index bd5e8122ea5e..fc2526b8bd55 100644
--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -6787,6 +6787,9 @@ static int nft_register_flowtable_net_hooks(struct net *net,
 
 	list_for_each_entry(hook, hook_list, list) {
 		list_for_each_entry(ft, &table->flowtables, list) {
+			if (!nft_is_active_next(net, ft))
+				continue;
+
 			list_for_each_entry(hook2, &ft->hook_list, list) {
 				if (hook->ops.dev == hook2->ops.dev &&
 				    hook->ops.pf == hook2->ops.pf) {
@@ -6846,6 +6849,7 @@ static int nft_flowtable_update(struct nft_ctx *ctx, const struct nlmsghdr *nlh,
 	struct nft_hook *hook, *next;
 	struct nft_trans *trans;
 	bool unregister = false;
+	u32 flags;
 	int err;
 
 	err = nft_flowtable_parse_hook(ctx, nla[NFTA_FLOWTABLE_HOOK],
@@ -6860,6 +6864,17 @@ static int nft_flowtable_update(struct nft_ctx *ctx, const struct nlmsghdr *nlh,
 		}
 	}
 
+	if (nla[NFTA_FLOWTABLE_FLAGS]) {
+		flags = ntohl(nla_get_be32(nla[NFTA_FLOWTABLE_FLAGS]));
+		if (flags & ~NFT_FLOWTABLE_MASK)
+			return -EOPNOTSUPP;
+		if ((flowtable->data.flags & NFT_FLOWTABLE_HW_OFFLOAD) ^
+		    (flags & NFT_FLOWTABLE_HW_OFFLOAD))
+			return -EOPNOTSUPP;
+	} else {
+		flags = flowtable->data.flags;
+	}
+
 	err = nft_register_flowtable_net_hooks(ctx->net, ctx->table,
 					       &flowtable_hook.list, flowtable);
 	if (err < 0)
@@ -6873,6 +6888,7 @@ static int nft_flowtable_update(struct nft_ctx *ctx, const struct nlmsghdr *nlh,
 		goto err_flowtable_update_hook;
 	}
 
+	nft_trans_flowtable_flags(trans) = flags;
 	nft_trans_flowtable(trans) = flowtable;
 	nft_trans_flowtable_update(trans) = true;
 	INIT_LIST_HEAD(&nft_trans_flowtable_hooks(trans));
@@ -6967,8 +6983,10 @@ static int nf_tables_newflowtable(struct net *net, struct sock *nlsk,
 	if (nla[NFTA_FLOWTABLE_FLAGS]) {
 		flowtable->data.flags =
 			ntohl(nla_get_be32(nla[NFTA_FLOWTABLE_FLAGS]));
-		if (flowtable->data.flags & ~NFT_FLOWTABLE_MASK)
+		if (flowtable->data.flags & ~NFT_FLOWTABLE_MASK) {
+			err = -EOPNOTSUPP;
 			goto err3;
+		}
 	}
 
 	write_pnet(&flowtable->data.net, net);
@@ -8179,6 +8197,8 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb)
 			break;
 		case NFT_MSG_NEWFLOWTABLE:
 			if (nft_trans_flowtable_update(trans)) {
+				nft_trans_flowtable(trans)->data.flags =
+					nft_trans_flowtable_flags(trans);
 				nf_tables_flowtable_notify(&trans->ctx,
 							   nft_trans_flowtable(trans),
 							   &nft_trans_flowtable_hooks(trans),
diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c
index bce6ca203d46..6bd31a7a27fc 100644
--- a/net/netfilter/x_tables.c
+++ b/net/netfilter/x_tables.c
@@ -1351,14 +1351,6 @@ struct xt_counters *xt_counters_alloc(unsigned int counters)
 }
 EXPORT_SYMBOL(xt_counters_alloc);
 
-struct xt_table_info
-*xt_table_get_private_protected(const struct xt_table *table)
-{
-	return rcu_dereference_protected(table->private,
-					 mutex_is_locked(&xt[table->af].mutex));
-}
-EXPORT_SYMBOL(xt_table_get_private_protected);
-
 struct xt_table_info *
 xt_replace_table(struct xt_table *table,
 	      unsigned int num_counters,
@@ -1366,6 +1358,7 @@ xt_replace_table(struct xt_table *table,
 	      int *error)
 {
 	struct xt_table_info *private;
+	unsigned int cpu;
 	int ret;
 
 	ret = xt_jumpstack_alloc(newinfo);
@@ -1375,20 +1368,47 @@ xt_replace_table(struct xt_table *table,
 	}
 
 	/* Do the substitution. */
-	private = xt_table_get_private_protected(table);
+	local_bh_disable();
+	private = table->private;
 
 	/* Check inside lock: is the old number correct? */
 	if (num_counters != private->number) {
 		pr_debug("num_counters != table->private->number (%u/%u)\n",
 			 num_counters, private->number);
+		local_bh_enable();
 		*error = -EAGAIN;
 		return NULL;
 	}
 
 	newinfo->initial_entries = private->initial_entries;
+	/*
+	 * Ensure contents of newinfo are visible before assigning to
+	 * private.
+	 */
+	smp_wmb();
+	table->private = newinfo;
+
+	/* make sure all cpus see new ->private value */
+	smp_mb();
 
-	rcu_assign_pointer(table->private, newinfo);
-	synchronize_rcu();
+	/*
+	 * Even though table entries have now been swapped, other CPU's
+	 * may still be using the old entries...
+	 */
+	local_bh_enable();
+
+	/* ... so wait for even xt_recseq on all cpus */
+	for_each_possible_cpu(cpu) {
+		seqcount_t *s = &per_cpu(xt_recseq, cpu);
+		u32 seq = raw_read_seqcount(s);
+
+		if (seq & 1) {
+			do {
+				cond_resched();
+				cpu_relax();
+			} while (seq == raw_read_seqcount(s));
+		}
+	}
 
 	audit_log_nfcfg(table->name, table->af, private->number,
 			!private->number ? AUDIT_XT_OP_REGISTER :
@@ -1424,12 +1444,12 @@ struct xt_table *xt_register_table(struct net *net,
 	}
 
 	/* Simplifies replace_table code. */
-	rcu_assign_pointer(table->private, bootstrap);
+	table->private = bootstrap;
 
 	if (!xt_replace_table(table, 0, newinfo, &ret))
 		goto unlock;
 
-	private = xt_table_get_private_protected(table);
+	private = table->private;
 	pr_debug("table->private->number = %u\n", private->number);
 
 	/* save number of initial entries */
@@ -1452,8 +1472,7 @@ void *xt_unregister_table(struct xt_table *table)
 	struct xt_table_info *private;
 
 	mutex_lock(&xt[table->af].mutex);
-	private = xt_table_get_private_protected(table);
-	RCU_INIT_POINTER(table->private, NULL);
+	private = table->private;
 	list_del(&table->list);
 	mutex_unlock(&xt[table->af].mutex);
 	audit_log_nfcfg(table->name, table->af, private->number,