Merge git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf

Pablo Neira Ayuso says:

====================
Netfilter fixes for net (v2)

The following patchset contains Netfilter fixes for net:

1) Move back the defrag users fields to the global netns_nf area.
   Kernel fails to boot if conntrack is builtin and kernel is booted
   with: nf_conntrack.enable_hooks=1. From Florian Westphal.

2) Rule event notification is missing relevant context such as
   the position handle and the NLM_F_APPEND flag.

3) Rule replacement is expanded to add + delete using the existing
   rule handle, reverse order of this operation so it makes sense
   from rule notification standpoint.

4) Propagate to userspace the NLM_F_CREATE and NLM_F_EXCL flags
   from the rule notification path.

Patches #2, #3 and #4 are used by 'nft monitor' and 'iptables-monitor'
userspace utilities which are not correctly representing the following
operations through netlink notifications:

- rule insertions
- rule addition/insertion from position handle
- create table/chain/set/map/flowtable/...
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

include/net/netfilter/ipv6/nf_defrag_ipv6.h

@@ -17,7 +17,6 @@ struct inet_frags_ctl;
 struct nft_ct_frag6_pernet {
 	struct ctl_table_header *nf_frag_frags_hdr;
 	struct fqdir	*fqdir;
-	unsigned int users;
 };
 
 #endif /* _NF_DEFRAG_IPV6_H */

include/net/netfilter/nf_tables.h

@@ -1202,7 +1202,7 @@ struct nft_object *nft_obj_lookup(const struct net *net,
 
 void nft_obj_notify(struct net *net, const struct nft_table *table,
 		    struct nft_object *obj, u32 portid, u32 seq,
-		    int event, int family, int report, gfp_t gfp);
+		    int event, u16 flags, int family, int report, gfp_t gfp);
 
 /**
  *	struct nft_object_type - stateful object type

include/net/netns/netfilter.h

@@ -27,5 +27,11 @@ struct netns_nf {
 #if IS_ENABLED(CONFIG_DECNET)
 	struct nf_hook_entries __rcu *hooks_decnet[NF_DN_NUMHOOKS];
 #endif
+#if IS_ENABLED(CONFIG_NF_DEFRAG_IPV4)
+	unsigned int defrag_ipv4_users;
+#endif
+#if IS_ENABLED(CONFIG_NF_DEFRAG_IPV6)
+	unsigned int defrag_ipv6_users;
+#endif
 };
 #endif

net/ipv4/netfilter/nf_defrag_ipv4.c

@@ -20,13 +20,8 @@
 #endif
 #include <net/netfilter/nf_conntrack_zones.h>
 
-static unsigned int defrag4_pernet_id __read_mostly;
 static DEFINE_MUTEX(defrag4_mutex);
 
-struct defrag4_pernet {
-	unsigned int users;
-};
-
 static int nf_ct_ipv4_gather_frags(struct net *net, struct sk_buff *skb,
 				   u_int32_t user)
 {
@@ -111,19 +106,15 @@ static const struct nf_hook_ops ipv4_defrag_ops[] = {
 
 static void __net_exit defrag4_net_exit(struct net *net)
 {
-	struct defrag4_pernet *nf_defrag = net_generic(net, defrag4_pernet_id);
-
-	if (nf_defrag->users) {
+	if (net->nf.defrag_ipv4_users) {
 		nf_unregister_net_hooks(net, ipv4_defrag_ops,
 					ARRAY_SIZE(ipv4_defrag_ops));
-		nf_defrag->users = 0;
+		net->nf.defrag_ipv4_users = 0;
 	}
 }
 
 static struct pernet_operations defrag4_net_ops = {
 	.exit = defrag4_net_exit,
-	.id   = &defrag4_pernet_id,
-	.size = sizeof(struct defrag4_pernet),
 };
 
 static int __init nf_defrag_init(void)
@@ -138,24 +129,23 @@ static void __exit nf_defrag_fini(void)
 
 int nf_defrag_ipv4_enable(struct net *net)
 {
-	struct defrag4_pernet *nf_defrag = net_generic(net, defrag4_pernet_id);
 	int err = 0;
 
 	mutex_lock(&defrag4_mutex);
-	if (nf_defrag->users == UINT_MAX) {
+	if (net->nf.defrag_ipv4_users == UINT_MAX) {
 		err = -EOVERFLOW;
 		goto out_unlock;
 	}
 
-	if (nf_defrag->users) {
-		nf_defrag->users++;
+	if (net->nf.defrag_ipv4_users) {
+		net->nf.defrag_ipv4_users++;
 		goto out_unlock;
 	}
 
 	err = nf_register_net_hooks(net, ipv4_defrag_ops,
 				    ARRAY_SIZE(ipv4_defrag_ops));
 	if (err == 0)
-		nf_defrag->users = 1;
+		net->nf.defrag_ipv4_users = 1;
 
  out_unlock:
 	mutex_unlock(&defrag4_mutex);
@@ -165,12 +155,10 @@ EXPORT_SYMBOL_GPL(nf_defrag_ipv4_enable);
 
 void nf_defrag_ipv4_disable(struct net *net)
 {
-	struct defrag4_pernet *nf_defrag = net_generic(net, defrag4_pernet_id);
-
 	mutex_lock(&defrag4_mutex);
-	if (nf_defrag->users) {
-		nf_defrag->users--;
-		if (nf_defrag->users == 0)
+	if (net->nf.defrag_ipv4_users) {
+		net->nf.defrag_ipv4_users--;
+		if (net->nf.defrag_ipv4_users == 0)
 			nf_unregister_net_hooks(net, ipv4_defrag_ops,
 						ARRAY_SIZE(ipv4_defrag_ops));
 	}

net/ipv6/netfilter/nf_conntrack_reasm.c

@@ -33,7 +33,7 @@
 
 static const char nf_frags_cache_name[] = "nf-frags";
 
-unsigned int nf_frag_pernet_id __read_mostly;
+static unsigned int nf_frag_pernet_id __read_mostly;
 static struct inet_frags nf_frags;
 
 static struct nft_ct_frag6_pernet *nf_frag_pernet(struct net *net)

net/ipv6/netfilter/nf_defrag_ipv6_hooks.c

@@ -25,8 +25,6 @@
 #include <net/netfilter/nf_conntrack_zones.h>
 #include <net/netfilter/ipv6/nf_defrag_ipv6.h>
 
-extern unsigned int nf_frag_pernet_id;
-
 static DEFINE_MUTEX(defrag6_mutex);
 
 static enum ip6_defrag_users nf_ct6_defrag_user(unsigned int hooknum,
@@ -91,12 +89,10 @@ static const struct nf_hook_ops ipv6_defrag_ops[] = {
 
 static void __net_exit defrag6_net_exit(struct net *net)
 {
-	struct nft_ct_frag6_pernet *nf_frag = net_generic(net, nf_frag_pernet_id);
-
-	if (nf_frag->users) {
+	if (net->nf.defrag_ipv6_users) {
 		nf_unregister_net_hooks(net, ipv6_defrag_ops,
 					ARRAY_SIZE(ipv6_defrag_ops));
-		nf_frag->users = 0;
+		net->nf.defrag_ipv6_users = 0;
 	}
 }
 
@@ -134,24 +130,23 @@ static void __exit nf_defrag_fini(void)
 
 int nf_defrag_ipv6_enable(struct net *net)
 {
-	struct nft_ct_frag6_pernet *nf_frag = net_generic(net, nf_frag_pernet_id);
 	int err = 0;
 
 	mutex_lock(&defrag6_mutex);
-	if (nf_frag->users == UINT_MAX) {
+	if (net->nf.defrag_ipv6_users == UINT_MAX) {
 		err = -EOVERFLOW;
 		goto out_unlock;
 	}
 
-	if (nf_frag->users) {
-		nf_frag->users++;
+	if (net->nf.defrag_ipv6_users) {
+		net->nf.defrag_ipv6_users++;
 		goto out_unlock;
 	}
 
 	err = nf_register_net_hooks(net, ipv6_defrag_ops,
 				    ARRAY_SIZE(ipv6_defrag_ops));
 	if (err == 0)
-		nf_frag->users = 1;
+		net->nf.defrag_ipv6_users = 1;
 
  out_unlock:
 	mutex_unlock(&defrag6_mutex);
@@ -161,12 +156,10 @@ EXPORT_SYMBOL_GPL(nf_defrag_ipv6_enable);
 
 void nf_defrag_ipv6_disable(struct net *net)
 {
-	struct nft_ct_frag6_pernet *nf_frag = net_generic(net, nf_frag_pernet_id);
-
 	mutex_lock(&defrag6_mutex);
-	if (nf_frag->users) {
-		nf_frag->users--;
-		if (nf_frag->users == 0)
+	if (net->nf.defrag_ipv6_users) {
+		net->nf.defrag_ipv6_users--;
+		if (net->nf.defrag_ipv6_users == 0)
 			nf_unregister_net_hooks(net, ipv6_defrag_ops,
 						ARRAY_SIZE(ipv6_defrag_ops));
 	}

net/netfilter/nf_tables_api.c

@@ -780,6 +780,7 @@ static void nf_tables_table_notify(const struct nft_ctx *ctx, int event)
 {
 	struct nftables_pernet *nft_net;
 	struct sk_buff *skb;
+	u16 flags = 0;
 	int err;
 
 	if (!ctx->report &&
@@ -790,8 +791,11 @@ static void nf_tables_table_notify(const struct nft_ctx *ctx, int event)
 	if (skb == NULL)
 		goto err;
 
+	if (ctx->flags & (NLM_F_CREATE | NLM_F_EXCL))
+		flags |= ctx->flags & (NLM_F_CREATE | NLM_F_EXCL);
+
 	err = nf_tables_fill_table_info(skb, ctx->net, ctx->portid, ctx->seq,
-					event, 0, ctx->family, ctx->table);
+					event, flags, ctx->family, ctx->table);
 	if (err < 0) {
 		kfree_skb(skb);
 		goto err;
@@ -1563,6 +1567,7 @@ static void nf_tables_chain_notify(const struct nft_ctx *ctx, int event)
 {
 	struct nftables_pernet *nft_net;
 	struct sk_buff *skb;
+	u16 flags = 0;
 	int err;
 
 	if (!ctx->report &&
@@ -1573,8 +1578,11 @@ static void nf_tables_chain_notify(const struct nft_ctx *ctx, int event)
 	if (skb == NULL)
 		goto err;
 
+	if (ctx->flags & (NLM_F_CREATE | NLM_F_EXCL))
+		flags |= ctx->flags & (NLM_F_CREATE | NLM_F_EXCL);
+
 	err = nf_tables_fill_chain_info(skb, ctx->net, ctx->portid, ctx->seq,
-					event, 0, ctx->family, ctx->table,
+					event, flags, ctx->family, ctx->table,
 					ctx->chain);
 	if (err < 0) {
 		kfree_skb(skb);
@@ -2866,8 +2874,7 @@ static int nf_tables_fill_rule_info(struct sk_buff *skb, struct net *net,
 				    u32 flags, int family,
 				    const struct nft_table *table,
 				    const struct nft_chain *chain,
-				    const struct nft_rule *rule,
-				    const struct nft_rule *prule)
+				    const struct nft_rule *rule, u64 handle)
 {
 	struct nlmsghdr *nlh;
 	const struct nft_expr *expr, *next;
@@ -2887,9 +2894,8 @@ static int nf_tables_fill_rule_info(struct sk_buff *skb, struct net *net,
 			 NFTA_RULE_PAD))
 		goto nla_put_failure;
 
-	if (event != NFT_MSG_DELRULE && prule) {
-		if (nla_put_be64(skb, NFTA_RULE_POSITION,
-				 cpu_to_be64(prule->handle),
+	if (event != NFT_MSG_DELRULE && handle) {
+		if (nla_put_be64(skb, NFTA_RULE_POSITION, cpu_to_be64(handle),
 				 NFTA_RULE_PAD))
 			goto nla_put_failure;
 	}
@@ -2925,7 +2931,10 @@ static void nf_tables_rule_notify(const struct nft_ctx *ctx,
 				  const struct nft_rule *rule, int event)
 {
 	struct nftables_pernet *nft_net = nft_pernet(ctx->net);
+	const struct nft_rule *prule;
 	struct sk_buff *skb;
+	u64 handle = 0;
+	u16 flags = 0;
 	int err;
 
 	if (!ctx->report &&
@@ -2936,9 +2945,20 @@ static void nf_tables_rule_notify(const struct nft_ctx *ctx,
 	if (skb == NULL)
 		goto err;
 
+	if (event == NFT_MSG_NEWRULE &&
+	    !list_is_first(&rule->list, &ctx->chain->rules) &&
+	    !list_is_last(&rule->list, &ctx->chain->rules)) {
+		prule = list_prev_entry(rule, list);
+		handle = prule->handle;
+	}
+	if (ctx->flags & (NLM_F_APPEND | NLM_F_REPLACE))
+		flags |= NLM_F_APPEND;
+	if (ctx->flags & (NLM_F_CREATE | NLM_F_EXCL))
+		flags |= ctx->flags & (NLM_F_CREATE | NLM_F_EXCL);
+
 	err = nf_tables_fill_rule_info(skb, ctx->net, ctx->portid, ctx->seq,
-				       event, 0, ctx->family, ctx->table,
-				       ctx->chain, rule, NULL);
+				       event, flags, ctx->family, ctx->table,
+				       ctx->chain, rule, handle);
 	if (err < 0) {
 		kfree_skb(skb);
 		goto err;
@@ -2964,6 +2984,7 @@ static int __nf_tables_dump_rules(struct sk_buff *skb,
 	struct net *net = sock_net(skb->sk);
 	const struct nft_rule *rule, *prule;
 	unsigned int s_idx = cb->args[0];
+	u64 handle;
 
 	prule = NULL;
 	list_for_each_entry_rcu(rule, &chain->rules, list) {
@@ -2975,12 +2996,17 @@ static int __nf_tables_dump_rules(struct sk_buff *skb,
 			memset(&cb->args[1], 0,
 					sizeof(cb->args) - sizeof(cb->args[0]));
 		}
+		if (prule)
+			handle = prule->handle;
+		else
+			handle = 0;
+
 		if (nf_tables_fill_rule_info(skb, net, NETLINK_CB(cb->skb).portid,
 					cb->nlh->nlmsg_seq,
 					NFT_MSG_NEWRULE,
 					NLM_F_MULTI | NLM_F_APPEND,
 					table->family,
-					table, chain, rule, prule) < 0)
+					table, chain, rule, handle) < 0)
 			return 1;
 
 		nl_dump_check_consistent(cb, nlmsg_hdr(skb));
@@ -3143,7 +3169,7 @@ static int nf_tables_getrule(struct sk_buff *skb, const struct nfnl_info *info,
 
 	err = nf_tables_fill_rule_info(skb2, net, NETLINK_CB(skb).portid,
 				       info->nlh->nlmsg_seq, NFT_MSG_NEWRULE, 0,
-				       family, table, chain, rule, NULL);
+				       family, table, chain, rule, 0);
 	if (err < 0)
 		goto err_fill_rule_info;
 
@@ -3403,17 +3429,15 @@ static int nf_tables_newrule(struct sk_buff *skb, const struct nfnl_info *info,
 	}
 
 	if (info->nlh->nlmsg_flags & NLM_F_REPLACE) {
+		err = nft_delrule(&ctx, old_rule);
+		if (err < 0)
+			goto err_destroy_flow_rule;
+
 		trans = nft_trans_rule_add(&ctx, NFT_MSG_NEWRULE, rule);
 		if (trans == NULL) {
 			err = -ENOMEM;
 			goto err_destroy_flow_rule;
 		}
-		err = nft_delrule(&ctx, old_rule);
-		if (err < 0) {
-			nft_trans_destroy(trans);
-			goto err_destroy_flow_rule;
-		}
-
 		list_add_tail_rcu(&rule->list, &old_rule->list);
 	} else {
 		trans = nft_trans_rule_add(&ctx, NFT_MSG_NEWRULE, rule);
@@ -3943,8 +3967,9 @@ static void nf_tables_set_notify(const struct nft_ctx *ctx,
 			         gfp_t gfp_flags)
 {
 	struct nftables_pernet *nft_net = nft_pernet(ctx->net);
-	struct sk_buff *skb;
 	u32 portid = ctx->portid;
+	struct sk_buff *skb;
+	u16 flags = 0;
 	int err;
 
 	if (!ctx->report &&
@@ -3955,7 +3980,10 @@ static void nf_tables_set_notify(const struct nft_ctx *ctx,
 	if (skb == NULL)
 		goto err;
 
-	err = nf_tables_fill_set(skb, ctx, set, event, 0);
+	if (ctx->flags & (NLM_F_CREATE | NLM_F_EXCL))
+		flags |= ctx->flags & (NLM_F_CREATE | NLM_F_EXCL);
+
+	err = nf_tables_fill_set(skb, ctx, set, event, flags);
 	if (err < 0) {
 		kfree_skb(skb);
 		goto err;
@@ -5231,12 +5259,13 @@ static int nf_tables_getsetelem(struct sk_buff *skb,
 static void nf_tables_setelem_notify(const struct nft_ctx *ctx,
 				     const struct nft_set *set,
 				     const struct nft_set_elem *elem,
-				     int event, u16 flags)
+				     int event)
 {
 	struct nftables_pernet *nft_net;
 	struct net *net = ctx->net;
 	u32 portid = ctx->portid;
 	struct sk_buff *skb;
+	u16 flags = 0;
 	int err;
 
 	if (!ctx->report && !nfnetlink_has_listeners(net, NFNLGRP_NFTABLES))
@@ -5246,6 +5275,9 @@ static void nf_tables_setelem_notify(const struct nft_ctx *ctx,
 	if (skb == NULL)
 		goto err;
 
+	if (ctx->flags & (NLM_F_CREATE | NLM_F_EXCL))
+		flags |= ctx->flags & (NLM_F_CREATE | NLM_F_EXCL);
+
 	err = nf_tables_fill_setelem_info(skb, ctx, 0, portid, event, flags,
 					  set, elem);
 	if (err < 0) {
@@ -6921,7 +6953,7 @@ static int nf_tables_delobj(struct sk_buff *skb, const struct nfnl_info *info,
 
 void nft_obj_notify(struct net *net, const struct nft_table *table,
 		    struct nft_object *obj, u32 portid, u32 seq, int event,
-		    int family, int report, gfp_t gfp)
+		    u16 flags, int family, int report, gfp_t gfp)
 {
 	struct nftables_pernet *nft_net = nft_pernet(net);
 	struct sk_buff *skb;
@@ -6946,8 +6978,9 @@ void nft_obj_notify(struct net *net, const struct nft_table *table,
 	if (skb == NULL)
 		goto err;
 
-	err = nf_tables_fill_obj_info(skb, net, portid, seq, event, 0, family,
-				      table, obj, false);
+	err = nf_tables_fill_obj_info(skb, net, portid, seq, event,
+				      flags & (NLM_F_CREATE | NLM_F_EXCL),
+				      family, table, obj, false);
 	if (err < 0) {
 		kfree_skb(skb);
 		goto err;
@@ -6964,7 +6997,7 @@ static void nf_tables_obj_notify(const struct nft_ctx *ctx,
 				 struct nft_object *obj, int event)
 {
 	nft_obj_notify(ctx->net, ctx->table, obj, ctx->portid, ctx->seq, event,
-		       ctx->family, ctx->report, GFP_KERNEL);
+		       ctx->flags, ctx->family, ctx->report, GFP_KERNEL);
 }
 
 /*
@@ -7745,6 +7778,7 @@ static void nf_tables_flowtable_notify(struct nft_ctx *ctx,
 {
 	struct nftables_pernet *nft_net = nft_pernet(ctx->net);
 	struct sk_buff *skb;
+	u16 flags = 0;
 	int err;
 
 	if (!ctx->report &&
@@ -7755,8 +7789,11 @@ static void nf_tables_flowtable_notify(struct nft_ctx *ctx,
 	if (skb == NULL)
 		goto err;
 
+	if (ctx->flags & (NLM_F_CREATE | NLM_F_EXCL))
+		flags |= ctx->flags & (NLM_F_CREATE | NLM_F_EXCL);
+
 	err = nf_tables_fill_flowtable_info(skb, ctx->net, ctx->portid,
-					    ctx->seq, event, 0,
+					    ctx->seq, event, flags,
 					    ctx->family, flowtable, hook_list);
 	if (err < 0) {
 		kfree_skb(skb);
@@ -8634,7 +8671,7 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb)
 			nft_setelem_activate(net, te->set, &te->elem);
 			nf_tables_setelem_notify(&trans->ctx, te->set,
 						 &te->elem,
-						 NFT_MSG_NEWSETELEM, 0);
+						 NFT_MSG_NEWSETELEM);
 			nft_trans_destroy(trans);
 			break;
 		case NFT_MSG_DELSETELEM:
@@ -8642,7 +8679,7 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb)
 
 			nf_tables_setelem_notify(&trans->ctx, te->set,
 						 &te->elem,
-						 NFT_MSG_DELSETELEM, 0);
+						 NFT_MSG_DELSETELEM);
 			nft_setelem_remove(net, te->set, &te->elem);
 			if (!nft_setelem_is_catchall(te->set, &te->elem)) {
 				atomic_dec(&te->set->nelems);

net/netfilter/nft_quota.c

@@ -60,7 +60,7 @@ static void nft_quota_obj_eval(struct nft_object *obj,
 	if (overquota &&
 	    !test_and_set_bit(NFT_QUOTA_DEPLETED_BIT, &priv->flags))
 		nft_obj_notify(nft_net(pkt), obj->key.table, obj, 0, 0,
-			       NFT_MSG_NEWOBJ, nft_pf(pkt), 0, GFP_ATOMIC);
+			       NFT_MSG_NEWOBJ, 0, nft_pf(pkt), 0, GFP_ATOMIC);
 }
 
 static int nft_quota_do_init(const struct nlattr * const tb[],