[PATCH nf,v2 1/2] netfilter: nf_tables: move set_update_list to nftables per-netns

[PATCH nf,v2 1/2] netfilter: nf_tables: move set_update_list to nftables per-netns

[Pablo Neira Ayuso]

This list is used to invoke the set .commit and .abort ops for the
rbtree and pipapo to run GC on expired elements and replace the current
datastructure view by the clone. For the rbtree, this also rebuild the
datapath b-search array.

From abort path, remove the set from the update_list if it is already
bound to rule, then the rule itself takes care of releasing the set and
its elements, otherwise, memleak is possible because set ops .abort
only deals with removing the set data structure, not the elements.

This is a preparation patch to call set .commit before processing the
transaction list for the rbtree, no functional changes are intended.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
v2: no changes.

 include/net/netfilter/nf_tables.h |  1 +
 net/netfilter/nf_tables_api.c     | 49 ++++++++++---------------------
 2 files changed, 16 insertions(+), 34 deletions(-)

diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h
index 3be612145c13..238f6ecb90e9 100644
--- a/include/net/netfilter/nf_tables.h
+++ b/include/net/netfilter/nf_tables.h
@@ -1949,6 +1949,7 @@ struct nftables_pernet {
 	struct list_head	binding_list;
 	struct list_head	module_list;
 	struct list_head	notify_list;
+	struct list_head	set_update_list;
 	struct mutex		commit_mutex;
 	u64			table_handle;
 	u64			tstamp;
diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
index f398792673db..110c0a31f8f2 100644
--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -595,10 +595,15 @@ static void nft_trans_commit_list_add_tail(struct net *net, struct nft_trans *tr
 static void nft_trans_commit_list_add_elem(struct net *net, struct nft_trans *trans)
 {
 	struct nftables_pernet *nft_net = nft_pernet(net);
+	struct nft_trans_elem *te;
 
 	WARN_ON_ONCE(trans->msg_type != NFT_MSG_NEWSETELEM &&
 		     trans->msg_type != NFT_MSG_DELSETELEM);
 
+	te = nft_trans_container_elem(trans);
+	if (te->set->ops->commit && list_empty(&te->set->pending_update))
+		list_add_tail(&te->set->pending_update, &nft_net->set_update_list);
+
 	if (nft_trans_try_collapse(nft_net, trans)) {
 		kfree(trans);
 		return;
@@ -10848,11 +10853,11 @@ static void nf_tables_commit_audit_log(struct list_head *adl, u32 generation)
 	}
 }
 
-static void nft_set_commit_update(struct list_head *set_update_list)
+static void nft_set_commit_update(struct nftables_pernet *nft_net)
 {
 	struct nft_set *set, *next;
 
-	list_for_each_entry_safe(set, next, set_update_list, pending_update) {
+	list_for_each_entry_safe(set, next, &nft_net->set_update_list, pending_update) {
 		list_del_init(&set->pending_update);
 
 		if (!set->ops->commit || set->dead)
@@ -10885,7 +10890,6 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb)
 	struct nft_trans_binding *trans_binding;
 	struct nft_trans *trans, *next;
 	unsigned int base_seq, gc_seq;
-	LIST_HEAD(set_update_list);
 	struct nft_trans_elem *te;
 	struct nft_chain *chain;
 	struct nft_table *table;
@@ -11097,27 +11101,13 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb)
 			break;
 		case NFT_MSG_NEWSETELEM:
 			te = nft_trans_container_elem(trans);
-
 			nft_trans_elems_add(&ctx, te);
-
-			if (te->set->ops->commit &&
-			    list_empty(&te->set->pending_update)) {
-				list_add_tail(&te->set->pending_update,
-					      &set_update_list);
-			}
 			nft_trans_destroy(trans);
 			break;
 		case NFT_MSG_DELSETELEM:
 		case NFT_MSG_DESTROYSETELEM:
 			te = nft_trans_container_elem(trans);
-
 			nft_trans_elems_remove(&ctx, te);
-
-			if (te->set->ops->commit &&
-			    list_empty(&te->set->pending_update)) {
-				list_add_tail(&te->set->pending_update,
-					      &set_update_list);
-			}
 			break;
 		case NFT_MSG_NEWOBJ:
 			if (nft_trans_obj_update(trans)) {
@@ -11186,7 +11176,7 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb)
 		}
 	}
 
-	nft_set_commit_update(&set_update_list);
+	nft_set_commit_update(nft_net);
 
 	nft_commit_notify(net, NETLINK_CB(skb).portid);
 	nf_tables_gen_notify(net, skb, NFT_MSG_NEWGEN);
@@ -11253,11 +11243,11 @@ static void nf_tables_abort_release(struct nft_trans *trans)
 	kfree(trans);
 }
 
-static void nft_set_abort_update(struct list_head *set_update_list)
+static void nft_set_abort_update(struct nftables_pernet *nft_net)
 {
 	struct nft_set *set, *next;
 
-	list_for_each_entry_safe(set, next, set_update_list, pending_update) {
+	list_for_each_entry_safe(set, next, &nft_net->set_update_list, pending_update) {
 		list_del_init(&set->pending_update);
 
 		if (!set->ops->abort)
@@ -11392,33 +11382,22 @@ static int __nf_tables_abort(struct net *net, enum nfnl_abort_action action)
 			nft_trans_destroy(trans);
 			break;
 		case NFT_MSG_NEWSETELEM:
+			te = nft_trans_container_elem(trans);
 			if (nft_trans_elem_set_bound(trans)) {
+				list_del_init(&te->set->pending_update);
 				nft_trans_destroy(trans);
 				break;
 			}
-			te = nft_trans_container_elem(trans);
 			if (!nft_trans_elems_new_abort(&ctx, te)) {
 				nft_trans_destroy(trans);
 				break;
 			}
-
-			if (te->set->ops->abort &&
-			    list_empty(&te->set->pending_update)) {
-				list_add_tail(&te->set->pending_update,
-					      &set_update_list);
-			}
 			break;
 		case NFT_MSG_DELSETELEM:
 		case NFT_MSG_DESTROYSETELEM:
 			te = nft_trans_container_elem(trans);
 
 			nft_trans_elems_destroy_abort(&ctx, te);
-
-			if (te->set->ops->abort &&
-			    list_empty(&te->set->pending_update)) {
-				list_add_tail(&te->set->pending_update,
-					      &set_update_list);
-			}
 			nft_trans_destroy(trans);
 			break;
 		case NFT_MSG_NEWOBJ:
@@ -11464,7 +11443,7 @@ static int __nf_tables_abort(struct net *net, enum nfnl_abort_action action)
 
 	WARN_ON_ONCE(!list_empty(&nft_net->commit_set_list));
 
-	nft_set_abort_update(&set_update_list);
+	nft_set_abort_update(nft_net);
 
 	synchronize_rcu();
 
@@ -12148,6 +12127,7 @@ static int __net_init nf_tables_init_net(struct net *net)
 	INIT_LIST_HEAD(&nft_net->binding_list);
 	INIT_LIST_HEAD(&nft_net->module_list);
 	INIT_LIST_HEAD(&nft_net->notify_list);
+	INIT_LIST_HEAD(&nft_net->set_update_list);
 	mutex_init(&nft_net->commit_mutex);
 	net->nft.base_seq = 1;
 	nft_net->gc_seq = 0;
@@ -12192,6 +12172,7 @@ static void __net_exit nf_tables_exit_net(struct net *net)
 	WARN_ON_ONCE(!list_empty(&nft_net->module_list));
 	WARN_ON_ONCE(!list_empty(&nft_net->notify_list));
 	WARN_ON_ONCE(!list_empty(&nft_net->destroy_list));
+	WARN_ON_ONCE(!list_empty(&nft_net->set_update_list));
 }
 
 static void nf_tables_exit_batch(struct list_head *net_exit_list)
-- 
2.47.3

[PATCH nf,v2 2/2] netfilter: nf_tables: call set ops .commit when building new ruleset blob

[Pablo Neira Ayuso]

The rbtree set only builds the b-search array after the new ruleset has
been published through set ops .commit.

This exposes an empty set for a short time span which results in a bogus
mismatch for the following batch:

 destroy table ip x
 table ip x {
	...
 }

The same problem also affects the pipapo set backend which also provides
a set ops .commit interface too.

This patch moves the set ops .commit call right before building and
publishing the chain blob. The commit path now performs an early
handling of the DELSETELEM command to remove stale elements from the
clone before it is published via rcu. Note that DELSETELEM notifications
are still delivered in order. NEWSETELEM commands are handled after the
set is published, since this clears the previous genbit to 1 to prepare
the element for the next control plane transaction. This comes at the
cost of one extra iteration over the transaction list.

Suggested-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
v2: fix double notification on deleted elements.

 net/netfilter/nf_tables_api.c | 56 ++++++++++++++++++++++++++++++-----
 1 file changed, 48 insertions(+), 8 deletions(-)

diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
index 110c0a31f8f2..7c33b523f678 100644
--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -7191,16 +7191,19 @@ static void nft_setelem_remove(const struct net *net,
 }
 
 static void nft_trans_elems_remove(const struct nft_ctx *ctx,
-				   const struct nft_trans_elem *te)
+				   const struct nft_trans_elem *te,
+				   bool notify)
 {
 	int i;
 
 	for (i = 0; i < te->nelems; i++) {
 		WARN_ON_ONCE(te->elems[i].update);
 
-		nf_tables_setelem_notify(ctx, te->set,
-					 te->elems[i].priv,
-					 te->nft_trans.msg_type);
+		if (notify) {
+			nf_tables_setelem_notify(ctx, te->set,
+						 te->elems[i].priv,
+						 te->nft_trans.msg_type);
+		}
 
 		nft_setelem_remove(ctx->net, te->set, te->elems[i].priv);
 		if (!nft_setelem_is_catchall(te->set, te->elems[i].priv)) {
@@ -7210,6 +7213,20 @@ static void nft_trans_elems_remove(const struct nft_ctx *ctx,
 	}
 }
 
+static void nft_trans_elems_remove_notify(const struct nft_ctx *ctx,
+					  const struct nft_trans_elem *te)
+{
+	int i;
+
+	for (i = 0; i < te->nelems; i++) {
+		WARN_ON_ONCE(te->elems[i].update);
+
+		nf_tables_setelem_notify(ctx, te->set,
+					 te->elems[i].priv,
+					 te->nft_trans.msg_type);
+	}
+}
+
 static bool nft_setelem_valid_key_end(const struct nft_set *set,
 				      struct nlattr **nla, u32 flags)
 {
@@ -10853,9 +10870,29 @@ static void nf_tables_commit_audit_log(struct list_head *adl, u32 generation)
 	}
 }
 
-static void nft_set_commit_update(struct nftables_pernet *nft_net)
+static void nft_set_commit_update(struct nft_ctx *ctx,
+				  struct nftables_pernet *nft_net)
 {
 	struct nft_set *set, *next;
+	struct nft_trans_elem *te;
+	struct nft_trans *trans;
+
+	if (list_empty(&nft_net->set_update_list))
+		return;
+
+	list_for_each_entry(trans, &nft_net->commit_list, list) {
+		nft_ctx_update(ctx, trans);
+
+		switch (trans->msg_type) {
+		case NFT_MSG_DELSETELEM:
+			te = nft_trans_container_elem(trans);
+			if (!te->set->ops->commit)
+				break;
+
+			nft_trans_elems_remove(ctx, te, false);
+			break;
+		}
+	}
 
 	list_for_each_entry_safe(set, next, &nft_net->set_update_list, pending_update) {
 		list_del_init(&set->pending_update);
@@ -10970,6 +11007,8 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb)
 	}
 
 	/* step 2.  Make rules_gen_X visible to packet path */
+	nft_set_commit_update(&ctx, nft_net);
+
 	list_for_each_entry(table, &nft_net->tables, list) {
 		list_for_each_entry(chain, &table->chains, list)
 			nf_tables_commit_chain(net, chain);
@@ -11107,7 +11146,10 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb)
 		case NFT_MSG_DELSETELEM:
 		case NFT_MSG_DESTROYSETELEM:
 			te = nft_trans_container_elem(trans);
-			nft_trans_elems_remove(&ctx, te);
+			if (te->set->ops->commit)
+				nft_trans_elems_remove_notify(&ctx, te);
+			else
+				nft_trans_elems_remove(&ctx, te, true);
 			break;
 		case NFT_MSG_NEWOBJ:
 			if (nft_trans_obj_update(trans)) {
@@ -11176,8 +11218,6 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb)
 		}
 	}
 
-	nft_set_commit_update(nft_net);
-
 	nft_commit_notify(net, NETLINK_CB(skb).portid);
 	nf_tables_gen_notify(net, skb, NFT_MSG_NEWGEN);
 	nf_tables_commit_audit_log(&adl, nft_base_seq(net));
-- 
2.47.3