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

Pablo Neira Ayuso says:

====================
Netfilter fixes for net

The following patchset contains Netfilter fixes for net:

1) Missing netlink attribute sanity check for NFTA_OSF_DREG,
from Florian Westphal.

2) Use bitmap infrastructure in ipset to fix KASAN slab-out-of-bounds
reads, from Jozsef Kadlecsik.

3) Missing initial CLOSED state in new sctp connection through
ctnetlink events, from Jiri Wiesner.

4) Missing check for NFT_CHAIN_HW_OFFLOAD in nf_tables offload
indirect block infrastructure, from wenxu.

5) Add __nft_chain_type_get() to sanity check family and chain type.

6) Autoload modules from the nf_tables abort path to fix races
reported by syzbot.

7) Remove unnecessary skb->csum update on inet_proto_csum_replace16(),
from Praveen Chaudhary.
====================

Signed-off-by: David S. Miller <[e-mail]>

include/linux/netfilter/ipset/ip_set.h

···
 	       sizeof(*addr));
 }
 
-/* Calculate the bytes required to store the inclusive range of a-b */
-static inline int
-bitmap_bytes(u32 a, u32 b)
-{
-	return 4 * ((((b - a + 8) / 8) + 3) / 4);
-}
-
 /* How often should the gc be run by default */
 #define IPSET_GC_TIME			(3 * 60)
 

include/linux/netfilter/nfnetlink.h

···
 	const struct nfnl_callback *cb;	/* callback for individual types */
 	struct module *owner;
 	int (*commit)(struct net *net, struct sk_buff *skb);
-	int (*abort)(struct net *net, struct sk_buff *skb);
+	int (*abort)(struct net *net, struct sk_buff *skb, bool autoload);
 	void (*cleanup)(struct net *net);
 	bool (*valid_genid)(struct net *net, u32 genid);
 };

include/net/netns/nftables.h

···
 struct netns_nftables {
 	struct list_head	tables;
 	struct list_head	commit_list;
+	struct list_head	module_list;
 	struct mutex		commit_mutex;
 	unsigned int		base_seq;
 	u8			gencursor;

net/core/utils.c

···
 }
 EXPORT_SYMBOL(inet_proto_csum_replace4);
 
+/**
+ * inet_proto_csum_replace16 - update layer 4 header checksum field
+ * @sum: Layer 4 header checksum field
+ * @skb: sk_buff for the packet
+ * @from: old IPv6 address
+ * @to: new IPv6 address
+ * @pseudohdr: True if layer 4 header checksum includes pseudoheader
+ *
+ * Update layer 4 header as per the update in IPv6 src/dst address.
+ *
+ * There is no need to update skb->csum in this function, because update in two
+ * fields a.) IPv6 src/dst address and b.) L4 header checksum cancels each other
+ * for skb->csum calculation. Whereas inet_proto_csum_replace4 function needs to
+ * update skb->csum, because update in 3 fields a.) IPv4 src/dst address,
+ * b.) IPv4 Header checksum and c.) L4 header checksum results in same diff as
+ * L4 Header checksum for skb->csum calculation.
+ */
 void inet_proto_csum_replace16(__sum16 *sum, struct sk_buff *skb,
 			       const __be32 *from, const __be32 *to,
 			       bool pseudohdr)
···
 	if (skb->ip_summed != CHECKSUM_PARTIAL) {
 		*sum = csum_fold(csum_partial(diff, sizeof(diff),
 				 ~csum_unfold(*sum)));
-		if (skb->ip_summed == CHECKSUM_COMPLETE && pseudohdr)
-			skb->csum = ~csum_partial(diff, sizeof(diff),
-						  ~skb->csum);
 	} else if (pseudohdr)
 		*sum = ~csum_fold(csum_partial(diff, sizeof(diff),
 				  csum_unfold(*sum)));

net/netfilter/ipset/ip_set_bitmap_gen.h

···
 
 	if (set->extensions & IPSET_EXT_DESTROY)
 		mtype_ext_cleanup(set);
-	memset(map->members, 0, map->memsize);
+	bitmap_zero(map->members, map->elements);
 	set->elements = 0;
 	set->ext_size = 0;
 }

net/netfilter/ipset/ip_set_bitmap_ip.c

···
 
 /* Type structure */
 struct bitmap_ip {
-	void *members;		/* the set members */
+	unsigned long *members;	/* the set members */
 	u32 first_ip;		/* host byte order, included in range */
 	u32 last_ip;		/* host byte order, included in range */
 	u32 elements;		/* number of max elements in the set */
···
 	    u32 first_ip, u32 last_ip,
 	    u32 elements, u32 hosts, u8 netmask)
 {
-	map->members = ip_set_alloc(map->memsize);
+	map->members = bitmap_zalloc(elements, GFP_KERNEL | __GFP_NOWARN);
 	if (!map->members)
 		return false;
 	map->first_ip = first_ip;
···
 	if (!map)
 		return -ENOMEM;
 
-	map->memsize = bitmap_bytes(0, elements - 1);
+	map->memsize = BITS_TO_LONGS(elements) * sizeof(unsigned long);
 	set->variant = &bitmap_ip;
 	if (!init_map_ip(set, map, first_ip, last_ip,
 			 elements, hosts, netmask)) {

net/netfilter/ipset/ip_set_bitmap_ipmac.c

···
 
 /* Type structure */
 struct bitmap_ipmac {
-	void *members;		/* the set members */
+	unsigned long *members;	/* the set members */
 	u32 first_ip;		/* host byte order, included in range */
 	u32 last_ip;		/* host byte order, included in range */
 	u32 elements;		/* number of max elements in the set */
···
 init_map_ipmac(struct ip_set *set, struct bitmap_ipmac *map,
 	       u32 first_ip, u32 last_ip, u32 elements)
 {
-	map->members = ip_set_alloc(map->memsize);
+	map->members = bitmap_zalloc(elements, GFP_KERNEL | __GFP_NOWARN);
 	if (!map->members)
 		return false;
 	map->first_ip = first_ip;
···
 	if (!map)
 		return -ENOMEM;
 
-	map->memsize = bitmap_bytes(0, elements - 1);
+	map->memsize = BITS_TO_LONGS(elements) * sizeof(unsigned long);
 	set->variant = &bitmap_ipmac;
 	if (!init_map_ipmac(set, map, first_ip, last_ip, elements)) {
 		kfree(map);

net/netfilter/ipset/ip_set_bitmap_port.c

···
 
 /* Type structure */
 struct bitmap_port {
-	void *members;		/* the set members */
+	unsigned long *members;	/* the set members */
 	u16 first_port;		/* host byte order, included in range */
 	u16 last_port;		/* host byte order, included in range */
 	u32 elements;		/* number of max elements in the set */
···
 init_map_port(struct ip_set *set, struct bitmap_port *map,
 	      u16 first_port, u16 last_port)
 {
-	map->members = ip_set_alloc(map->memsize);
+	map->members = bitmap_zalloc(map->elements, GFP_KERNEL | __GFP_NOWARN);
 	if (!map->members)
 		return false;
 	map->first_port = first_port;
···
 		return -ENOMEM;
 
 	map->elements = elements;
-	map->memsize = bitmap_bytes(0, map->elements);
+	map->memsize = BITS_TO_LONGS(elements) * sizeof(unsigned long);
 	set->variant = &bitmap_port;
 	if (!init_map_port(set, map, first_port, last_port)) {
 		kfree(map);

net/netfilter/nf_conntrack_proto_sctp.c

···
 	{
 /*	ORIGINAL	*/
 /*                  sNO, sCL, sCW, sCE, sES, sSS, sSR, sSA, sHS, sHA */
-/* init         */ {sCW, sCW, sCW, sCE, sES, sSS, sSR, sSA, sCW, sHA},
+/* init         */ {sCL, sCL, sCW, sCE, sES, sSS, sSR, sSA, sCW, sHA},
 /* init_ack     */ {sCL, sCL, sCW, sCE, sES, sSS, sSR, sSA, sCL, sHA},
 /* abort        */ {sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL},
 /* shutdown     */ {sCL, sCL, sCW, sCE, sSS, sSS, sSR, sSA, sCL, sSS},
···
 /*	REPLY	*/
 /*                  sNO, sCL, sCW, sCE, sES, sSS, sSR, sSA, sHS, sHA */
 /* init         */ {sIV, sCL, sCW, sCE, sES, sSS, sSR, sSA, sIV, sHA},/* INIT in sCL Big TODO */
-/* init_ack     */ {sIV, sCL, sCW, sCE, sES, sSS, sSR, sSA, sIV, sHA},
+/* init_ack     */ {sIV, sCW, sCW, sCE, sES, sSS, sSR, sSA, sIV, sHA},
 /* abort        */ {sIV, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sIV, sCL},
 /* shutdown     */ {sIV, sCL, sCW, sCE, sSR, sSS, sSR, sSA, sIV, sSR},
 /* shutdown_ack */ {sIV, sCL, sCW, sCE, sES, sSA, sSA, sSA, sIV, sHA},
···
 			ct->proto.sctp.vtag[IP_CT_DIR_REPLY] = sh->vtag;
 		}
 
-		ct->proto.sctp.state = new_state;
+		ct->proto.sctp.state = SCTP_CONNTRACK_NONE;
 	}
 
 	return true;

net/netfilter/nf_tables_api.c

···
 static const struct nft_chain_type *chain_type[NFPROTO_NUMPROTO][NFT_CHAIN_T_MAX];
 
 static const struct nft_chain_type *
+__nft_chain_type_get(u8 family, enum nft_chain_types type)
+{
+	if (family >= NFPROTO_NUMPROTO ||
+	    type >= NFT_CHAIN_T_MAX)
+		return NULL;
+
+	return chain_type[family][type];
+}
+
+static const struct nft_chain_type *
 __nf_tables_chain_type_lookup(const struct nlattr *nla, u8 family)
 {
+	const struct nft_chain_type *type;
 	int i;
 
 	for (i = 0; i < NFT_CHAIN_T_MAX; i++) {
-		if (chain_type[family][i] != NULL &&
-		    !nla_strcmp(nla, chain_type[family][i]->name))
-			return chain_type[family][i];
+		type = __nft_chain_type_get(family, i);
+		if (!type)
+			continue;
+		if (!nla_strcmp(nla, type->name))
+			return type;
 	}
 	return NULL;
 }
 
-/*
- * Loading a module requires dropping mutex that guards the transaction.
- * A different client might race to start a new transaction meanwhile. Zap the
- * list of pending transaction and then restore it once the mutex is grabbed
- * again. Users of this function return EAGAIN which implicitly triggers the
- * transaction abort path to clean up the list of pending transactions.
- */
+struct nft_module_request {
+	struct list_head	list;
+	char			module[MODULE_NAME_LEN];
+	bool			done;
+};
+
 #ifdef CONFIG_MODULES
-static void nft_request_module(struct net *net, const char *fmt, ...)
+static int nft_request_module(struct net *net, const char *fmt, ...)
 {
 	char module_name[MODULE_NAME_LEN];
-	LIST_HEAD(commit_list);
+	struct nft_module_request *req;
 	va_list args;
 	int ret;
-
-	list_splice_init(&net->nft.commit_list, &commit_list);
 
 	va_start(args, fmt);
 	ret = vsnprintf(module_name, MODULE_NAME_LEN, fmt, args);
 	va_end(args);
 	if (ret >= MODULE_NAME_LEN)
-		return;
+		return 0;
 
-	mutex_unlock(&net->nft.commit_mutex);
-	request_module("%s", module_name);
-	mutex_lock(&net->nft.commit_mutex);
+	list_for_each_entry(req, &net->nft.module_list, list) {
+		if (!strcmp(req->module, module_name)) {
+			if (req->done)
+				return 0;
 
-	WARN_ON_ONCE(!list_empty(&net->nft.commit_list));
-	list_splice(&commit_list, &net->nft.commit_list);
+			/* A request to load this module already exists. */
+			return -EAGAIN;
+		}
+	}
+
+	req = kmalloc(sizeof(*req), GFP_KERNEL);
+	if (!req)
+		return -ENOMEM;
+
+	req->done = false;
+	strlcpy(req->module, module_name, MODULE_NAME_LEN);
+	list_add_tail(&req->list, &net->nft.module_list);
+
+	return -EAGAIN;
 }
 #endif
 
···
 	lockdep_nfnl_nft_mutex_not_held();
 #ifdef CONFIG_MODULES
 	if (autoload) {
-		nft_request_module(net, "nft-chain-%u-%.*s", family,
-				   nla_len(nla), (const char *)nla_data(nla));
-		type = __nf_tables_chain_type_lookup(nla, family);
-		if (type != NULL)
+		if (nft_request_module(net, "nft-chain-%u-%.*s", family,
+				       nla_len(nla),
+				       (const char *)nla_data(nla)) == -EAGAIN)
 			return ERR_PTR(-EAGAIN);
 	}
 #endif
···
 
 void nft_register_chain_type(const struct nft_chain_type *ctype)
 {
-	if (WARN_ON(ctype->family >= NFPROTO_NUMPROTO))
-		return;
-
 	nfnl_lock(NFNL_SUBSYS_NFTABLES);
-	if (WARN_ON(chain_type[ctype->family][ctype->type] != NULL)) {
+	if (WARN_ON(__nft_chain_type_get(ctype->family, ctype->type))) {
 		nfnl_unlock(NFNL_SUBSYS_NFTABLES);
 		return;
 	}
···
 	hook->num = ntohl(nla_get_be32(ha[NFTA_HOOK_HOOKNUM]));
 	hook->priority = ntohl(nla_get_be32(ha[NFTA_HOOK_PRIORITY]));
 
-	type = chain_type[family][NFT_CHAIN_T_DEFAULT];
+	type = __nft_chain_type_get(family, NFT_CHAIN_T_DEFAULT);
+	if (!type)
+		return -EOPNOTSUPP;
+
 	if (nla[NFTA_CHAIN_TYPE]) {
 		type = nf_tables_chain_type_lookup(net, nla[NFTA_CHAIN_TYPE],
 						   family, autoload);
···
 static int nft_expr_type_request_module(struct net *net, u8 family,
 					struct nlattr *nla)
 {
-	nft_request_module(net, "nft-expr-%u-%.*s", family,
-			   nla_len(nla), (char *)nla_data(nla));
-	if (__nft_expr_type_get(family, nla))
+	if (nft_request_module(net, "nft-expr-%u-%.*s", family,
+			       nla_len(nla), (char *)nla_data(nla)) == -EAGAIN)
 		return -EAGAIN;
 
 	return 0;
···
 		if (nft_expr_type_request_module(net, family, nla) == -EAGAIN)
 			return ERR_PTR(-EAGAIN);
 
-		nft_request_module(net, "nft-expr-%.*s",
-				   nla_len(nla), (char *)nla_data(nla));
-		if (__nft_expr_type_get(family, nla))
+		if (nft_request_module(net, "nft-expr-%.*s",
+				       nla_len(nla),
+				       (char *)nla_data(nla)) == -EAGAIN)
 			return ERR_PTR(-EAGAIN);
 	}
 #endif
···
 			err = PTR_ERR(ops);
 #ifdef CONFIG_MODULES
 			if (err == -EAGAIN)
-				nft_expr_type_request_module(ctx->net,
-							     ctx->family,
-							     tb[NFTA_EXPR_NAME]);
+				if (nft_expr_type_request_module(ctx->net,
+								 ctx->family,
+								 tb[NFTA_EXPR_NAME]) != -EAGAIN)
+					err = -ENOENT;
 #endif
 			goto err1;
 		}
···
 	lockdep_nfnl_nft_mutex_not_held();
 #ifdef CONFIG_MODULES
 	if (list_empty(&nf_tables_set_types)) {
-		nft_request_module(ctx->net, "nft-set");
-		if (!list_empty(&nf_tables_set_types))
+		if (nft_request_module(ctx->net, "nft-set") == -EAGAIN)
 			return ERR_PTR(-EAGAIN);
 	}
 #endif
···
 	lockdep_nfnl_nft_mutex_not_held();
 #ifdef CONFIG_MODULES
 	if (type == NULL) {
-		nft_request_module(net, "nft-obj-%u", objtype);
-		if (__nft_obj_type_get(objtype))
+		if (nft_request_module(net, "nft-obj-%u", objtype) == -EAGAIN)
 			return ERR_PTR(-EAGAIN);
 	}
 #endif
···
 	lockdep_nfnl_nft_mutex_not_held();
 #ifdef CONFIG_MODULES
 	if (type == NULL) {
-		nft_request_module(net, "nf-flowtable-%u", family);
-		if (__nft_flowtable_type_get(family))
+		if (nft_request_module(net, "nf-flowtable-%u", family) == -EAGAIN)
 			return ERR_PTR(-EAGAIN);
 	}
 #endif
···
 	list_del_rcu(&chain->list);
 }
 
+static void nf_tables_module_autoload_cleanup(struct net *net)
+{
+	struct nft_module_request *req, *next;
+
+	WARN_ON_ONCE(!list_empty(&net->nft.commit_list));
+	list_for_each_entry_safe(req, next, &net->nft.module_list, list) {
+		WARN_ON_ONCE(!req->done);
+		list_del(&req->list);
+		kfree(req);
+	}
+}
+
 static void nf_tables_commit_release(struct net *net)
 {
 	struct nft_trans *trans;
···
 	 * to prevent expensive synchronize_rcu() in commit phase.
 	 */
 	if (list_empty(&net->nft.commit_list)) {
+		nf_tables_module_autoload_cleanup(net);
 		mutex_unlock(&net->nft.commit_mutex);
 		return;
 	}
···
 	list_splice_tail_init(&net->nft.commit_list, &nf_tables_destroy_list);
 	spin_unlock(&nf_tables_destroy_list_lock);
 
+	nf_tables_module_autoload_cleanup(net);
 	mutex_unlock(&net->nft.commit_mutex);
 
 	schedule_work(&trans_destroy_work);
···
 	return 0;
 }
 
+static void nf_tables_module_autoload(struct net *net)
+{
+	struct nft_module_request *req, *next;
+	LIST_HEAD(module_list);
+
+	list_splice_init(&net->nft.module_list, &module_list);
+	mutex_unlock(&net->nft.commit_mutex);
+	list_for_each_entry_safe(req, next, &module_list, list) {
+		if (req->done) {
+			list_del(&req->list);
+			kfree(req);
+		} else {
+			request_module("%s", req->module);
+			req->done = true;
+		}
+	}
+	mutex_lock(&net->nft.commit_mutex);
+	list_splice(&module_list, &net->nft.module_list);
+}
+
 static void nf_tables_abort_release(struct nft_trans *trans)
 {
 	switch (trans->msg_type) {
···
 	kfree(trans);
 }
 
-static int __nf_tables_abort(struct net *net)
+static int __nf_tables_abort(struct net *net, bool autoload)
 {
 	struct nft_trans *trans, *next;
 	struct nft_trans_elem *te;
···
 		nf_tables_abort_release(trans);
 	}
 
+	if (autoload)
+		nf_tables_module_autoload(net);
+	else
+		nf_tables_module_autoload_cleanup(net);
+
 	return 0;
 }
 
···
 	nft_validate_state_update(net, NFT_VALIDATE_SKIP);
 }
 
-static int nf_tables_abort(struct net *net, struct sk_buff *skb)
+static int nf_tables_abort(struct net *net, struct sk_buff *skb, bool autoload)
 {
-	int ret = __nf_tables_abort(net);
+	int ret = __nf_tables_abort(net, autoload);
 
 	mutex_unlock(&net->nft.commit_mutex);
 
···
 {
 	INIT_LIST_HEAD(&net->nft.tables);
 	INIT_LIST_HEAD(&net->nft.commit_list);
+	INIT_LIST_HEAD(&net->nft.module_list);
 	mutex_init(&net->nft.commit_mutex);
 	net->nft.base_seq = 1;
 	net->nft.validate_state = NFT_VALIDATE_SKIP;
···
 {
 	mutex_lock(&net->nft.commit_mutex);
 	if (!list_empty(&net->nft.commit_list))
-		__nf_tables_abort(net);
+		__nf_tables_abort(net, false);
 	__nft_release_tables(net);
 	mutex_unlock(&net->nft.commit_mutex);
 	WARN_ON_ONCE(!list_empty(&net->nft.tables));

net/netfilter/nf_tables_offload.c

···
 
 	mutex_lock(&net->nft.commit_mutex);
 	chain = __nft_offload_get_chain(dev);
-	if (chain) {
+	if (chain && chain->flags & NFT_CHAIN_HW_OFFLOAD) {
 		struct nft_base_chain *basechain;
 
 		basechain = nft_base_chain(chain);

net/netfilter/nfnetlink.c

···
 	}
 done:
 	if (status & NFNL_BATCH_REPLAY) {
-		ss->abort(net, oskb);
+		ss->abort(net, oskb, true);
 		nfnl_err_reset(&err_list);
 		kfree_skb(skb);
 		module_put(ss->owner);
···
 			status |= NFNL_BATCH_REPLAY;
 			goto done;
 		} else if (err) {
-			ss->abort(net, oskb);
+			ss->abort(net, oskb, false);
 			netlink_ack(oskb, nlmsg_hdr(oskb), err, NULL);
 		}
 	} else {
-		ss->abort(net, oskb);
+		ss->abort(net, oskb, false);
 	}
 	if (ss->cleanup)
 		ss->cleanup(net);

net/netfilter/nft_osf.c

···
 	int err;
 	u8 ttl;
 
+	if (!tb[NFTA_OSF_DREG])
+		return -EINVAL;
+
 	if (tb[NFTA_OSF_TTL]) {
 		ttl = nla_get_u8(tb[NFTA_OSF_TTL]);
 		if (ttl > 2)