Skip to contentSkip to conversation

Netfilter updates for net-next

PATCH New Pablo Neira Ayuso 1 participant 14 messages Jun 9, 2021

Subsystems (3)
lore
Share & tools
lore

Pablo Neira Ayuso

author maintainer
#0 lore raw

cover Netfilter updates for net-next

Hi,

The following patchset contains Netfilter updates for net-next:

1) Add nfgenmsg field to nfnetlink's struct nfnl_info and use it.

2) Remove nft_ctx_init_from_elemattr() and nft_ctx_init_from_setattr() helper functions.

3) Add the nf_ct_pernet() helper function to fetch the conntrack pernetns data area.

4) Expose TCP and UDP flowtable offload timeouts through sysctl, from Oz Shlomo.

5) Add nfnetlink_hook subsystem to fetch the netfilter hook pipeline configuration, from Florian Westphal. This also includes a new field to annotate the hook type as metadata.

6) Fix unsafe memory access to non-linear skbuff in the new SCTP chunk support for nft_exthdr, from Phil Sutter.

Please, pull these changes from:

git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next.git

Thank you!

----------------------------------------------------------------

The following changes since commit 1a42624aecba438f1d114430a14b640cdfa51c87:

net: dsa: xrs700x: allow HSR/PRP supervision dupes for node_table (2021-06-04 14:49:28 -0700)

are available in the Git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next.git HEAD

for you to fetch changes up to c5c6accd7b7e10434d6afda4f6a5107c480bb4fb:

netfilter: nf_tables: move base hook annotation to init helper (2021-06-09 21:29:23 +0200)

----------------------------------------------------------------
Colin Ian King (1):
netfilter: nfnetlink_hook: fix array index out-of-bounds error

Florian Westphal (4):
netfilter: annotate nf_tables base hook ops
netfilter: add new hook nfnl subsystem
netfilter: nfnetlink_hook: add depends-on nftables
netfilter: nf_tables: move base hook annotation to init helper

Oz Shlomo (3):
netfilter: conntrack: Introduce tcp offload timeout configuration
netfilter: conntrack: Introduce udp offload timeout configuration
netfilter: flowtable: Set offload timeouts according to proto values

Pablo Neira Ayuso (4):
netfilter: nfnetlink: add struct nfgenmsg to struct nfnl_info and use it
netfilter: nf_tables: remove nft_ctx_init_from_elemattr()
netfilter: nf_tables: remove nft_ctx_init_from_setattr()
netfilter: nftables: add nf_ct_pernet() helper function

Phil Sutter (1):
netfilter: nft_exthdr: Fix for unsafe packet data read

 include/linux/netfilter.h                     |   8 +-
 include/linux/netfilter/nfnetlink.h           |   1 +
 include/net/netfilter/nf_conntrack.h          |   7 +
 include/net/netfilter/nf_flow_table.h         |   2 +
 include/net/netns/conntrack.h                 |   8 +
 include/uapi/linux/netfilter/nfnetlink.h      |   3 +-
 include/uapi/linux/netfilter/nfnetlink_hook.h |  55 ++++
 net/netfilter/Kconfig                         |  10 +
 net/netfilter/Makefile                        |   1 +
 net/netfilter/nf_conntrack_core.c             |  22 +-
 net/netfilter/nf_conntrack_ecache.c           |   8 +-
 net/netfilter/nf_conntrack_expect.c           |  12 +-
 net/netfilter/nf_conntrack_helper.c           |   6 +-
 net/netfilter/nf_conntrack_netlink.c          |  23 +-
 net/netfilter/nf_conntrack_proto.c            |   6 +-
 net/netfilter/nf_conntrack_proto_tcp.c        |   5 +
 net/netfilter/nf_conntrack_proto_udp.c        |   5 +
 net/netfilter/nf_conntrack_standalone.c       |  54 +++-
 net/netfilter/nf_flow_table_core.c            |  47 +++-
 net/netfilter/nf_flow_table_offload.c         |   4 +-
 net/netfilter/nf_tables_api.c                 | 202 ++++++--------
 net/netfilter/nfnetlink.c                     |   3 +
 net/netfilter/nfnetlink_hook.c                | 375 ++++++++++++++++++++++++++
 net/netfilter/nfnetlink_log.c                 |   5 +-
 net/netfilter/nfnetlink_queue.c               |   9 +-
 net/netfilter/nft_compat.c                    |  17 +-
 net/netfilter/nft_exthdr.c                    |   4 +-
 27 files changed, 697 insertions(+), 205 deletions(-)
 create mode 100644 include/uapi/linux/netfilter/nfnetlink_hook.h
 create mode 100644 net/netfilter/nfnetlink_hook.c
author maintainer
#1 lore raw

PATCH 1/13 netfilter: nfnetlink: add struct nfgenmsg to struct nfnl_info and use it

Update the nfnl_info structure to add a pointer to the nfnetlink header. This simplifies the existing codebase since this header is usually accessed. Update existing clients to use this new field.

---
 include/linux/netfilter/nfnetlink.h  |  1 +
 net/netfilter/nf_conntrack_netlink.c | 23 +++++-------
 net/netfilter/nf_tables_api.c        | 55 ++++++++++------------------
 net/netfilter/nfnetlink.c            |  2 +
 net/netfilter/nfnetlink_log.c        |  5 +--
 net/netfilter/nfnetlink_queue.c      |  9 ++---
 net/netfilter/nft_compat.c           | 17 +++------
 7 files changed, 42 insertions(+), 70 deletions(-)
include/linux/netfilter/nfnetlink.h +1 -0
index 515ce53aa20d..241e005f290a 100644
@@ -11,6 +11,7 @@ struct nfnl_info {
 	struct net		*net;
 	struct sock		*sk;
 	const struct nlmsghdr	*nlh;
+	const struct nfgenmsg	*nfmsg;
 	struct netlink_ext_ack	*extack;
 };
 
net/netfilter/nf_conntrack_netlink.c +9 -14
index 220f51f055ab..4e1a9dba7077 100644
@@ -1528,7 +1528,7 @@ static int ctnetlink_del_conntrack(struct sk_buff *skb,
 				   const struct nfnl_info *info,
 				   const struct nlattr * const cda[])
 {
-	struct nfgenmsg *nfmsg = nlmsg_data(info->nlh);
+	u8 family = info->nfmsg->nfgen_family;
 	struct nf_conntrack_tuple_hash *h;
 	struct nf_conntrack_tuple tuple;
 	struct nf_conntrack_zone zone;
@@ -1541,12 +1541,12 @@ static int ctnetlink_del_conntrack(struct sk_buff *skb,
 
 	if (cda[CTA_TUPLE_ORIG])
 		err = ctnetlink_parse_tuple(cda, &tuple, CTA_TUPLE_ORIG,
-					    nfmsg->nfgen_family, &zone);
+					    family, &zone);
 	else if (cda[CTA_TUPLE_REPLY])
 		err = ctnetlink_parse_tuple(cda, &tuple, CTA_TUPLE_REPLY,
-					    nfmsg->nfgen_family, &zone);
+					    family, &zone);
 	else {
-		u_int8_t u3 = nfmsg->version ? nfmsg->nfgen_family : AF_UNSPEC;
+		u_int8_t u3 = info->nfmsg->version ? family : AF_UNSPEC;
 
 		return ctnetlink_flush_conntrack(info->net, cda,
 						 NETLINK_CB(skb).portid,
@@ -1586,8 +1586,7 @@ static int ctnetlink_get_conntrack(struct sk_buff *skb,
 				   const struct nfnl_info *info,
 				   const struct nlattr * const cda[])
 {
-	struct nfgenmsg *nfmsg = nlmsg_data(info->nlh);
-	u_int8_t u3 = nfmsg->nfgen_family;
+	u_int8_t u3 = info->nfmsg->nfgen_family;
 	struct nf_conntrack_tuple_hash *h;
 	struct nf_conntrack_tuple tuple;
 	struct nf_conntrack_zone zone;
@@ -2363,10 +2362,9 @@ static int ctnetlink_new_conntrack(struct sk_buff *skb,
 				   const struct nfnl_info *info,
 				   const struct nlattr * const cda[])
 {
-	struct nfgenmsg *nfmsg = nlmsg_data(info->nlh);
 	struct nf_conntrack_tuple otuple, rtuple;
 	struct nf_conntrack_tuple_hash *h = NULL;
-	u_int8_t u3 = nfmsg->nfgen_family;
+	u_int8_t u3 = info->nfmsg->nfgen_family;
 	struct nf_conntrack_zone zone;
 	struct nf_conn *ct;
 	int err;
@@ -3259,8 +3257,7 @@ static int ctnetlink_get_expect(struct sk_buff *skb,
 				const struct nfnl_info *info,
 				const struct nlattr * const cda[])
 {
-	struct nfgenmsg *nfmsg = nlmsg_data(info->nlh);
-	u_int8_t u3 = nfmsg->nfgen_family;
+	u_int8_t u3 = info->nfmsg->nfgen_family;
 	struct nf_conntrack_tuple tuple;
 	struct nf_conntrack_expect *exp;
 	struct nf_conntrack_zone zone;
@@ -3349,8 +3346,7 @@ static int ctnetlink_del_expect(struct sk_buff *skb,
 				const struct nfnl_info *info,
 				const struct nlattr * const cda[])
 {
-	struct nfgenmsg *nfmsg = nlmsg_data(info->nlh);
-	u_int8_t u3 = nfmsg->nfgen_family;
+	u_int8_t u3 = info->nfmsg->nfgen_family;
 	struct nf_conntrack_expect *exp;
 	struct nf_conntrack_tuple tuple;
 	struct nf_conntrack_zone zone;
@@ -3601,8 +3597,7 @@ static int ctnetlink_new_expect(struct sk_buff *skb,
 				const struct nfnl_info *info,
 				const struct nlattr * const cda[])
 {
-	struct nfgenmsg *nfmsg = nlmsg_data(info->nlh);
-	u_int8_t u3 = nfmsg->nfgen_family;
+	u_int8_t u3 = info->nfmsg->nfgen_family;
 	struct nf_conntrack_tuple tuple;
 	struct nf_conntrack_expect *exp;
 	struct nf_conntrack_zone zone;
net/netfilter/nf_tables_api.c +19 -36

218 lines not on this page · raw message

net/netfilter/nfnetlink.c +2 -0
index e8dbd8379027..028a1f39318b 100644
@@ -256,6 +256,7 @@ static int nfnetlink_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh,
 			.net	= net,
 			.sk	= nfnlnet->nfnl,
 			.nlh	= nlh,
+			.nfmsg	= nlmsg_data(nlh),
 			.extack	= extack,
 		};
 
@@ -491,6 +492,7 @@ static void nfnetlink_rcv_batch(struct sk_buff *skb, struct nlmsghdr *nlh,
 				.net	= net,
 				.sk	= nfnlnet->nfnl,
 				.nlh	= nlh,
+				.nfmsg	= nlmsg_data(nlh),
 				.extack	= &extack,
 			};
 
net/netfilter/nfnetlink_log.c +2 -3
index 587086b18c36..691ef4cffdd9 100644
@@ -871,15 +871,14 @@ static int nfulnl_recv_config(struct sk_buff *skb, const struct nfnl_info *info,
 			      const struct nlattr * const nfula[])
 {
 	struct nfnl_log_net *log = nfnl_log_pernet(info->net);
-	struct nfgenmsg *nfmsg = nlmsg_data(info->nlh);
-	u_int16_t group_num = ntohs(nfmsg->res_id);
+	u_int16_t group_num = ntohs(info->nfmsg->res_id);
 	struct nfulnl_msg_config_cmd *cmd = NULL;
 	struct nfulnl_instance *inst;
 	u16 flags = 0;
 	int ret = 0;
 
 	if (nfula[NFULA_CFG_CMD]) {
-		u_int8_t pf = nfmsg->nfgen_family;
+		u_int8_t pf = info->nfmsg->nfgen_family;
 		cmd = nla_data(nfula[NFULA_CFG_CMD]);
 
 		/* Commands without queue context */
net/netfilter/nfnetlink_queue.c +3 -6
index f37a575ebd7f..f774de0fc24f 100644
@@ -1051,8 +1051,7 @@ static int nfqnl_recv_verdict_batch(struct sk_buff *skb,
 				    const struct nlattr * const nfqa[])
 {
 	struct nfnl_queue_net *q = nfnl_queue_pernet(info->net);
-	struct nfgenmsg *nfmsg = nlmsg_data(info->nlh);
-	u16 queue_num = ntohs(nfmsg->res_id);
+	u16 queue_num = ntohs(info->nfmsg->res_id);
 	struct nf_queue_entry *entry, *tmp;
 	struct nfqnl_msg_verdict_hdr *vhdr;
 	struct nfqnl_instance *queue;
@@ -1160,8 +1159,7 @@ static int nfqnl_recv_verdict(struct sk_buff *skb, const struct nfnl_info *info,
 			      const struct nlattr * const nfqa[])
 {
 	struct nfnl_queue_net *q = nfnl_queue_pernet(info->net);
-	struct nfgenmsg *nfmsg = nlmsg_data(info->nlh);
-	u_int16_t queue_num = ntohs(nfmsg->res_id);
+	u_int16_t queue_num = ntohs(info->nfmsg->res_id);
 	struct nfqnl_msg_verdict_hdr *vhdr;
 	enum ip_conntrack_info ctinfo;
 	struct nfqnl_instance *queue;
@@ -1243,8 +1241,7 @@ static int nfqnl_recv_config(struct sk_buff *skb, const struct nfnl_info *info,
 			     const struct nlattr * const nfqa[])
 {
 	struct nfnl_queue_net *q = nfnl_queue_pernet(info->net);
-	struct nfgenmsg *nfmsg = nlmsg_data(info->nlh);
-	u_int16_t queue_num = ntohs(nfmsg->res_id);
+	u_int16_t queue_num = ntohs(info->nfmsg->res_id);
 	struct nfqnl_msg_config_cmd *cmd = NULL;
 	struct nfqnl_instance *queue;
 	__u32 flags = 0, mask = 0;
net/netfilter/nft_compat.c +6 -11
index 3144a9ad2f6a..639c337c885b 100644
@@ -625,7 +625,7 @@ static int nfnl_compat_get_rcu(struct sk_buff *skb,
 			       const struct nfnl_info *info,
 			       const struct nlattr * const tb[])
 {
-	struct nfgenmsg *nfmsg;
+	u8 family = info->nfmsg->nfgen_family;
 	const char *name, *fmt;
 	struct sk_buff *skb2;
 	int ret = 0, target;
@@ -640,9 +640,7 @@ static int nfnl_compat_get_rcu(struct sk_buff *skb,
 	rev = ntohl(nla_get_be32(tb[NFTA_COMPAT_REV]));
 	target = ntohl(nla_get_be32(tb[NFTA_COMPAT_TYPE]));
 
-	nfmsg = nlmsg_data(info->nlh);
-
-	switch(nfmsg->nfgen_family) {
+	switch(family) {
 	case AF_INET:
 		fmt = "ipt_%s";
 		break;
@@ -656,8 +654,7 @@ static int nfnl_compat_get_rcu(struct sk_buff *skb,
 		fmt = "arpt_%s";
 		break;
 	default:
-		pr_err("nft_compat: unsupported protocol %d\n",
-			nfmsg->nfgen_family);
+		pr_err("nft_compat: unsupported protocol %d\n", family);
 		return -EINVAL;
 	}
 
@@ -665,9 +662,8 @@ static int nfnl_compat_get_rcu(struct sk_buff *skb,
 		return -EINVAL;
 
 	rcu_read_unlock();
-	try_then_request_module(xt_find_revision(nfmsg->nfgen_family, name,
-						 rev, target, &ret),
-						 fmt, name);
+	try_then_request_module(xt_find_revision(family, name, rev, target, &ret),
+				fmt, name);
 	if (ret < 0)
 		goto out_put;
 
@@ -682,8 +678,7 @@ static int nfnl_compat_get_rcu(struct sk_buff *skb,
 				  info->nlh->nlmsg_seq,
 				  NFNL_MSG_TYPE(info->nlh->nlmsg_type),
 				  NFNL_MSG_COMPAT_GET,
-				  nfmsg->nfgen_family,
-				  name, ret, target) <= 0) {
+				  family, name, ret, target) <= 0) {
 		kfree_skb(skb2);
 		goto out_put;
 	}
Signature · 1 line
2.30.2
author maintainer
#2 lore raw

PATCH 2/13 netfilter: nf_tables: remove nft_ctx_init_from_elemattr()

In reply to Pablo Neira Ayuso

Replace nft_ctx_init_from_elemattr() by nft_table_lookup() and set up the context structure right before it is really needed.

Moreover, nft_ctx_init_from_elemattr() is setting up the context structure for codepaths where this is not really needed at all.

This helper function is also not helping to consolidate code, removing it saves us 4 LoC.

---
 net/netfilter/nf_tables_api.c | 72 +++++++++++++++++------------------
 1 file changed, 34 insertions(+), 38 deletions(-)
net/netfilter/nf_tables_api.c +34 -38
index b2b4e03ce036..2fbcb2543795 100644
@@ -4703,28 +4703,6 @@ static const struct nla_policy nft_set_elem_list_policy[NFTA_SET_ELEM_LIST_MAX +
 	[NFTA_SET_ELEM_LIST_SET_ID]	= { .type = NLA_U32 },
 };
 
-static int nft_ctx_init_from_elemattr(struct nft_ctx *ctx, struct net *net,
-				      const struct sk_buff *skb,
-				      const struct nlmsghdr *nlh,
-				      const struct nlattr * const nla[],
-				      struct netlink_ext_ack *extack,
-				      u8 genmask, u32 nlpid)
-{
-	const struct nfgenmsg *nfmsg = nlmsg_data(nlh);
-	int family = nfmsg->nfgen_family;
-	struct nft_table *table;
-
-	table = nft_table_lookup(net, nla[NFTA_SET_ELEM_LIST_TABLE], family,
-				 genmask, nlpid);
-	if (IS_ERR(table)) {
-		NL_SET_BAD_ATTR(extack, nla[NFTA_SET_ELEM_LIST_TABLE]);
-		return PTR_ERR(table);
-	}
-
-	nft_ctx_init(ctx, net, skb, nlh, family, table, NULL, nla);
-	return 0;
-}
-
 static int nft_set_elem_expr_dump(struct sk_buff *skb,
 				  const struct nft_set *set,
 				  const struct nft_set_ext *ext)
@@ -5182,21 +5160,27 @@ static int nf_tables_getsetelem(struct sk_buff *skb,
 {
 	struct netlink_ext_ack *extack = info->extack;
 	u8 genmask = nft_genmask_cur(info->net);
+	u8 family = info->nfmsg->nfgen_family;
 	struct net *net = info->net;
+	struct nft_table *table;
 	struct nft_set *set;
 	struct nlattr *attr;
 	struct nft_ctx ctx;
 	int rem, err = 0;
 
-	err = nft_ctx_init_from_elemattr(&ctx, net, skb, info->nlh, nla, extack,
-					 genmask, NETLINK_CB(skb).portid);
-	if (err < 0)
-		return err;
+	table = nft_table_lookup(net, nla[NFTA_SET_ELEM_LIST_TABLE], family,
+				 genmask, NETLINK_CB(skb).portid);
+	if (IS_ERR(table)) {
+		NL_SET_BAD_ATTR(extack, nla[NFTA_SET_ELEM_LIST_TABLE]);
+		return PTR_ERR(table);
+	}
 
-	set = nft_set_lookup(ctx.table, nla[NFTA_SET_ELEM_LIST_SET], genmask);
+	set = nft_set_lookup(table, nla[NFTA_SET_ELEM_LIST_SET], genmask);
 	if (IS_ERR(set))
 		return PTR_ERR(set);
 
+	nft_ctx_init(&ctx, net, skb, info->nlh, family, table, NULL, nla);
+
 	if (info->nlh->nlmsg_flags & NLM_F_DUMP) {
 		struct netlink_dump_control c = {
 			.start = nf_tables_dump_set_start,
@@ -5965,8 +5949,10 @@ static int nf_tables_newsetelem(struct sk_buff *skb,
 	struct nftables_pernet *nft_net = nft_pernet(info->net);
 	struct netlink_ext_ack *extack = info->extack;
 	u8 genmask = nft_genmask_next(info->net);
+	u8 family = info->nfmsg->nfgen_family;
 	struct net *net = info->net;
 	const struct nlattr *attr;
+	struct nft_table *table;
 	struct nft_set *set;
 	struct nft_ctx ctx;
 	int rem, err;
@@ -5974,12 +5960,14 @@ static int nf_tables_newsetelem(struct sk_buff *skb,
 	if (nla[NFTA_SET_ELEM_LIST_ELEMENTS] == NULL)
 		return -EINVAL;
 
-	err = nft_ctx_init_from_elemattr(&ctx, net, skb, info->nlh, nla, extack,
-					 genmask, NETLINK_CB(skb).portid);
-	if (err < 0)
-		return err;
+	table = nft_table_lookup(net, nla[NFTA_SET_ELEM_LIST_TABLE], family,
+				 genmask, NETLINK_CB(skb).portid);
+	if (IS_ERR(table)) {
+		NL_SET_BAD_ATTR(extack, nla[NFTA_SET_ELEM_LIST_TABLE]);
+		return PTR_ERR(table);
+	}
 
-	set = nft_set_lookup_global(net, ctx.table, nla[NFTA_SET_ELEM_LIST_SET],
+	set = nft_set_lookup_global(net, table, nla[NFTA_SET_ELEM_LIST_SET],
 				    nla[NFTA_SET_ELEM_LIST_SET_ID], genmask);
 	if (IS_ERR(set))
 		return PTR_ERR(set);
@@ -5987,6 +5975,8 @@ static int nf_tables_newsetelem(struct sk_buff *skb,
 	if (!list_empty(&set->bindings) && set->flags & NFT_SET_CONSTANT)
 		return -EBUSY;
 
+	nft_ctx_init(&ctx, net, skb, info->nlh, family, table, NULL, nla);
+
 	nla_for_each_nested(attr, nla[NFTA_SET_ELEM_LIST_ELEMENTS], rem) {
 		err = nft_add_set_elem(&ctx, set, attr, info->nlh->nlmsg_flags);
 		if (err < 0)
@@ -5994,7 +5984,7 @@ static int nf_tables_newsetelem(struct sk_buff *skb,
 	}
 
 	if (nft_net->validate_state == NFT_VALIDATE_DO)
-		return nft_table_validate(net, ctx.table);
+		return nft_table_validate(net, table);
 
 	return 0;
 }
@@ -6232,23 +6222,29 @@ static int nf_tables_delsetelem(struct sk_buff *skb,
 {
 	struct netlink_ext_ack *extack = info->extack;
 	u8 genmask = nft_genmask_next(info->net);
+	u8 family = info->nfmsg->nfgen_family;
 	struct net *net = info->net;
 	const struct nlattr *attr;
+	struct nft_table *table;
 	struct nft_set *set;
 	struct nft_ctx ctx;
 	int rem, err = 0;
 
-	err = nft_ctx_init_from_elemattr(&ctx, net, skb, info->nlh, nla, extack,
-					 genmask, NETLINK_CB(skb).portid);
-	if (err < 0)
-		return err;
+	table = nft_table_lookup(net, nla[NFTA_SET_ELEM_LIST_TABLE], family,
+				 genmask, NETLINK_CB(skb).portid);
+	if (IS_ERR(table)) {
+		NL_SET_BAD_ATTR(extack, nla[NFTA_SET_ELEM_LIST_TABLE]);
+		return PTR_ERR(table);
+	}
 
-	set = nft_set_lookup(ctx.table, nla[NFTA_SET_ELEM_LIST_SET], genmask);
+	set = nft_set_lookup(table, nla[NFTA_SET_ELEM_LIST_SET], genmask);
 	if (IS_ERR(set))
 		return PTR_ERR(set);
 	if (!list_empty(&set->bindings) && set->flags & NFT_SET_CONSTANT)
 		return -EBUSY;
 
+	nft_ctx_init(&ctx, net, skb, info->nlh, family, table, NULL, nla);
+
 	if (!nla[NFTA_SET_ELEM_LIST_ELEMENTS])
 		return nft_set_flush(&ctx, set, genmask);
 
Signature · 1 line
2.30.2
author maintainer
#3 lore raw

PATCH 3/13 netfilter: nf_tables: remove nft_ctx_init_from_setattr()

In reply to Pablo Neira Ayuso

Replace nft_ctx_init_from_setattr() by nft_table_lookup().

This patch also disentangles nf_tables_delset() where NFTA_SET_TABLE is required while nft_ctx_init_from_setattr() allows it to be optional.

From the nf_tables_delset() path, this also allows to set up the context structure when it is needed.

Removing this helper function saves us 14 LoC, so it is not helping to consolidate code.

---
 net/netfilter/nf_tables_api.c | 64 ++++++++++++++---------------------
 1 file changed, 25 insertions(+), 39 deletions(-)
net/netfilter/nf_tables_api.c +25 -39
index 2fbcb2543795..6c2000a11c7e 100644
@@ -3639,30 +3639,6 @@ static const struct nla_policy nft_set_desc_policy[NFTA_SET_DESC_MAX + 1] = {
 	[NFTA_SET_DESC_CONCAT]		= { .type = NLA_NESTED },
 };
 
-static int nft_ctx_init_from_setattr(struct nft_ctx *ctx, struct net *net,
-				     const struct sk_buff *skb,
-				     const struct nlmsghdr *nlh,
-				     const struct nlattr * const nla[],
-				     struct netlink_ext_ack *extack,
-				     u8 genmask, u32 nlpid)
-{
-	const struct nfgenmsg *nfmsg = nlmsg_data(nlh);
-	int family = nfmsg->nfgen_family;
-	struct nft_table *table = NULL;
-
-	if (nla[NFTA_SET_TABLE] != NULL) {
-		table = nft_table_lookup(net, nla[NFTA_SET_TABLE], family,
-					 genmask, nlpid);
-		if (IS_ERR(table)) {
-			NL_SET_BAD_ATTR(extack, nla[NFTA_SET_TABLE]);
-			return PTR_ERR(table);
-		}
-	}
-
-	nft_ctx_init(ctx, net, skb, nlh, family, table, NULL, nla);
-	return 0;
-}
-
 static struct nft_set *nft_set_lookup(const struct nft_table *table,
 				      const struct nlattr *nla, u8 genmask)
 {
@@ -4044,17 +4020,24 @@ static int nf_tables_getset(struct sk_buff *skb, const struct nfnl_info *info,
 {
 	struct netlink_ext_ack *extack = info->extack;
 	u8 genmask = nft_genmask_cur(info->net);
+	u8 family = info->nfmsg->nfgen_family;
+	struct nft_table *table = NULL;
 	struct net *net = info->net;
 	const struct nft_set *set;
 	struct sk_buff *skb2;
 	struct nft_ctx ctx;
 	int err;
 
-	/* Verify existence before starting dump */
-	err = nft_ctx_init_from_setattr(&ctx, net, skb, info->nlh, nla, extack,
-					genmask, 0);
-	if (err < 0)
-		return err;
+	if (nla[NFTA_SET_TABLE]) {
+		table = nft_table_lookup(net, nla[NFTA_SET_TABLE], family,
+					 genmask, 0);
+		if (IS_ERR(table)) {
+			NL_SET_BAD_ATTR(extack, nla[NFTA_SET_TABLE]);
+			return PTR_ERR(table);
+		}
+	}
+
+	nft_ctx_init(&ctx, net, skb, info->nlh, family, table, NULL, nla);
 
 	if (info->nlh->nlmsg_flags & NLM_F_DUMP) {
 		struct netlink_dump_control c = {
@@ -4074,7 +4057,7 @@ static int nf_tables_getset(struct sk_buff *skb, const struct nfnl_info *info,
 	if (!nla[NFTA_SET_TABLE])
 		return -EINVAL;
 
-	set = nft_set_lookup(ctx.table, nla[NFTA_SET_NAME], genmask);
+	set = nft_set_lookup(table, nla[NFTA_SET_NAME], genmask);
 	if (IS_ERR(set))
 		return PTR_ERR(set);
 
@@ -4467,28 +4450,29 @@ static int nf_tables_delset(struct sk_buff *skb, const struct nfnl_info *info,
 {
 	struct netlink_ext_ack *extack = info->extack;
 	u8 genmask = nft_genmask_next(info->net);
+	u8 family = info->nfmsg->nfgen_family;
 	struct net *net = info->net;
 	const struct nlattr *attr;
+	struct nft_table *table;
 	struct nft_set *set;
 	struct nft_ctx ctx;
-	int err;
 
 	if (info->nfmsg->nfgen_family == NFPROTO_UNSPEC)
 		return -EAFNOSUPPORT;
-	if (nla[NFTA_SET_TABLE] == NULL)
-		return -EINVAL;
 
-	err = nft_ctx_init_from_setattr(&ctx, net, skb, info->nlh, nla, extack,
-					genmask, NETLINK_CB(skb).portid);
-	if (err < 0)
-		return err;
+	table = nft_table_lookup(net, nla[NFTA_SET_TABLE], family,
+				 genmask, NETLINK_CB(skb).portid);
+	if (IS_ERR(table)) {
+		NL_SET_BAD_ATTR(extack, nla[NFTA_SET_TABLE]);
+		return PTR_ERR(table);
+	}
 
 	if (nla[NFTA_SET_HANDLE]) {
 		attr = nla[NFTA_SET_HANDLE];
-		set = nft_set_lookup_byhandle(ctx.table, attr, genmask);
+		set = nft_set_lookup_byhandle(table, attr, genmask);
 	} else {
 		attr = nla[NFTA_SET_NAME];
-		set = nft_set_lookup(ctx.table, attr, genmask);
+		set = nft_set_lookup(table, attr, genmask);
 	}
 
 	if (IS_ERR(set)) {
@@ -4502,6 +4486,8 @@ static int nf_tables_delset(struct sk_buff *skb, const struct nfnl_info *info,
 		return -EBUSY;
 	}
 
+	nft_ctx_init(&ctx, net, skb, info->nlh, family, table, NULL, nla);
+
 	return nft_delset(&ctx, set);
 }
 
Signature · 1 line
2.30.2
author maintainer
#4 lore raw

PATCH 4/13 netfilter: nftables: add nf_ct_pernet() helper function

In reply to Pablo Neira Ayuso

Consolidate call to net_generic(net, nf_conntrack_net_id) in this wrapper function.

---
 include/net/netfilter/nf_conntrack.h    |  7 +++++++
 net/netfilter/nf_conntrack_core.c       | 22 +++++++++-------------
 net/netfilter/nf_conntrack_ecache.c     |  8 +++-----
 net/netfilter/nf_conntrack_expect.c     | 12 +++++-------
 net/netfilter/nf_conntrack_helper.c     |  6 ++----
 net/netfilter/nf_conntrack_proto.c      |  6 ++----
 net/netfilter/nf_conntrack_standalone.c |  8 +++-----
 7 files changed, 31 insertions(+), 38 deletions(-)
include/net/netfilter/nf_conntrack.h +7 -0
index 06dc6db70d18..cc663c68ddc4 100644
@@ -346,6 +346,13 @@ nf_ct_set(struct sk_buff *skb, struct nf_conn *ct, enum ip_conntrack_info info)
 	skb_set_nfct(skb, (unsigned long)ct | info);
 }
 
+extern unsigned int nf_conntrack_net_id;
+
+static inline struct nf_conntrack_net *nf_ct_pernet(const struct net *net)
+{
+	return net_generic(net, nf_conntrack_net_id);
+}
+
 #define NF_CT_STAT_INC(net, count)	  __this_cpu_inc((net)->ct.stat->count)
 #define NF_CT_STAT_INC_ATOMIC(net, count) this_cpu_inc((net)->ct.stat->count)
 #define NF_CT_STAT_ADD_ATOMIC(net, count, v) this_cpu_add((net)->ct.stat->count, (v))
net/netfilter/nf_conntrack_core.c +9 -13
index e0befcf8113a..96ba19fc8155 100644
@@ -55,8 +55,6 @@
 
 #include "nf_internals.h"
 
-extern unsigned int nf_conntrack_net_id;
-
 __cacheline_aligned_in_smp spinlock_t nf_conntrack_locks[CONNTRACK_LOCKS];
 EXPORT_SYMBOL_GPL(nf_conntrack_locks);
 
@@ -87,8 +85,6 @@ static __read_mostly bool nf_conntrack_locks_all;
 
 static struct conntrack_gc_work conntrack_gc_work;
 
-extern unsigned int nf_conntrack_net_id;
-
 void nf_conntrack_lock(spinlock_t *lock) __acquires(lock)
 {
 	/* 1) Acquire the lock */
@@ -1404,7 +1400,7 @@ static void gc_worker(struct work_struct *work)
 				continue;
 
 			net = nf_ct_net(tmp);
-			cnet = net_generic(net, nf_conntrack_net_id);
+			cnet = nf_ct_pernet(net);
 			if (atomic_read(&cnet->count) < nf_conntrack_max95)
 				continue;
 
@@ -1484,7 +1480,7 @@ __nf_conntrack_alloc(struct net *net,
 		     const struct nf_conntrack_tuple *repl,
 		     gfp_t gfp, u32 hash)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 	unsigned int ct_count;
 	struct nf_conn *ct;
 
@@ -1556,7 +1552,7 @@ void nf_conntrack_free(struct nf_conn *ct)
 
 	nf_ct_ext_destroy(ct);
 	kmem_cache_free(nf_conntrack_cachep, ct);
-	cnet = net_generic(net, nf_conntrack_net_id);
+	cnet = nf_ct_pernet(net);
 
 	smp_mb__before_atomic();
 	atomic_dec(&cnet->count);
@@ -1614,7 +1610,7 @@ init_conntrack(struct net *net, struct nf_conn *tmpl,
 			     GFP_ATOMIC);
 
 	local_bh_disable();
-	cnet = net_generic(net, nf_conntrack_net_id);
+	cnet = nf_ct_pernet(net);
 	if (cnet->expect_count) {
 		spin_lock(&nf_conntrack_expect_lock);
 		exp = nf_ct_find_expectation(net, zone, tuple);
@@ -2317,7 +2313,7 @@ __nf_ct_unconfirmed_destroy(struct net *net)
 
 void nf_ct_unconfirmed_destroy(struct net *net)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
 	might_sleep();
 
@@ -2333,7 +2329,7 @@ void nf_ct_iterate_cleanup_net(struct net *net,
 			       int (*iter)(struct nf_conn *i, void *data),
 			       void *data, u32 portid, int report)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 	struct iter_data d;
 
 	might_sleep();
@@ -2367,7 +2363,7 @@ nf_ct_iterate_destroy(int (*iter)(struct nf_conn *i, void *data), void *data)
 
 	down_read(&net_rwsem);
 	for_each_net(net) {
-		struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+		struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
 		if (atomic_read(&cnet->count) == 0)
 			continue;
@@ -2449,7 +2445,7 @@ void nf_conntrack_cleanup_net_list(struct list_head *net_exit_list)
 i_see_dead_people:
 	busy = 0;
 	list_for_each_entry(net, net_exit_list, exit_list) {
-		struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+		struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
 		nf_ct_iterate_cleanup(kill_all, net, 0, 0);
 		if (atomic_read(&cnet->count) != 0)
@@ -2733,7 +2729,7 @@ void nf_conntrack_init_end(void)
 
 int nf_conntrack_init_net(struct net *net)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 	int ret = -ENOMEM;
 	int cpu;
 
net/netfilter/nf_conntrack_ecache.c +3 -5
index 759d87aef95f..296e4a171bd1 100644
@@ -27,8 +27,6 @@
 #include <net/netfilter/nf_conntrack_ecache.h>
 #include <net/netfilter/nf_conntrack_extend.h>
 
-extern unsigned int nf_conntrack_net_id;
-
 static DEFINE_MUTEX(nf_ct_ecache_mutex);
 
 #define ECACHE_RETRY_WAIT (HZ/10)
@@ -348,7 +346,7 @@ EXPORT_SYMBOL_GPL(nf_ct_expect_unregister_notifier);
 
 void nf_conntrack_ecache_work(struct net *net, enum nf_ct_ecache_state state)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
 	if (state == NFCT_ECACHE_DESTROY_FAIL &&
 	    !delayed_work_pending(&cnet->ecache_dwork)) {
@@ -371,7 +369,7 @@ static const struct nf_ct_ext_type event_extend = {
 
 void nf_conntrack_ecache_pernet_init(struct net *net)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
 	net->ct.sysctl_events = nf_ct_events;
 	cnet->ct_net = &net->ct;
@@ -380,7 +378,7 @@ void nf_conntrack_ecache_pernet_init(struct net *net)
 
 void nf_conntrack_ecache_pernet_fini(struct net *net)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
 	cancel_delayed_work_sync(&cnet->ecache_dwork);
 }
net/netfilter/nf_conntrack_expect.c +5 -7
index efdd391b3f72..1e851bc2e61a 100644
@@ -43,8 +43,6 @@ unsigned int nf_ct_expect_max __read_mostly;
 static struct kmem_cache *nf_ct_expect_cachep __read_mostly;
 static unsigned int nf_ct_expect_hashrnd __read_mostly;
 
-extern unsigned int nf_conntrack_net_id;
-
 /* nf_conntrack_expect helper functions */
 void nf_ct_unlink_expect_report(struct nf_conntrack_expect *exp,
 				u32 portid, int report)
@@ -58,7 +56,7 @@ void nf_ct_unlink_expect_report(struct nf_conntrack_expect *exp,
 
 	hlist_del_rcu(&exp->hnode);
 
-	cnet = net_generic(net, nf_conntrack_net_id);
+	cnet = nf_ct_pernet(net);
 	cnet->expect_count--;
 
 	hlist_del_rcu(&exp->lnode);
@@ -123,7 +121,7 @@ __nf_ct_expect_find(struct net *net,
 		    const struct nf_conntrack_zone *zone,
 		    const struct nf_conntrack_tuple *tuple)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 	struct nf_conntrack_expect *i;
 	unsigned int h;
 
@@ -164,7 +162,7 @@ nf_ct_find_expectation(struct net *net,
 		       const struct nf_conntrack_zone *zone,
 		       const struct nf_conntrack_tuple *tuple)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 	struct nf_conntrack_expect *i, *exp = NULL;
 	unsigned int h;
 
@@ -397,7 +395,7 @@ static void nf_ct_expect_insert(struct nf_conntrack_expect *exp)
 	master_help->expecting[exp->class]++;
 
 	hlist_add_head_rcu(&exp->hnode, &nf_ct_expect_hash[h]);
-	cnet = net_generic(net, nf_conntrack_net_id);
+	cnet = nf_ct_pernet(net);
 	cnet->expect_count++;
 
 	NF_CT_STAT_INC(net, expect_create);
@@ -468,7 +466,7 @@ static inline int __nf_ct_expect_check(struct nf_conntrack_expect *expect,
 		}
 	}
 
-	cnet = net_generic(net, nf_conntrack_net_id);
+	cnet = nf_ct_pernet(net);
 	if (cnet->expect_count >= nf_ct_expect_max) {
 		net_warn_ratelimited("nf_conntrack: expectation table full\n");
 		ret = -EMFILE;
net/netfilter/nf_conntrack_helper.c +2 -4
index ac396cc8bfae..ae4488a13c70 100644
@@ -43,8 +43,6 @@ MODULE_PARM_DESC(nf_conntrack_helper,
 static DEFINE_MUTEX(nf_ct_nat_helpers_mutex);
 static struct list_head nf_ct_nat_helpers __read_mostly;
 
-extern unsigned int nf_conntrack_net_id;
-
 /* Stupid hash, but collision free for the default registrations of the
  * helpers currently in the kernel. */
 static unsigned int helper_hash(const struct nf_conntrack_tuple *tuple)
@@ -214,7 +212,7 @@ EXPORT_SYMBOL_GPL(nf_ct_helper_ext_add);
 static struct nf_conntrack_helper *
 nf_ct_lookup_helper(struct nf_conn *ct, struct net *net)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
 	if (!cnet->sysctl_auto_assign_helper) {
 		if (cnet->auto_assign_helper_warned)
@@ -560,7 +558,7 @@ static const struct nf_ct_ext_type helper_extend = {
 
 void nf_conntrack_helper_pernet_init(struct net *net)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
 	cnet->sysctl_auto_assign_helper = nf_ct_auto_assign_helper;
 }
net/netfilter/nf_conntrack_proto.c +2 -4
index 89e5bac384d7..fbc1fa36d2c2 100644
@@ -42,8 +42,6 @@
 #include <net/ipv6.h>
 #include <net/inet_frag.h>
 
-extern unsigned int nf_conntrack_net_id;
-
 static DEFINE_MUTEX(nf_ct_proto_mutex);
 
 #ifdef CONFIG_SYSCTL
@@ -446,7 +444,7 @@ static struct nf_ct_bridge_info *nf_ct_bridge_info;
 
 static int nf_ct_netns_do_get(struct net *net, u8 nfproto)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 	bool fixup_needed = false, retry = true;
 	int err = 0;
 retry:
@@ -531,7 +529,7 @@ static int nf_ct_netns_do_get(struct net *net, u8 nfproto)
 
 static void nf_ct_netns_do_put(struct net *net, u8 nfproto)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
 	mutex_lock(&nf_ct_proto_mutex);
 	switch (nfproto) {
net/netfilter/nf_conntrack_standalone.c +3 -5
index aaa55246d0ca..bce93656fad9 100644
@@ -512,9 +512,7 @@ static void nf_conntrack_standalone_fini_proc(struct net *net)
 
 u32 nf_conntrack_count(const struct net *net)
 {
-	const struct nf_conntrack_net *cnet;
-
-	cnet = net_generic(net, nf_conntrack_net_id);
+	const struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
 	return atomic_read(&cnet->count);
 }
@@ -1032,7 +1030,7 @@ static void nf_conntrack_standalone_init_gre_sysctl(struct net *net,
 
 static int nf_conntrack_standalone_init_sysctl(struct net *net)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 	struct nf_udp_net *un = nf_udp_pernet(net);
 	struct ctl_table *table;
 
@@ -1085,7 +1083,7 @@ static int nf_conntrack_standalone_init_sysctl(struct net *net)
 
 static void nf_conntrack_standalone_fini_sysctl(struct net *net)
 {
-	struct nf_conntrack_net *cnet = net_generic(net, nf_conntrack_net_id);
+	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 	struct ctl_table *table;
 
 	table = cnet->sysctl_header->ctl_table_arg;
Signature · 1 line
2.30.2
author maintainer
#5 lore raw

PATCH 5/13 netfilter: conntrack: Introduce tcp offload timeout configuration

In reply to Pablo Neira Ayuso

From: Oz Shlomo <[email protected]>

TCP connections may be offloaded from nf conntrack to nf flow table. Offloaded connections are aged after 30 seconds of inactivity. Once aged, ownership is returned to conntrack with a hard coded pickup time of 120 seconds, after which the connection may be deleted. eted. The current aging intervals may be too aggressive for some users.

Provide users with the ability to control the nf flow table offload aging and pickup time intervals via sysctl parameter as a pre-step for configuring the nf flow table GC timeout intervals.

---
 include/net/netns/conntrack.h           |  4 ++++
 net/netfilter/nf_conntrack_proto_tcp.c  |  5 +++++
 net/netfilter/nf_conntrack_standalone.c | 24 ++++++++++++++++++++++++
 3 files changed, 33 insertions(+)
include/net/netns/conntrack.h +4 -0
index ad0a95c2335e..3a391e27ec60 100644
@@ -27,6 +27,10 @@ struct nf_tcp_net {
 	u8 tcp_loose;
 	u8 tcp_be_liberal;
 	u8 tcp_max_retrans;
+#if IS_ENABLED(CONFIG_NF_FLOW_TABLE)
+	unsigned int offload_timeout;
+	unsigned int offload_pickup;
+#endif
 };
 
 enum udp_conntrack {
net/netfilter/nf_conntrack_proto_tcp.c +5 -0
index 34e22416a721..de840fc41a2e 100644
@@ -1441,6 +1441,11 @@ void nf_conntrack_tcp_init_net(struct net *net)
 	 * will be started.
 	 */
 	tn->tcp_max_retrans = 3;
+
+#if IS_ENABLED(CONFIG_NF_FLOW_TABLE)
+	tn->offload_timeout = 30 * HZ;
+	tn->offload_pickup = 120 * HZ;
+#endif
 }
 
 const struct nf_conntrack_l4proto nf_conntrack_l4proto_tcp =
net/netfilter/nf_conntrack_standalone.c +24 -0
index bce93656fad9..67b0fcd1a787 100644
@@ -573,6 +573,10 @@ enum nf_ct_sysctl_index {
 	NF_SYSCTL_CT_PROTO_TIMEOUT_TCP_CLOSE,
 	NF_SYSCTL_CT_PROTO_TIMEOUT_TCP_RETRANS,
 	NF_SYSCTL_CT_PROTO_TIMEOUT_TCP_UNACK,
+#if IS_ENABLED(CONFIG_NF_FLOW_TABLE)
+	NF_SYSCTL_CT_PROTO_TIMEOUT_TCP_OFFLOAD,
+	NF_SYSCTL_CT_PROTO_TIMEOUT_TCP_OFFLOAD_PICKUP,
+#endif
 	NF_SYSCTL_CT_PROTO_TCP_LOOSE,
 	NF_SYSCTL_CT_PROTO_TCP_LIBERAL,
 	NF_SYSCTL_CT_PROTO_TCP_MAX_RETRANS,
@@ -760,6 +764,20 @@ static struct ctl_table nf_ct_sysctl_table[] = {
 		.mode		= 0644,
 		.proc_handler	= proc_dointvec_jiffies,
 	},
+#if IS_ENABLED(CONFIG_NF_FLOW_TABLE)
+	[NF_SYSCTL_CT_PROTO_TIMEOUT_TCP_OFFLOAD] = {
+		.procname	= "nf_flowtable_tcp_timeout",
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_jiffies,
+	},
+	[NF_SYSCTL_CT_PROTO_TIMEOUT_TCP_OFFLOAD_PICKUP] = {
+		.procname	= "nf_flowtable_tcp_pickup",
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_jiffies,
+	},
+#endif
 	[NF_SYSCTL_CT_PROTO_TCP_LOOSE] = {
 		.procname	= "nf_conntrack_tcp_loose",
 		.maxlen		= sizeof(u8),
@@ -969,6 +987,12 @@ static void nf_conntrack_standalone_init_tcp_sysctl(struct net *net,
 	XASSIGN(LIBERAL, &tn->tcp_be_liberal);
 	XASSIGN(MAX_RETRANS, &tn->tcp_max_retrans);
 #undef XASSIGN
+
+#if IS_ENABLED(CONFIG_NF_FLOW_TABLE)
+	table[NF_SYSCTL_CT_PROTO_TIMEOUT_TCP_OFFLOAD].data = &tn->offload_timeout;
+	table[NF_SYSCTL_CT_PROTO_TIMEOUT_TCP_OFFLOAD_PICKUP].data = &tn->offload_pickup;
+#endif
+
 }
 
 static void nf_conntrack_standalone_init_sctp_sysctl(struct net *net,
Signature · 1 line
2.30.2
author maintainer
#6 lore raw

PATCH 6/13 netfilter: conntrack: Introduce udp offload timeout configuration

In reply to Pablo Neira Ayuso

From: Oz Shlomo <[email protected]>

UDP connections may be offloaded from nf conntrack to nf flow table. Offloaded connections are aged after 30 seconds of inactivity. Once aged, ownership is returned to conntrack with a hard coded pickup time of 30 seconds, after which the connection may be deleted. eted. The current aging intervals may be too aggressive for some users.

Provide users with the ability to control the nf flow table offload aging and pickup time intervals via sysctl parameter as a pre-step for configuring the nf flow table GC timeout intervals.

---
 include/net/netns/conntrack.h           |  4 ++++
 net/netfilter/nf_conntrack_proto_udp.c  |  5 +++++
 net/netfilter/nf_conntrack_standalone.c | 22 ++++++++++++++++++++++
 3 files changed, 31 insertions(+)
include/net/netns/conntrack.h +4 -0
index 3a391e27ec60..c3094b83a525 100644
@@ -41,6 +41,10 @@ enum udp_conntrack {
 
 struct nf_udp_net {
 	unsigned int timeouts[UDP_CT_MAX];
+#if IS_ENABLED(CONFIG_NF_FLOW_TABLE)
+	unsigned int offload_timeout;
+	unsigned int offload_pickup;
+#endif
 };
 
 struct nf_icmp_net {
net/netfilter/nf_conntrack_proto_udp.c +5 -0
index af402f458ee0..68911fcaa0f1 100644
@@ -270,6 +270,11 @@ void nf_conntrack_udp_init_net(struct net *net)
 
 	for (i = 0; i < UDP_CT_MAX; i++)
 		un->timeouts[i] = udp_timeouts[i];
+
+#if IS_ENABLED(CONFIG_NF_FLOW_TABLE)
+	un->offload_timeout = 30 * HZ;
+	un->offload_pickup = 30 * HZ;
+#endif
 }
 
 const struct nf_conntrack_l4proto nf_conntrack_l4proto_udp =
net/netfilter/nf_conntrack_standalone.c +22 -0
index 67b0fcd1a787..f57a951c9b5e 100644
@@ -582,6 +582,10 @@ enum nf_ct_sysctl_index {
 	NF_SYSCTL_CT_PROTO_TCP_MAX_RETRANS,
 	NF_SYSCTL_CT_PROTO_TIMEOUT_UDP,
 	NF_SYSCTL_CT_PROTO_TIMEOUT_UDP_STREAM,
+#if IS_ENABLED(CONFIG_NF_FLOW_TABLE)
+	NF_SYSCTL_CT_PROTO_TIMEOUT_UDP_OFFLOAD,
+	NF_SYSCTL_CT_PROTO_TIMEOUT_UDP_OFFLOAD_PICKUP,
+#endif
 	NF_SYSCTL_CT_PROTO_TIMEOUT_ICMP,
 	NF_SYSCTL_CT_PROTO_TIMEOUT_ICMPV6,
 #ifdef CONFIG_NF_CT_PROTO_SCTP
@@ -812,6 +816,20 @@ static struct ctl_table nf_ct_sysctl_table[] = {
 		.mode		= 0644,
 		.proc_handler	= proc_dointvec_jiffies,
 	},
+#if IS_ENABLED(CONFIG_NFT_FLOW_OFFLOAD)
+	[NF_SYSCTL_CT_PROTO_TIMEOUT_UDP_OFFLOAD] = {
+		.procname	= "nf_flowtable_udp_timeout",
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_jiffies,
+	},
+	[NF_SYSCTL_CT_PROTO_TIMEOUT_UDP_OFFLOAD_PICKUP] = {
+		.procname	= "nf_flowtable_udp_pickup",
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_jiffies,
+	},
+#endif
 	[NF_SYSCTL_CT_PROTO_TIMEOUT_ICMP] = {
 		.procname	= "nf_conntrack_icmp_timeout",
 		.maxlen		= sizeof(unsigned int),
@@ -1081,6 +1099,10 @@ static int nf_conntrack_standalone_init_sysctl(struct net *net)
 	table[NF_SYSCTL_CT_PROTO_TIMEOUT_ICMPV6].data = &nf_icmpv6_pernet(net)->timeout;
 	table[NF_SYSCTL_CT_PROTO_TIMEOUT_UDP].data = &un->timeouts[UDP_CT_UNREPLIED];
 	table[NF_SYSCTL_CT_PROTO_TIMEOUT_UDP_STREAM].data = &un->timeouts[UDP_CT_REPLIED];
+#if IS_ENABLED(CONFIG_NF_FLOW_TABLE)
+	table[NF_SYSCTL_CT_PROTO_TIMEOUT_UDP_OFFLOAD].data = &un->offload_timeout;
+	table[NF_SYSCTL_CT_PROTO_TIMEOUT_UDP_OFFLOAD_PICKUP].data = &un->offload_pickup;
+#endif
 
 	nf_conntrack_standalone_init_tcp_sysctl(net, table);
 	nf_conntrack_standalone_init_sctp_sysctl(net, table);
Signature · 1 line
2.30.2
author maintainer
#7 lore raw

PATCH 7/13 netfilter: flowtable: Set offload timeouts according to proto values

In reply to Pablo Neira Ayuso

From: Oz Shlomo <[email protected]>

Currently the aging period for tcp/udp connections is hard coded to 30 seconds. Aged tcp/udp connections configure a hard coded 120/30 seconds pickup timeout for conntrack.
This configuration may be too aggressive or permissive for some users.

Dynamically configure the nf flow table GC timeout intervals according to the user defined values.

---
 include/net/netfilter/nf_flow_table.h |  2 ++
 net/netfilter/nf_flow_table_core.c    | 47 +++++++++++++++++++++------
 net/netfilter/nf_flow_table_offload.c |  4 +--
 3 files changed, 41 insertions(+), 12 deletions(-)
include/net/netfilter/nf_flow_table.h +2 -0
index 48ef7460ff30..a3647fadf1cc 100644
@@ -177,6 +177,8 @@ struct flow_offload {
 #define NF_FLOW_TIMEOUT (30 * HZ)
 #define nf_flowtable_time_stamp	(u32)jiffies
 
+unsigned long flow_offload_get_timeout(struct flow_offload *flow);
+
 static inline __s32 nf_flow_timeout_delta(unsigned int timeout)
 {
 	return (__s32)(timeout - nf_flowtable_time_stamp);
net/netfilter/nf_flow_table_core.c +37 -10

79 lines not on this page · raw message

net/netfilter/nf_flow_table_offload.c +2 -2
index 528b2f172684..f92006cec94c 100644
@@ -937,7 +937,7 @@ static void flow_offload_work_stats(struct flow_offload_work *offload)
 
 	lastused = max_t(u64, stats[0].lastused, stats[1].lastused);
 	offload->flow->timeout = max_t(u64, offload->flow->timeout,
-				       lastused + NF_FLOW_TIMEOUT);
+				       lastused + flow_offload_get_timeout(offload->flow));
 
 	if (offload->flowtable->flags & NF_FLOWTABLE_COUNTER) {
 		if (stats[0].pkts)
@@ -1041,7 +1041,7 @@ void nf_flow_offload_stats(struct nf_flowtable *flowtable,
 	__s32 delta;
 
 	delta = nf_flow_timeout_delta(flow->timeout);
-	if ((delta >= (9 * NF_FLOW_TIMEOUT) / 10))
+	if ((delta >= (9 * flow_offload_get_timeout(flow)) / 10))
 		return;
 
 	offload = nf_flow_offload_work_alloc(flowtable, flow, FLOW_CLS_STATS);
Signature · 1 line
2.30.2
author maintainer
#8 lore raw

PATCH 8/13 netfilter: annotate nf_tables base hook ops

In reply to Pablo Neira Ayuso

From: Florian Westphal <[email protected]>

This will allow a followup patch to treat the 'ops->priv' pointer as nft_chain argument without having to first walk the table/chains to check if there is a matching base chain pointer.

---
 include/linux/netfilter.h     | 8 +++++++-
 net/netfilter/nf_tables_api.c | 4 +++-
 2 files changed, 10 insertions(+), 2 deletions(-)
include/linux/netfilter.h +7 -1
index f161569fbe2f..3fda1a508733 100644
@@ -77,12 +77,18 @@ struct nf_hook_state {
 typedef unsigned int nf_hookfn(void *priv,
 			       struct sk_buff *skb,
 			       const struct nf_hook_state *state);
+enum nf_hook_ops_type {
+	NF_HOOK_OP_UNDEFINED,
+	NF_HOOK_OP_NF_TABLES,
+};
+
 struct nf_hook_ops {
 	/* User fills in from here down. */
 	nf_hookfn		*hook;
 	struct net_device	*dev;
 	void			*priv;
-	u_int8_t		pf;
+	u8			pf;
+	enum nf_hook_ops_type	hook_ops_type:8;
 	unsigned int		hooknum;
 	/* Hooks are ordered in ascending priority. */
 	int			priority;
net/netfilter/nf_tables_api.c +3 -1
index 6c2000a11c7e..c9308241b688 100644
@@ -2168,8 +2168,10 @@ static int nf_tables_addchain(struct nft_ctx *ctx, u8 family, u8 genmask,
 	}
 
 	nft_trans_chain_policy(trans) = NFT_CHAIN_POLICY_UNSET;
-	if (nft_is_base_chain(chain))
+	if (nft_is_base_chain(chain)) {
+		basechain->ops.hook_ops_type = NF_HOOK_OP_NF_TABLES;
 		nft_trans_chain_policy(trans) = policy;
+	}
 
 	err = nft_chain_add(table, chain);
 	if (err < 0) {
Signature · 1 line
2.30.2
author maintainer
#9 lore raw

PATCH 9/13 netfilter: add new hook nfnl subsystem

In reply to Pablo Neira Ayuso

From: Florian Westphal <[email protected]>

This nfnl subsystem allows to dump the list of all active netfiler hooks, e.g. defrag, conntrack, nf/ip/arp/ip6tables and so on.

This helps to see what kind of features are currently enabled in the network stack.

Sample output from nft tool using this infra:

 $ nft list hook ip input
 family ip hook input {
   +0000000010 nft_do_chain_inet [nf_tables] # nft table firewalld INPUT
   +0000000100 nf_nat_ipv4_local_in [nf_nat]
   +2147483647 ipv4_confirm [nf_conntrack]
 }
---
 include/uapi/linux/netfilter/nfnetlink.h      |   3 +-
 include/uapi/linux/netfilter/nfnetlink_hook.h |  55 +++
 net/netfilter/Kconfig                         |   9 +
 net/netfilter/Makefile                        |   1 +
 net/netfilter/nfnetlink.c                     |   1 +
 net/netfilter/nfnetlink_hook.c                | 375 ++++++++++++++++++
 6 files changed, 443 insertions(+), 1 deletion(-)
 create mode 100644 include/uapi/linux/netfilter/nfnetlink_hook.h
 create mode 100644 net/netfilter/nfnetlink_hook.c
include/uapi/linux/netfilter/nfnetlink.h +2 -1
index 5bc960f220b3..6cd58cd2a6f0 100644
@@ -60,7 +60,8 @@ struct nfgenmsg {
 #define NFNL_SUBSYS_CTHELPER		9
 #define NFNL_SUBSYS_NFTABLES		10
 #define NFNL_SUBSYS_NFT_COMPAT		11
-#define NFNL_SUBSYS_COUNT		12
+#define NFNL_SUBSYS_HOOK		12
+#define NFNL_SUBSYS_COUNT		13
 
 /* Reserved control nfnetlink messages */
 #define NFNL_MSG_BATCH_BEGIN		NLMSG_MIN_TYPE
include/uapi/linux/netfilter/nfnetlink_hook.h new file +55 -0

55 lines not on this page · raw message

net/netfilter/Kconfig +9 -0
index 172d74560632..c81321372198 100644
@@ -19,6 +19,15 @@ config NETFILTER_FAMILY_BRIDGE
 config NETFILTER_FAMILY_ARP
 	bool
 
+config NETFILTER_NETLINK_HOOK
+	tristate "Netfilter base hook dump support"
+	depends on NETFILTER_ADVANCED
+	select NETFILTER_NETLINK
+	help
+	  If this option is enabled, the kernel will include support
+	  to list the base netfilter hooks via NFNETLINK.
+	  This is helpful for debugging.
+
 config NETFILTER_NETLINK_ACCT
 	tristate "Netfilter NFACCT over NFNETLINK interface"
 	depends on NETFILTER_ADVANCED
net/netfilter/Makefile +1 -0
index e80e010354b1..87112dad1fd4 100644
@@ -22,6 +22,7 @@ obj-$(CONFIG_NETFILTER_NETLINK_ACCT) += nfnetlink_acct.o
 obj-$(CONFIG_NETFILTER_NETLINK_QUEUE) += nfnetlink_queue.o
 obj-$(CONFIG_NETFILTER_NETLINK_LOG) += nfnetlink_log.o
 obj-$(CONFIG_NETFILTER_NETLINK_OSF) += nfnetlink_osf.o
+obj-$(CONFIG_NETFILTER_NETLINK_HOOK) += nfnetlink_hook.o
 
 # connection tracking
 obj-$(CONFIG_NF_CONNTRACK) += nf_conntrack.o
net/netfilter/nfnetlink.c +1 -0
index 028a1f39318b..7e2c8dd01408 100644
@@ -68,6 +68,7 @@ static const char *const nfnl_lockdep_names[NFNL_SUBSYS_COUNT] = {
 	[NFNL_SUBSYS_CTHELPER] = "nfnl_subsys_cthelper",
 	[NFNL_SUBSYS_NFTABLES] = "nfnl_subsys_nftables",
 	[NFNL_SUBSYS_NFT_COMPAT] = "nfnl_subsys_nftcompat",
+	[NFNL_SUBSYS_HOOK] = "nfnl_subsys_hook",
 };
 
 static const int nfnl_group2type[NFNLGRP_MAX+1] = {
net/netfilter/nfnetlink_hook.c new file +375 -0

375 lines not on this page · raw message

Signature · 1 line
2.30.2
author maintainer
#10 lore raw

PATCH 10/13 netfilter: nft_exthdr: Fix for unsafe packet data read

In reply to Pablo Neira Ayuso

From: Phil Sutter <[email protected]>

While iterating through an SCTP packet's chunks, skb_header_pointer() is called for the minimum expected chunk header size. If (that part of) the skbuff is non-linear, the following memcpy() may read data past temporary buffer '_sch'. Use skb_copy_bits() instead which does the right thing in this situation.

Fixes 133dc203d77df ("netfilter: nft_exthdr: Support SCTP chunks")
---
 net/netfilter/nft_exthdr.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
net/netfilter/nft_exthdr.c +3 -1
index 1b0579cb62d0..7f705b5c09de 100644
@@ -327,7 +327,9 @@ static void nft_exthdr_sctp_eval(const struct nft_expr *expr,
 				break;
 
 			dest[priv->len / NFT_REG32_SIZE] = 0;
-			memcpy(dest, (char *)sch + priv->offset, priv->len);
+			if (skb_copy_bits(pkt->skb, offset + priv->offset,
+					  dest, priv->len) < 0)
+				break;
 			return;
 		}
 		offset += SCTP_PAD4(ntohs(sch->length));
Signature · 1 line
2.30.2
author maintainer
#11 lore raw

PATCH 11/13 netfilter: nfnetlink_hook: fix array index out-of-bounds error

In reply to Pablo Neira Ayuso

From: Colin Ian King <[email protected]>

Currently the array net->nf.hooks_ipv6 is accessed by index hook before hook is sanity checked. Fix this by moving the sanity check to before the array access.

Addresses-Coverity: ("Out-of-bounds access")

Fixes e2cf17d3774c ("netfilter: add new hook nfnl subsystem")
---
 net/netfilter/nfnetlink_hook.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/netfilter/nfnetlink_hook.c +1 -1
index 04586dfa2acd..58fda6ac663b 100644
@@ -181,9 +181,9 @@ nfnl_hook_entries_head(u8 pf, unsigned int hook, struct net *net, const char *de
 		hook_head = rcu_dereference(net->nf.hooks_ipv4[hook]);
 		break;
 	case NFPROTO_IPV6:
-		hook_head = rcu_dereference(net->nf.hooks_ipv6[hook]);
 		if (hook >= ARRAY_SIZE(net->nf.hooks_ipv6))
 			return ERR_PTR(-EINVAL);
+		hook_head = rcu_dereference(net->nf.hooks_ipv6[hook]);
 		break;
 	case NFPROTO_ARP:
 #ifdef CONFIG_NETFILTER_FAMILY_ARP
Signature · 1 line
2.30.2
author maintainer
#12 lore raw

PATCH 12/13 netfilter: nfnetlink_hook: add depends-on nftables

In reply to Pablo Neira Ayuso

From: Florian Westphal <[email protected]>

nfnetlink_hook.c: In function 'nfnl_hook_put_nft_chain_info':
nfnetlink_hook.c:76:7: error: implicit declaration of 'nft_is_active'

This macro is only defined when NF_TABLES is enabled. While its possible to also add an ifdef-guard, the infrastructure is currently not useful without nf_tables.

Fixes 252956528caa ("netfilter: add new hook nfnl subsystem")
---
 net/netfilter/Kconfig | 1 +
 1 file changed, 1 insertion(+)
net/netfilter/Kconfig +1 -0
index c81321372198..54395266339d 100644
@@ -22,6 +22,7 @@ config NETFILTER_FAMILY_ARP
 config NETFILTER_NETLINK_HOOK
 	tristate "Netfilter base hook dump support"
 	depends on NETFILTER_ADVANCED
+	depends on NF_TABLES
 	select NETFILTER_NETLINK
 	help
 	  If this option is enabled, the kernel will include support
Signature · 1 line
2.30.2
author maintainer
#13 lore raw

PATCH 13/13 netfilter: nf_tables: move base hook annotation to init helper

In reply to Pablo Neira Ayuso

From: Florian Westphal <[email protected]>

coverity scanner says:

2187  if (nft_is_base_chain(chain)) {
vvv   CID 1505166:  Memory - corruptions  (UNINIT)
vvv   Using uninitialized value "basechain".
2188  basechain->ops.hook_ops_type = NF_HOOK_OP_NF_TABLES;

... I don't see how nft_is_base_chain() can evaluate to true while basechain pointer is garbage.

However, it seems better to place the NF_HOOK_OP_NF_TABLES annotation in nft_basechain_hook_init() instead.

Addresses-Coverity-ID: 1505166 ("Memory - corruptions")

Fixes 65b8b7bfc5284f ("netfilter: annotate nf_tables base hook ops")
---
 net/netfilter/nf_tables_api.c | 15 +++++++--------
 1 file changed, 7 insertions(+), 8 deletions(-)
net/netfilter/nf_tables_api.c +7 -8
index c9308241b688..caaff7ab9e73 100644
@@ -1997,11 +1997,12 @@ static void nft_basechain_hook_init(struct nf_hook_ops *ops, u8 family,
 				    const struct nft_chain_hook *hook,
 				    struct nft_chain *chain)
 {
-	ops->pf		= family;
-	ops->hooknum	= hook->num;
-	ops->priority	= hook->priority;
-	ops->priv	= chain;
-	ops->hook	= hook->type->hooks[ops->hooknum];
+	ops->pf			= family;
+	ops->hooknum		= hook->num;
+	ops->priority		= hook->priority;
+	ops->priv		= chain;
+	ops->hook		= hook->type->hooks[ops->hooknum];
+	ops->hook_ops_type	= NF_HOOK_OP_NF_TABLES;
 }
 
 static int nft_basechain_init(struct nft_base_chain *basechain, u8 family,
@@ -2168,10 +2169,8 @@ static int nf_tables_addchain(struct nft_ctx *ctx, u8 family, u8 genmask,
 	}
 
 	nft_trans_chain_policy(trans) = NFT_CHAIN_POLICY_UNSET;
-	if (nft_is_base_chain(chain)) {
-		basechain->ops.hook_ops_type = NF_HOOK_OP_NF_TABLES;
+	if (nft_is_base_chain(chain))
 		nft_trans_chain_policy(trans) = policy;
-	}
 
 	err = nft_chain_add(table, chain);
 	if (err < 0) {
Signature · 1 line
2.30.2