Skip to contentSkip to conversation

netfilter: nf_tables: Implement jump limit for nft_table_validate

PATCH Superseded Shaun Brady 3 participants 8 messages May 20, 2025 – Jul 28, 2025

Subsystems (4)

v5 of 6 Superseded Latest is v6 →

lore
Share & tools
lore
#0 lore raw

PATCH v5 1/2 netfilter: nf_tables: Implement jump limit for nft_table_validate

Observing https://bugzilla.netfilter.org/show_bug.cgi?id=1665, I was able to reproduce the bug using linux-stable. Summarized, when adding large/repeated jump chains, while still staying under the NFT_JUMP_STACK_SIZE (currently 16), the kernel soons locks up.

From the bug report:
table='loop-test'
nft add table inet $table
nft add chain inet $table test0 '{type filter hook input priority 0; policy accept;}' for((i=1;i<16;i++));do nft add chain inet $table test$i;done
nft add rule inet $table test0 jump test1
for((i=1;i<15;i++));do nft add rule inet $table test$i jump test$((i+1));done nft add rule inet $table test15 tcp dport 8080 drop

  After the jump rule is added for 3 to 5 times, the system freezes and even softlockup occurs.
  for((i=1;i<15;i++));do nft add rule inet $table test$i jump test$((i+1));done
  for((i=1;i<15;i++));do nft add rule inet $table test$i jump test$((i+1));done
  for((i=1;i<15;i++));do nft add rule inet $table test$i jump test$((i+1));done

This patch is a different approach than the original proposed approach found in the bug report. Additionally, the limit, namespace specific, is only applied to non-init-net namespaces, with the common use case being to protect against rogue containers.

Add a new counter, total_jump_counter, to nft_ctx. On every call to nft_table_validate() (rule addition time, versus packet inspection time) start the counter at the current sum of all jump counts in all other tables, sans 4 vs 6 differences.

Increment said counter for every jump encountered during table validation. If the counter ever exceeds the namespaces jump limit *during validation*, gracefully reject the rule with -EMLINK (the same behavior as exceeding NFT_JUMP_STACK_SIZE).

This allows immediate feedback to the user about a bad chain, versus the original idea (from the bug report) of allowing the addition to the table. It keeps the in memory ruleset consistent, versus catching the failure during packet inspection at some unknown point in the future and arbitrarily denying the packet. Most importantly, it removes the original problem of a kernel crash.

The compile time limit NFT_DEFAULT_MAX_TABLE_JUMPS of 65536 was chosen to account for any normal use case, and when this value (and associated stressing loop table) was tested against a 1CPU/256MB machine, the system remained functional.

A sysctl entry net/netfilter/nf_max_table_jumps_netns for the limit was also added for any use cases that may exceed this limit. It is network namespace specific. As it is a control limit, for namespaces with an owner that is non-init_user_ns, this sysctl is read only.

Example output of nft when patch is applied (and count is reached):

Error: Could not process rule: Too many links
add rule inet loop-test test14 jump test15
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

v2: nf_max_table_jumps renamed to nf_max_table_jumps_netns;
Limit namespace specific, only applies to non-init netns; Limit raised to 16k
v3: Fix improper conditional compile; removed unneeded pernet subsys;
Include NFPROTO_IPV4|6 for _INET matches; add selftest script
v4: Simplify table jump count logic in for netns
v5: Move jump family test to helper function; move test to second commit

---
 Documentation/networking/netfilter-sysctl.rst |   9 ++
 include/net/netfilter/nf_tables.h             |   4 +
 include/net/netns/netfilter.h                 |   2 +
 net/netfilter/nf_tables_api.c                 | 125 +++++++++++++++++-
 net/netfilter/nft_immediate.c                 |   1 +
 5 files changed, 139 insertions(+), 2 deletions(-)
Documentation/networking/netfilter-sysctl.rst +9 -0
index beb6d7b275d4..8a87c72f0672 100644
@@ -15,3 +15,12 @@ nf_log_all_netns - BOOLEAN
 	with LOG target; this aims to prevent containers from flooding host
 	kernel log. If enabled, this target also works in other network
 	namespaces. This variable is only accessible from init_net.
+
+nf_max_table_jumps_netns - INTEGER (count)
+	default 65536
+
+	The maximum numbers of jumps a netns can have across its tables.
+	This only applies to non-init_net namespaces, and is read only for
+	non-init_user_ns namespaces.  Meeting or exceeding this value will
+	cause additional rules to not be added, with EMLINK being return to
+	the user.
include/net/netfilter/nf_tables.h +4 -0
index 803d5f1601f9..61b8ae421e4f 100644
@@ -205,6 +205,7 @@ static inline void nft_data_copy(u32 *dst, const struct nft_data *src,
  *	@nla: netlink attributes
  *	@portid: netlink portID of the original message
  *	@seq: netlink sequence number
+ *	@total_jump_count: Found jumps for netns
  *	@flags: modifiers to new request
  *	@family: protocol family
  *	@level: depth of the chains
@@ -218,6 +219,7 @@ struct nft_ctx {
 	const struct nlattr * const 	*nla;
 	u32				portid;
 	u32				seq;
+	u32				total_jump_count;
 	u16				flags;
 	u8				family;
 	u8				level;
@@ -1275,6 +1277,7 @@ static inline void nft_use_inc_restore(u32 *use)
  *	@hgenerator: handle generator state
  *	@handle: table handle
  *	@use: number of chain references to this table
+ *	@total_jump_count: jumps as per last validate
  *	@family:address family
  *	@flags: table flag (see enum nft_table_flags)
  *	@genmask: generation mask
@@ -1294,6 +1297,7 @@ struct nft_table {
 	u64				hgenerator;
 	u64				handle;
 	u32				use;
+	u32				total_jump_count;
 	u16				family:6,
 					flags:8,
 					genmask:2;
include/net/netns/netfilter.h +2 -0
index a6a0bf4a247e..c0dcd85d108c 100644
@@ -15,6 +15,7 @@ struct netns_nf {
 	const struct nf_logger __rcu *nf_loggers[NFPROTO_NUMPROTO];
 #ifdef CONFIG_SYSCTL
 	struct ctl_table_header *nf_log_dir_header;
+	struct ctl_table_header *nf_limit_control_dir_header;
 #ifdef CONFIG_LWTUNNEL
 	struct ctl_table_header *nf_lwtnl_dir_header;
 #endif
@@ -33,5 +34,6 @@ struct netns_nf {
 #if IS_ENABLED(CONFIG_NF_DEFRAG_IPV6)
 	unsigned int defrag_ipv6_users;
 #endif
+	u32 nf_max_table_jumps_netns __maybe_unused;
 };
 #endif
net/netfilter/nf_tables_api.c +123 -2

187 lines not on this page · raw message

net/netfilter/nft_immediate.c +1 -0
index 02ee5fb69871..b21736e389a4 100644
@@ -260,6 +260,7 @@ static int nft_immediate_validate(const struct nft_ctx *ctx,
 	case NFT_JUMP:
 	case NFT_GOTO:
 		pctx->level++;
+		pctx->total_jump_count++;
 		err = nft_chain_validate(ctx, data->verdict.chain);
 		if (err < 0)
 			return err;
Signature · 1 line
2.49.0
#1 lore raw

PATCH v5 2/2 Add test for nft_max_table_jumps_netns sysctl

Introduce test for recently added jump limit functionality. Tests sysctl behavior with regard to netns, as well as calling user_ns.

---
 .../testing/selftests/net/netfilter/Makefile  |   1 +
 .../netfilter/nft_max_table_jumps_netns.sh    | 227 ++++++++++++++++++
 2 files changed, 228 insertions(+)
 create mode 100755 tools/testing/selftests/net/netfilter/nft_max_table_jumps_netns.sh
tools/testing/selftests/net/netfilter/Makefile +1 -0
index ffe161fac8b5..bc7df8feb0f7 100644
@@ -23,6 +23,7 @@ TEST_PROGS += nft_concat_range.sh
 TEST_PROGS += nft_conntrack_helper.sh
 TEST_PROGS += nft_fib.sh
 TEST_PROGS += nft_flowtable.sh
+TEST_PROGS += nft_max_table_jumps_netns.sh
 TEST_PROGS += nft_meta.sh
 TEST_PROGS += nft_nat.sh
 TEST_PROGS += nft_nat_zones.sh
tools/testing/selftests/net/netfilter/nft_max_table_jumps_netns.sh new file +227 -0

227 lines not on this page · raw message

Signature · 1 line
2.49.0
#3 lore raw
Quoted 2 lines · Florian Westphal

LGTM, thanks.

Much thanks to you for your feedback and patience.

Double checking... is there anything else for me to do in regards to this patch at this time, or is it in the hands of the merge process of the team(s)?

Just want to make sure no one is expecting action from me.

Thanks!!

SB

maintainer
#4 lore raw
Quoted 9 lines · Shaun Brady
Quoted 2 lines · Florian Westphal

LGTM, thanks.

Much thanks to you for your feedback and patience.

Double checking... is there anything else for me to do in regards to this patch at this time, or is it in the hands of the merge process of the team(s)?

I don't think there is anyhing left do do here except waiting. As 6.15 was just released I'd expect that there will be a backlog until the development trees have been re-synced with linus tree.

#5 lore raw
In reply to Shaun Brady

On Mon, May 19, 2025 at 11:08:41PM -0400, Shaun Brady wrote:
[...]

Quoted 10 lines
net/netfilter/nf_tables_api.c +5 -1
@@ -4039,14 +4049,62 @@ int nft_chain_validate(const struct nft_ctx *ctx, const struct nft_chain *chain)
 }
 EXPORT_SYMBOL_GPL(nft_chain_validate);
 
-static int nft_table_validate(struct net *net, const struct nft_table *table)
+/** nft_families_inc_jump - Determine if tables should add to the total jump
+ * count for a netns.
+ *
+ * @table: table of interest
+ * @sibling_table: a 'sibling' table to compare against

net/netfilter/nf_tables_api.c int nft_chain_validate(const struct nft_ctx *ctx, const struct nft_chain *chain) · line 4066

Quoted 10 lines
net/netfilter/nf_tables_api.c +10 -0
+ *
+ * Compare attributes of the tables to determine if the sibling tables
+ * total_jump_count should be added to the working context (done by caller).
+ * Mostly concerned with family compatibilities, but also identifying equality
+ * (a tables own addition will be recalculated soon).
+ *
+ * Ex: v4 tables do not apply to v6 packets
+ */
+static bool nft_families_inc_jump(struct nft_table *table, struct nft_table *sibling_table)
+{

I think we mentioned about NFPROTO_BRIDGE here too.

Quoted 9 lines
net/netfilter/nf_tables_api.c +9 -0
+	/* Invert parameters to on require one test for two cases (the reverse) */
+	if (sibling_table->family > table->family) /* include/uapi/linux/netfilter.h */
+		return nft_families_inc_jump(sibling_table, table);
+
+	/* We found ourselves; don't add current jump count (will be counted dynamically) */
+	if (sibling_table == table)
+		return false;
+
+	switch (table->family) {

net/netfilter/nf_tables_api.c int nft_chain_validate(const struct nft_ctx *ctx, const struct nft_chain *chain) · line 4085

Quoted 10 lines
net/netfilter/nf_tables_api.c +10 -0
+	case NFPROTO_IPV4:
+		if (sibling_table->family == NFPROTO_IPV6)
+			return false;
+		break;
+	}
+
+	return true;
+}
+
+static int nft_table_validate(struct net *net, struct nft_table *table)

This function is called from abort path too. I suspect total_jump_count for this table will not be OK in such case. And this selftest does cover many cases.

Quoted 7 lines
net/netfilter/nf_tables_api.c +2 -0
 {
 	struct nft_chain *chain;
+	struct nftables_pernet *nft_net;
 	struct nft_ctx ctx = {
 		.net	= net,
 		.family	= table->family,
+		.total_jump_count = 0,

net/netfilter/nf_tables_api.c int nft_chain_validate(const struct nft_ctx *ctx, const struct nft_chain *chain) · line 4102

Quoted 10 lines
net/netfilter/nf_tables_api.c +8 -0
 	};
 	int err;
+	u32 total_jumps_before_validate;
+	struct nft_table *sibling_table;
+
+	nft_net = nft_pernet(net);
+
+	if (!net_eq(net, &init_net)) {
+		list_for_each_entry(sibling_table, &nft_net->tables, list) {
+			if(nft_families_inc_jump(table, sibling_table))
                          ^
                  coding style
Quoted 38 lines
net/netfilter/nf_tables_api.c +17 -0
+				ctx.total_jump_count += sibling_table->total_jump_count;
+		}
+	}
+
+	total_jumps_before_validate = ctx.total_jump_count;
 
 	list_for_each_entry(chain, &table->chains, list) {
 		if (!nft_is_base_chain(chain))
@@ -4060,6 +4118,8 @@ static int nft_table_validate(struct net *net, const struct nft_table *table)
 		cond_resched();
 	}
 
+	table->total_jump_count = ctx.total_jump_count - total_jumps_before_validate;
+
 	return 0;
 }
 
@@ -4084,6 +4144,7 @@ int nft_setelem_validate(const struct nft_ctx *ctx, struct nft_set *set,
 	case NFT_JUMP:
 	case NFT_GOTO:
 		pctx->level++;
+		pctx->total_jump_count++;
 		err = nft_chain_validate(ctx, data->verdict.chain);
 		if (err < 0)
 			return err;
@@ -11916,6 +11977,59 @@ static struct notifier_block nft_nl_notifier = {
 	.notifier_call  = nft_rcv_nl_event,
 };
 
+#ifdef CONFIG_SYSCTL
+static struct ctl_table nf_limit_control_sysctl_table[] = {
+	{
+		.procname	= "nf_max_table_jumps_netns",
+		.data		= &init_net.nf.nf_max_table_jumps_netns,
+		.maxlen		= sizeof(init_net.nf.nf_max_table_jumps_netns),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec,
+	},

net/netfilter/nf_tables_api.c static struct notifier_block nft_nl_notifier = { · line 11998

Quoted 10 lines
net/netfilter/nf_tables_api.c +10 -0
+};
+
+static int netfilter_limit_control_sysctl_init(struct net *net)
+{
+	struct ctl_table *tbl;
+
+	net->nf.nf_max_table_jumps_netns = NFT_DEFAULT_MAX_TABLE_JUMPS;
+	tbl = nf_limit_control_sysctl_table;
+	if (!net_eq(net, &init_net)) {
+		tbl = kmemdup(tbl, sizeof(nf_limit_control_sysctl_table), GFP_KERNEL);

Not checking error:

                if (!tbl)
                        ...
Quoted 18 lines
net/netfilter/nf_tables_api.c +18 -0
+		tbl->data = &net->nf.nf_max_table_jumps_netns;
+		if (net->user_ns != &init_user_ns)
+			tbl->mode &= ~0222;
+	}
+
+	net->nf.nf_limit_control_dir_header = register_net_sysctl_sz(
+		net, "net/netfilter", tbl, ARRAY_SIZE(nf_limit_control_sysctl_table));
+
+	if (!net->nf.nf_limit_control_dir_header)
+		goto err_alloc;
+
+	return 0;
+
+err_alloc:
+	if (tbl != nf_limit_control_sysctl_table)
+		kfree(tbl);
+	return -ENOMEM;
+}
#6 lore raw
Quoted 6 lines · Pablo Neira Ayuso

net/netfilter/nf_tables_api.c int nft_chain_validate(const struct nft_ctx *ctx, const struct nft_chain *chain) · line 4066

Quoted 2 lines · Shaun Brady
net/netfilter/nf_tables_api.c +2 -0
+static bool nft_families_inc_jump(struct nft_table *table, struct nft_table *sibling_table)
+{

I think we mentioned about NFPROTO_BRIDGE here too.

NFPROTO_BRIDGE was indeed brought up, which brought us to the current implementation of nft_families_inc_jump.

---- excerpt ----

Quoted 2 lines

Maybe it is better to have a global limit for all tables, regardless the family, in a non-init-netns?

Looks like it would be simpler.

The only cases where processing is disjunct is ipv4 vs. ipv6.

---- /excerpt ----

The current implementation includes all tables which are not:
1) the same table we're working with (tbl == other_tbl) (self)
2) tables where the protocol is tbl.family = v4, other_tbl.family = v6 (and vice versa)

I believe from this jumps from NFPROTO_BRIDGE to high protocols will be counted.

Quoted 6 lines

Review

Quoted 2 lines
(quoted diff) +2 -0
+
+static int nft_table_validate(struct net *net, struct nft_table *table)

This function is called from abort path too. I suspect total_jump_count for this table will not be OK in such case. And this selftest does cover many cases.

The abort case is something I did not consider. Would nft_table_validate be called after a transaction rollback, and thusly a likely set of tables that previously validated successfully?

Another potential save is that I don't update nft_table.total_jump_count until after a successful validation of the table.

I might ask, how is the abort state different for validation, or is the abort state when someone literally kills an update (and thus would be a rollback?)?

Would a test case that attempts to demonstrate safety of the value suffice?

Quoted 4 lines

net/netfilter/nf_tables_api.c int nft_chain_validate(const struct nft_ctx *ctx, const struct nft_chain *chain) · line 4102

Quoted 2 lines
net/netfilter/nf_tables_api.c +1 -0
+			if(nft_families_inc_jump(table, sibling_table))
                          ^
                  coding style

I see it (or really, don't). Will fix in v6.

Quoted 8 lines

net/netfilter/nf_tables_api.c static struct notifier_block nft_nl_notifier = { · line 11998

Quoted 3 lines
net/netfilter/nf_tables_api.c +2 -0
+	if (!net_eq(net, &init_net)) {
+		tbl = kmemdup(tbl, sizeof(nf_limit_control_sysctl_table), GFP_KERNEL);

Not checking error:

                if (!tbl)
                        ...

Good catch. I think I can jump down to the same err_alloc label below.

Thanks for the feedback!

SB

#7 lore raw
In reply to Pablo Neira Ayuso
Quoted 5 lines · Pablo Neira Ayuso

This function is called from abort path too. I suspect total_jump_count for this table will not be OK in such case. And this selftest does cover many cases.

Your suspicion was correct. It took a bit to run over the abort path that used validation (I think it would just be userland sending abort), but it looked like it updated the total before the roll back.

My solution was to send in the abort state, and when action != NFNL_ABORT_VALIDATE, we don't update the totals.

Another alternative would be to just run validate again at the end of the abort (after the rollback), but that seemed wrong to do the costly validate twice.

See v6 patch (just sent) that also addresses your other feedback.

Thanks!

SB