[PATCH net 0/7] Netfilter fixes for net

[PATCH net 0/7] Netfilter fixes for net

[Pablo Neira Ayuso]

Hi,

The following patchset contains Netfilter fixes for net:

1) Fix NAT IPv6 offload in the flowtable.

2) icmpv6 is printed as unknown in /proc/net/nf_conntrack.

3) Use div64_u64() in nft_limit, from Eric Dumazet.

4) Use pre_exit to unregister ebtables and arptables hooks,
   from Florian Westphal.

5) Fix out-of-bound memset in x_tables compat match/target,
   also from Florian.

6) Clone set elements expression to ensure proper initialization.

Please, pull these changes from:

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

Thanks!

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

The following changes since commit 9adc89af724f12a03b47099cd943ed54e877cd59:

  net: let skb_orphan_partial wake-up waiters. (2021-03-30 13:57:28 -0700)

are available in the Git repository at:

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

for you to fetch changes up to 4d8f9065830e526c83199186c5f56a6514f457d2:

  netfilter: nftables: clone set element expression template (2021-04-13 00:19:05 +0200)

----------------------------------------------------------------
Eric Dumazet (1):
      netfilter: nft_limit: avoid possible divide error in nft_limit_init

Florian Westphal (3):
      netfilter: bridge: add pre_exit hooks for ebtable unregistration
      netfilter: arp_tables: add pre_exit hook for table unregister
      netfilter: x_tables: fix compat match/target pad out-of-bound write

Pablo Neira Ayuso (3):
      netfilter: flowtable: fix NAT IPv6 offload mangling
      netfilter: conntrack: do not print icmpv6 as unknown via /proc
      netfilter: nftables: clone set element expression template

 include/linux/netfilter_arp/arp_tables.h  |  5 ++--
 include/linux/netfilter_bridge/ebtables.h |  5 ++--
 net/bridge/netfilter/ebtable_broute.c     |  8 +++++-
 net/bridge/netfilter/ebtable_filter.c     |  8 +++++-
 net/bridge/netfilter/ebtable_nat.c        |  8 +++++-
 net/bridge/netfilter/ebtables.c           | 30 ++++++++++++++++++--
 net/ipv4/netfilter/arp_tables.c           | 11 ++++++--
 net/ipv4/netfilter/arptable_filter.c      | 10 ++++++-
 net/ipv4/netfilter/ip_tables.c            |  2 ++
 net/ipv6/netfilter/ip6_tables.c           |  2 ++
 net/netfilter/nf_conntrack_standalone.c   |  1 +
 net/netfilter/nf_flow_table_offload.c     |  6 ++--
 net/netfilter/nf_tables_api.c             | 46 +++++++++++++++++++++++--------
 net/netfilter/nft_limit.c                 |  4 +--
 net/netfilter/x_tables.c                  | 10 ++-----
 15 files changed, 118 insertions(+), 38 deletions(-)

[PATCH net 1/7] netfilter: flowtable: fix NAT IPv6 offload mangling

[Pablo Neira Ayuso]

Fix out-of-bound access in the address array.

Fixes: 5c27d8d76ce8 ("netfilter: nf_flow_table_offload: add IPv6 support")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nf_flow_table_offload.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/net/netfilter/nf_flow_table_offload.c b/net/netfilter/nf_flow_table_offload.c
index 2a6993fa40d7..1c5460e7bce8 100644
--- a/net/netfilter/nf_flow_table_offload.c
+++ b/net/netfilter/nf_flow_table_offload.c
@@ -305,12 +305,12 @@ static void flow_offload_ipv6_mangle(struct nf_flow_rule *flow_rule,
 				     const __be32 *addr, const __be32 *mask)
 {
 	struct flow_action_entry *entry;
-	int i;
+	int i, j;
 
-	for (i = 0; i < sizeof(struct in6_addr) / sizeof(u32); i += sizeof(u32)) {
+	for (i = 0, j = 0; i < sizeof(struct in6_addr) / sizeof(u32); i += sizeof(u32), j++) {
 		entry = flow_action_entry_next(flow_rule);
 		flow_offload_mangle(entry, FLOW_ACT_MANGLE_HDR_TYPE_IP6,
-				    offset + i, &addr[i], mask);
+				    offset + i, &addr[j], mask);
 	}
 }
 
-- 
2.20.1

Re: [PATCH net 1/7] netfilter: flowtable: fix NAT IPv6 offload mangling

[patchwork-bot+netdevbpf]

Hello:

This series was applied to netdev/net.git (refs/heads/master):

On Tue, 13 Apr 2021 00:30:53 +0200 you wrote:
> Fix out-of-bound access in the address array.
> 
> Fixes: 5c27d8d76ce8 ("netfilter: nf_flow_table_offload: add IPv6 support")
> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
> ---
>  net/netfilter/nf_flow_table_offload.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)

Here is the summary with links:
  - [net,1/7] netfilter: flowtable: fix NAT IPv6 offload mangling
    https://git.kernel.org/netdev/net/c/0e07e25b481a
  - [net,2/7] netfilter: conntrack: do not print icmpv6 as unknown via /proc
    https://git.kernel.org/netdev/net/c/fbea31808ca1
  - [net,3/7] netfilter: nft_limit: avoid possible divide error in nft_limit_init
    https://git.kernel.org/netdev/net/c/b895bdf5d643
  - [net,4/7] netfilter: bridge: add pre_exit hooks for ebtable unregistration
    https://git.kernel.org/netdev/net/c/7ee3c61dcd28
  - [net,5/7] netfilter: arp_tables: add pre_exit hook for table unregister
    https://git.kernel.org/netdev/net/c/d163a925ebbc
  - [net,6/7] netfilter: x_tables: fix compat match/target pad out-of-bound write
    https://git.kernel.org/netdev/net/c/b29c457a6511
  - [net,7/7] netfilter: nftables: clone set element expression template
    https://git.kernel.org/netdev/net/c/4d8f9065830e

You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html

[PATCH net 2/7] netfilter: conntrack: do not print icmpv6 as unknown via /proc

[Pablo Neira Ayuso]

/proc/net/nf_conntrack shows icmpv6 as unknown.

Fixes: 09ec82f5af99 ("netfilter: conntrack: remove protocol name from l4proto struct")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nf_conntrack_standalone.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/net/netfilter/nf_conntrack_standalone.c b/net/netfilter/nf_conntrack_standalone.c
index 0ee702d374b0..c6c0cb465664 100644
--- a/net/netfilter/nf_conntrack_standalone.c
+++ b/net/netfilter/nf_conntrack_standalone.c
@@ -266,6 +266,7 @@ static const char* l4proto_name(u16 proto)
 	case IPPROTO_GRE: return "gre";
 	case IPPROTO_SCTP: return "sctp";
 	case IPPROTO_UDPLITE: return "udplite";
+	case IPPROTO_ICMPV6: return "icmpv6";
 	}
 
 	return "unknown";
-- 
2.20.1

[PATCH net 3/7] netfilter: nft_limit: avoid possible divide error in nft_limit_init

[Pablo Neira Ayuso]

From: Eric Dumazet <edumazet@google.com>

div_u64() divides u64 by u32.

nft_limit_init() wants to divide u64 by u64, use the appropriate
math function (div64_u64)

divide error: 0000 [#1] PREEMPT SMP KASAN
CPU: 1 PID: 8390 Comm: syz-executor188 Not tainted 5.12.0-rc4-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
RIP: 0010:div_u64_rem include/linux/math64.h:28 [inline]
RIP: 0010:div_u64 include/linux/math64.h:127 [inline]
RIP: 0010:nft_limit_init+0x2a2/0x5e0 net/netfilter/nft_limit.c:85
Code: ef 4c 01 eb 41 0f 92 c7 48 89 de e8 38 a5 22 fa 4d 85 ff 0f 85 97 02 00 00 e8 ea 9e 22 fa 4c 0f af f3 45 89 ed 31 d2 4c 89 f0 <49> f7 f5 49 89 c6 e8 d3 9e 22 fa 48 8d 7d 48 48 b8 00 00 00 00 00
RSP: 0018:ffffc90009447198 EFLAGS: 00010246
RAX: 0000000000000000 RBX: 0000200000000000 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffffffff875152e6 RDI: 0000000000000003
RBP: ffff888020f80908 R08: 0000200000000000 R09: 0000000000000000
R10: ffffffff875152d8 R11: 0000000000000000 R12: ffffc90009447270
R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000
FS:  000000000097a300(0000) GS:ffff8880b9d00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00000000200001c4 CR3: 0000000026a52000 CR4: 00000000001506e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 nf_tables_newexpr net/netfilter/nf_tables_api.c:2675 [inline]
 nft_expr_init+0x145/0x2d0 net/netfilter/nf_tables_api.c:2713
 nft_set_elem_expr_alloc+0x27/0x280 net/netfilter/nf_tables_api.c:5160
 nf_tables_newset+0x1997/0x3150 net/netfilter/nf_tables_api.c:4321
 nfnetlink_rcv_batch+0x85a/0x21b0 net/netfilter/nfnetlink.c:456
 nfnetlink_rcv_skb_batch net/netfilter/nfnetlink.c:580 [inline]
 nfnetlink_rcv+0x3af/0x420 net/netfilter/nfnetlink.c:598
 netlink_unicast_kernel net/netlink/af_netlink.c:1312 [inline]
 netlink_unicast+0x533/0x7d0 net/netlink/af_netlink.c:1338
 netlink_sendmsg+0x856/0xd90 net/netlink/af_netlink.c:1927
 sock_sendmsg_nosec net/socket.c:654 [inline]
 sock_sendmsg+0xcf/0x120 net/socket.c:674
 ____sys_sendmsg+0x6e8/0x810 net/socket.c:2350
 ___sys_sendmsg+0xf3/0x170 net/socket.c:2404
 __sys_sendmsg+0xe5/0x1b0 net/socket.c:2433
 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
 entry_SYSCALL_64_after_hwframe+0x44/0xae

Fixes: c26844eda9d4 ("netfilter: nf_tables: Fix nft limit burst handling")
Fixes: 3e0f64b7dd31 ("netfilter: nft_limit: fix packet ratelimiting")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Diagnosed-by: Luigi Rizzo <lrizzo@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nft_limit.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/net/netfilter/nft_limit.c b/net/netfilter/nft_limit.c
index 0e2c315c3b5e..82ec27bdf941 100644
--- a/net/netfilter/nft_limit.c
+++ b/net/netfilter/nft_limit.c
@@ -76,13 +76,13 @@ static int nft_limit_init(struct nft_limit *limit,
 		return -EOVERFLOW;
 
 	if (pkts) {
-		tokens = div_u64(limit->nsecs, limit->rate) * limit->burst;
+		tokens = div64_u64(limit->nsecs, limit->rate) * limit->burst;
 	} else {
 		/* The token bucket size limits the number of tokens can be
 		 * accumulated. tokens_max specifies the bucket size.
 		 * tokens_max = unit * (rate + burst) / rate.
 		 */
-		tokens = div_u64(limit->nsecs * (limit->rate + limit->burst),
+		tokens = div64_u64(limit->nsecs * (limit->rate + limit->burst),
 				 limit->rate);
 	}
 
-- 
2.20.1

[PATCH net 4/7] netfilter: bridge: add pre_exit hooks for ebtable unregistration

[Pablo Neira Ayuso]

From: Florian Westphal <fw@strlen.de>

Just like ip/ip6/arptables, the hooks have to be removed, then
synchronize_rcu() has to be called to make sure no more packets are being
processed before the ruleset data is released.

Place the hook unregistration in the pre_exit hook, then call the new
ebtables pre_exit function from there.

Years ago, when first netns support got added for netfilter+ebtables,
this used an older (now removed) netfilter hook unregister API, that did
a unconditional synchronize_rcu().

Now that all is done with call_rcu, ebtable_{filter,nat,broute} pernet exit
handlers may free the ebtable ruleset while packets are still in flight.

This can only happens on module removal, not during netns exit.

The new function expects the table name, not the table struct.

This is because upcoming patch set (targeting -next) will remove all
net->xt.{nat,filter,broute}_table instances, this makes it necessary
to avoid external references to those member variables.

The existing APIs will be converted, so follow the upcoming scheme of
passing name + hook type instead.

Fixes: aee12a0a3727e ("ebtables: remove nf_hook_register usage")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 include/linux/netfilter_bridge/ebtables.h |  5 ++--
 net/bridge/netfilter/ebtable_broute.c     |  8 +++++-
 net/bridge/netfilter/ebtable_filter.c     |  8 +++++-
 net/bridge/netfilter/ebtable_nat.c        |  8 +++++-
 net/bridge/netfilter/ebtables.c           | 30 ++++++++++++++++++++---
 5 files changed, 51 insertions(+), 8 deletions(-)

diff --git a/include/linux/netfilter_bridge/ebtables.h b/include/linux/netfilter_bridge/ebtables.h
index 2f5c4e6ecd8a..3a956145a25c 100644
--- a/include/linux/netfilter_bridge/ebtables.h
+++ b/include/linux/netfilter_bridge/ebtables.h
@@ -110,8 +110,9 @@ extern int ebt_register_table(struct net *net,
 			      const struct ebt_table *table,
 			      const struct nf_hook_ops *ops,
 			      struct ebt_table **res);
-extern void ebt_unregister_table(struct net *net, struct ebt_table *table,
-				 const struct nf_hook_ops *);
+extern void ebt_unregister_table(struct net *net, struct ebt_table *table);
+void ebt_unregister_table_pre_exit(struct net *net, const char *tablename,
+				   const struct nf_hook_ops *ops);
 extern unsigned int ebt_do_table(struct sk_buff *skb,
 				 const struct nf_hook_state *state,
 				 struct ebt_table *table);
diff --git a/net/bridge/netfilter/ebtable_broute.c b/net/bridge/netfilter/ebtable_broute.c
index 66e7af165494..32bc2821027f 100644
--- a/net/bridge/netfilter/ebtable_broute.c
+++ b/net/bridge/netfilter/ebtable_broute.c
@@ -105,14 +105,20 @@ static int __net_init broute_net_init(struct net *net)
 				  &net->xt.broute_table);
 }
 
+static void __net_exit broute_net_pre_exit(struct net *net)
+{
+	ebt_unregister_table_pre_exit(net, "broute", &ebt_ops_broute);
+}
+
 static void __net_exit broute_net_exit(struct net *net)
 {
-	ebt_unregister_table(net, net->xt.broute_table, &ebt_ops_broute);
+	ebt_unregister_table(net, net->xt.broute_table);
 }
 
 static struct pernet_operations broute_net_ops = {
 	.init = broute_net_init,
 	.exit = broute_net_exit,
+	.pre_exit = broute_net_pre_exit,
 };
 
 static int __init ebtable_broute_init(void)
diff --git a/net/bridge/netfilter/ebtable_filter.c b/net/bridge/netfilter/ebtable_filter.c
index 78cb9b21022d..bcf982e12f16 100644
--- a/net/bridge/netfilter/ebtable_filter.c
+++ b/net/bridge/netfilter/ebtable_filter.c
@@ -99,14 +99,20 @@ static int __net_init frame_filter_net_init(struct net *net)
 				  &net->xt.frame_filter);
 }
 
+static void __net_exit frame_filter_net_pre_exit(struct net *net)
+{
+	ebt_unregister_table_pre_exit(net, "filter", ebt_ops_filter);
+}
+
 static void __net_exit frame_filter_net_exit(struct net *net)
 {
-	ebt_unregister_table(net, net->xt.frame_filter, ebt_ops_filter);
+	ebt_unregister_table(net, net->xt.frame_filter);
 }
 
 static struct pernet_operations frame_filter_net_ops = {
 	.init = frame_filter_net_init,
 	.exit = frame_filter_net_exit,
+	.pre_exit = frame_filter_net_pre_exit,
 };
 
 static int __init ebtable_filter_init(void)
diff --git a/net/bridge/netfilter/ebtable_nat.c b/net/bridge/netfilter/ebtable_nat.c
index 0888936ef853..0d092773f816 100644
--- a/net/bridge/netfilter/ebtable_nat.c
+++ b/net/bridge/netfilter/ebtable_nat.c
@@ -99,14 +99,20 @@ static int __net_init frame_nat_net_init(struct net *net)
 				  &net->xt.frame_nat);
 }
 
+static void __net_exit frame_nat_net_pre_exit(struct net *net)
+{
+	ebt_unregister_table_pre_exit(net, "nat", ebt_ops_nat);
+}
+
 static void __net_exit frame_nat_net_exit(struct net *net)
 {
-	ebt_unregister_table(net, net->xt.frame_nat, ebt_ops_nat);
+	ebt_unregister_table(net, net->xt.frame_nat);
 }
 
 static struct pernet_operations frame_nat_net_ops = {
 	.init = frame_nat_net_init,
 	.exit = frame_nat_net_exit,
+	.pre_exit = frame_nat_net_pre_exit,
 };
 
 static int __init ebtable_nat_init(void)
diff --git a/net/bridge/netfilter/ebtables.c b/net/bridge/netfilter/ebtables.c
index ebe33b60efd6..d481ff24a150 100644
--- a/net/bridge/netfilter/ebtables.c
+++ b/net/bridge/netfilter/ebtables.c
@@ -1232,10 +1232,34 @@ int ebt_register_table(struct net *net, const struct ebt_table *input_table,
 	return ret;
 }
 
-void ebt_unregister_table(struct net *net, struct ebt_table *table,
-			  const struct nf_hook_ops *ops)
+static struct ebt_table *__ebt_find_table(struct net *net, const char *name)
+{
+	struct ebt_table *t;
+
+	mutex_lock(&ebt_mutex);
+
+	list_for_each_entry(t, &net->xt.tables[NFPROTO_BRIDGE], list) {
+		if (strcmp(t->name, name) == 0) {
+			mutex_unlock(&ebt_mutex);
+			return t;
+		}
+	}
+
+	mutex_unlock(&ebt_mutex);
+	return NULL;
+}
+
+void ebt_unregister_table_pre_exit(struct net *net, const char *name, const struct nf_hook_ops *ops)
+{
+	struct ebt_table *table = __ebt_find_table(net, name);
+
+	if (table)
+		nf_unregister_net_hooks(net, ops, hweight32(table->valid_hooks));
+}
+EXPORT_SYMBOL(ebt_unregister_table_pre_exit);
+
+void ebt_unregister_table(struct net *net, struct ebt_table *table)
 {
-	nf_unregister_net_hooks(net, ops, hweight32(table->valid_hooks));
 	__ebt_unregister_table(net, table);
 }
 
-- 
2.20.1

[PATCH net 5/7] netfilter: arp_tables: add pre_exit hook for table unregister

[Pablo Neira Ayuso]

From: Florian Westphal <fw@strlen.de>

Same problem that also existed in iptables/ip(6)tables, when
arptable_filter is removed there is no longer a wait period before the
table/ruleset is free'd.

Unregister the hook in pre_exit, then remove the table in the exit
function.
This used to work correctly because the old nf_hook_unregister API
did unconditional synchronize_net.

The per-net hook unregister function uses call_rcu instead.

Fixes: b9e69e127397 ("netfilter: xtables: don't hook tables by default")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 include/linux/netfilter_arp/arp_tables.h |  5 +++--
 net/ipv4/netfilter/arp_tables.c          |  9 +++++++--
 net/ipv4/netfilter/arptable_filter.c     | 10 +++++++++-
 3 files changed, 19 insertions(+), 5 deletions(-)

diff --git a/include/linux/netfilter_arp/arp_tables.h b/include/linux/netfilter_arp/arp_tables.h
index 7d3537c40ec9..26a13294318c 100644
--- a/include/linux/netfilter_arp/arp_tables.h
+++ b/include/linux/netfilter_arp/arp_tables.h
@@ -52,8 +52,9 @@ extern void *arpt_alloc_initial_table(const struct xt_table *);
 int arpt_register_table(struct net *net, const struct xt_table *table,
 			const struct arpt_replace *repl,
 			const struct nf_hook_ops *ops, struct xt_table **res);
-void arpt_unregister_table(struct net *net, struct xt_table *table,
-			   const struct nf_hook_ops *ops);
+void arpt_unregister_table(struct net *net, struct xt_table *table);
+void arpt_unregister_table_pre_exit(struct net *net, struct xt_table *table,
+				    const struct nf_hook_ops *ops);
 extern unsigned int arpt_do_table(struct sk_buff *skb,
 				  const struct nf_hook_state *state,
 				  struct xt_table *table);
diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c
index d1e04d2b5170..6c26533480dd 100644
--- a/net/ipv4/netfilter/arp_tables.c
+++ b/net/ipv4/netfilter/arp_tables.c
@@ -1539,10 +1539,15 @@ int arpt_register_table(struct net *net,
 	return ret;
 }
 
-void arpt_unregister_table(struct net *net, struct xt_table *table,
-			   const struct nf_hook_ops *ops)
+void arpt_unregister_table_pre_exit(struct net *net, struct xt_table *table,
+				    const struct nf_hook_ops *ops)
 {
 	nf_unregister_net_hooks(net, ops, hweight32(table->valid_hooks));
+}
+EXPORT_SYMBOL(arpt_unregister_table_pre_exit);
+
+void arpt_unregister_table(struct net *net, struct xt_table *table)
+{
 	__arpt_unregister_table(net, table);
 }
 
diff --git a/net/ipv4/netfilter/arptable_filter.c b/net/ipv4/netfilter/arptable_filter.c
index c216b9ad3bb2..6c300ba5634e 100644
--- a/net/ipv4/netfilter/arptable_filter.c
+++ b/net/ipv4/netfilter/arptable_filter.c
@@ -56,16 +56,24 @@ static int __net_init arptable_filter_table_init(struct net *net)
 	return err;
 }
 
+static void __net_exit arptable_filter_net_pre_exit(struct net *net)
+{
+	if (net->ipv4.arptable_filter)
+		arpt_unregister_table_pre_exit(net, net->ipv4.arptable_filter,
+					       arpfilter_ops);
+}
+
 static void __net_exit arptable_filter_net_exit(struct net *net)
 {
 	if (!net->ipv4.arptable_filter)
 		return;
-	arpt_unregister_table(net, net->ipv4.arptable_filter, arpfilter_ops);
+	arpt_unregister_table(net, net->ipv4.arptable_filter);
 	net->ipv4.arptable_filter = NULL;
 }
 
 static struct pernet_operations arptable_filter_net_ops = {
 	.exit = arptable_filter_net_exit,
+	.pre_exit = arptable_filter_net_pre_exit,
 };
 
 static int __init arptable_filter_init(void)
-- 
2.20.1

[PATCH net 6/7] netfilter: x_tables: fix compat match/target pad out-of-bound write

[Pablo Neira Ayuso]

From: Florian Westphal <fw@strlen.de>

xt_compat_match/target_from_user doesn't check that zeroing the area
to start of next rule won't write past end of allocated ruleset blob.

Remove this code and zero the entire blob beforehand.

Reported-by: syzbot+cfc0247ac173f597aaaa@syzkaller.appspotmail.com
Reported-by: Andy Nguyen <theflow@google.com>
Fixes: 9fa492cdc160c ("[NETFILTER]: x_tables: simplify compat API")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/ipv4/netfilter/arp_tables.c |  2 ++
 net/ipv4/netfilter/ip_tables.c  |  2 ++
 net/ipv6/netfilter/ip6_tables.c |  2 ++
 net/netfilter/x_tables.c        | 10 ++--------
 4 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c
index 6c26533480dd..d6d45d820d79 100644
--- a/net/ipv4/netfilter/arp_tables.c
+++ b/net/ipv4/netfilter/arp_tables.c
@@ -1193,6 +1193,8 @@ static int translate_compat_table(struct net *net,
 	if (!newinfo)
 		goto out_unlock;
 
+	memset(newinfo->entries, 0, size);
+
 	newinfo->number = compatr->num_entries;
 	for (i = 0; i < NF_ARP_NUMHOOKS; i++) {
 		newinfo->hook_entry[i] = compatr->hook_entry[i];
diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c
index f15bc21d7301..f77ea0dbe656 100644
--- a/net/ipv4/netfilter/ip_tables.c
+++ b/net/ipv4/netfilter/ip_tables.c
@@ -1428,6 +1428,8 @@ translate_compat_table(struct net *net,
 	if (!newinfo)
 		goto out_unlock;
 
+	memset(newinfo->entries, 0, size);
+
 	newinfo->number = compatr->num_entries;
 	for (i = 0; i < NF_INET_NUMHOOKS; i++) {
 		newinfo->hook_entry[i] = compatr->hook_entry[i];
diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c
index 2e2119bfcf13..eb2b5404806c 100644
--- a/net/ipv6/netfilter/ip6_tables.c
+++ b/net/ipv6/netfilter/ip6_tables.c
@@ -1443,6 +1443,8 @@ translate_compat_table(struct net *net,
 	if (!newinfo)
 		goto out_unlock;
 
+	memset(newinfo->entries, 0, size);
+
 	newinfo->number = compatr->num_entries;
 	for (i = 0; i < NF_INET_NUMHOOKS; i++) {
 		newinfo->hook_entry[i] = compatr->hook_entry[i];
diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c
index 6bd31a7a27fc..92e9d4ebc5e8 100644
--- a/net/netfilter/x_tables.c
+++ b/net/netfilter/x_tables.c
@@ -733,7 +733,7 @@ void xt_compat_match_from_user(struct xt_entry_match *m, void **dstptr,
 {
 	const struct xt_match *match = m->u.kernel.match;
 	struct compat_xt_entry_match *cm = (struct compat_xt_entry_match *)m;
-	int pad, off = xt_compat_match_offset(match);
+	int off = xt_compat_match_offset(match);
 	u_int16_t msize = cm->u.user.match_size;
 	char name[sizeof(m->u.user.name)];
 
@@ -743,9 +743,6 @@ void xt_compat_match_from_user(struct xt_entry_match *m, void **dstptr,
 		match->compat_from_user(m->data, cm->data);
 	else
 		memcpy(m->data, cm->data, msize - sizeof(*cm));
-	pad = XT_ALIGN(match->matchsize) - match->matchsize;
-	if (pad > 0)
-		memset(m->data + match->matchsize, 0, pad);
 
 	msize += off;
 	m->u.user.match_size = msize;
@@ -1116,7 +1113,7 @@ void xt_compat_target_from_user(struct xt_entry_target *t, void **dstptr,
 {
 	const struct xt_target *target = t->u.kernel.target;
 	struct compat_xt_entry_target *ct = (struct compat_xt_entry_target *)t;
-	int pad, off = xt_compat_target_offset(target);
+	int off = xt_compat_target_offset(target);
 	u_int16_t tsize = ct->u.user.target_size;
 	char name[sizeof(t->u.user.name)];
 
@@ -1126,9 +1123,6 @@ void xt_compat_target_from_user(struct xt_entry_target *t, void **dstptr,
 		target->compat_from_user(t->data, ct->data);
 	else
 		memcpy(t->data, ct->data, tsize - sizeof(*ct));
-	pad = XT_ALIGN(target->targetsize) - target->targetsize;
-	if (pad > 0)
-		memset(t->data + target->targetsize, 0, pad);
 
 	tsize += off;
 	t->u.user.target_size = tsize;
-- 
2.20.1

[PATCH net 7/7] netfilter: nftables: clone set element expression template

[Pablo Neira Ayuso]

memcpy() breaks when using connlimit in set elements. Use
nft_expr_clone() to initialize the connlimit expression list, otherwise
connlimit garbage collector crashes when walking on the list head copy.

[  493.064656] Workqueue: events_power_efficient nft_rhash_gc [nf_tables]
[  493.064685] RIP: 0010:find_or_evict+0x5a/0x90 [nf_conncount]
[  493.064694] Code: 2b 43 40 83 f8 01 77 0d 48 c7 c0 f5 ff ff ff 44 39 63 3c 75 df 83 6d 18 01 48 8b 43 08 48 89 de 48 8b 13 48 8b 3d ee 2f 00 00 <48> 89 42 08 48 89 10 48 b8 00 01 00 00 00 00 ad de 48 89 03 48 83
[  493.064699] RSP: 0018:ffffc90000417dc0 EFLAGS: 00010297
[  493.064704] RAX: 0000000000000000 RBX: ffff888134f38410 RCX: 0000000000000000
[  493.064708] RDX: 0000000000000000 RSI: ffff888134f38410 RDI: ffff888100060cc0
[  493.064711] RBP: ffff88812ce594a8 R08: ffff888134f38438 R09: 00000000ebb9025c
[  493.064714] R10: ffffffff8219f838 R11: 0000000000000017 R12: 0000000000000001
[  493.064718] R13: ffffffff82146740 R14: ffff888134f38410 R15: 0000000000000000
[  493.064721] FS:  0000000000000000(0000) GS:ffff88840e440000(0000) knlGS:0000000000000000
[  493.064725] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  493.064729] CR2: 0000000000000008 CR3: 00000001330aa002 CR4: 00000000001706e0
[  493.064733] Call Trace:
[  493.064737]  nf_conncount_gc_list+0x8f/0x150 [nf_conncount]
[  493.064746]  nft_rhash_gc+0x106/0x390 [nf_tables]

Reported-by: Laura Garcia Liebana <nevola@gmail.com>
Fixes: 409444522976 ("netfilter: nf_tables: add elements with stateful expressions")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nf_tables_api.c | 46 ++++++++++++++++++++++++++---------
 1 file changed, 34 insertions(+), 12 deletions(-)

diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
index f57f1a6ba96f..589d2f6978d3 100644
--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -5295,16 +5295,35 @@ int nft_set_elem_expr_clone(const struct nft_ctx *ctx, struct nft_set *set,
 	return -ENOMEM;
 }
 
-static void nft_set_elem_expr_setup(const struct nft_set_ext *ext, int i,
-				    struct nft_expr *expr_array[])
+static int nft_set_elem_expr_setup(struct nft_ctx *ctx,
+				   const struct nft_set_ext *ext,
+				   struct nft_expr *expr_array[],
+				   u32 num_exprs)
 {
 	struct nft_set_elem_expr *elem_expr = nft_set_ext_expr(ext);
-	struct nft_expr *expr = nft_setelem_expr_at(elem_expr, elem_expr->size);
+	struct nft_expr *expr;
+	int i, err;
+
+	for (i = 0; i < num_exprs; i++) {
+		expr = nft_setelem_expr_at(elem_expr, elem_expr->size);
+		err = nft_expr_clone(expr, expr_array[i]);
+		if (err < 0)
+			goto err_elem_expr_setup;
+
+		elem_expr->size += expr_array[i]->ops->size;
+		nft_expr_destroy(ctx, expr_array[i]);
+		expr_array[i] = NULL;
+	}
+
+	return 0;
+
+err_elem_expr_setup:
+	for (; i < num_exprs; i++) {
+		nft_expr_destroy(ctx, expr_array[i]);
+		expr_array[i] = NULL;
+	}
 
-	memcpy(expr, expr_array[i], expr_array[i]->ops->size);
-	elem_expr->size += expr_array[i]->ops->size;
-	kfree(expr_array[i]);
-	expr_array[i] = NULL;
+	return -ENOMEM;
 }
 
 static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set,
@@ -5556,12 +5575,15 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set,
 		*nft_set_ext_obj(ext) = obj;
 		obj->use++;
 	}
-	for (i = 0; i < num_exprs; i++)
-		nft_set_elem_expr_setup(ext, i, expr_array);
+	err = nft_set_elem_expr_setup(ctx, ext, expr_array, num_exprs);
+	if (err < 0)
+		goto err_elem_expr;
 
 	trans = nft_trans_elem_alloc(ctx, NFT_MSG_NEWSETELEM, set);
-	if (trans == NULL)
-		goto err_trans;
+	if (trans == NULL) {
+		err = -ENOMEM;
+		goto err_elem_expr;
+	}
 
 	ext->genmask = nft_genmask_cur(ctx->net) | NFT_SET_ELEM_BUSY_MASK;
 	err = set->ops->insert(ctx->net, set, &elem, &ext2);
@@ -5605,7 +5627,7 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set,
 	set->ops->remove(ctx->net, set, &elem);
 err_element_clash:
 	kfree(trans);
-err_trans:
+err_elem_expr:
 	if (obj)
 		obj->use--;
 
-- 
2.20.1