[PATCH net 0/8] Netfilter/IPVS fixes for net

[PATCH net 0/8] Netfilter/IPVS fixes for net

[Pablo Neira Ayuso]

Hi,

The following patchset contains Netfilter fixes for net:

1) Crash due to missing initialization of timer data in
   xt_IDLETIMER, from Juhee Kang.

2) NF_CONNTRACK_SECMARK should be bool in Kconfig, from Vegard Nossum.

3) Skip netdev events on netns removal, from Florian Westphal.

4) Add testcase to show port shadowing via UDP, also from Florian.

5) Remove pr_debug() code in ip6t_rt, this fixes a crash due to
   unsafe access to non-linear skbuff, from Xin Long.

6) Make net/ipv4/vs/debug_level read-only from non-init netns,
   from Antoine Tenart.

7) Remove bogus invocation to bash in selftests/netfilter/nft_flowtable.sh
   also from Florian.

There will be a relatively simple conflict between net-next and net
after this pull-request, as reported by Stephen Rothwell.

diff --cc net/netfilter/ipvs/ip_vs_ctl.c
index 29ec3ef63edc,cbea5a68afb5..000000000000
--- a/net/netfilter/ipvs/ip_vs_ctl.c
+++ b/net/netfilter/ipvs/ip_vs_ctl.c
@@@ -4090,11 -4096,8 +4096,13 @@@ static int __net_init ip_vs_control_net
	tbl[idx++].data = &ipvs->sysctl_conn_reuse_mode;
	tbl[idx++].data = &ipvs->sysctl_schedule_icmp;
	tbl[idx++].data = &ipvs->sysctl_ignore_tunneled;
+	ipvs->sysctl_run_estimation = 1;
+	tbl[idx++].data = &ipvs->sysctl_run_estimation;
 +#ifdef CONFIG_IP_VS_DEBUG
 +	/* Global sysctls must be ro in non-init netns */
 +	if (!net_eq(net, &init_net))
 +		tbl[idx++].mode = 0444;
 +#endif

Please, pull these changes from:

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

Thanks.

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

The following changes since commit 8d6c414cd2fb74aa6812e9bfec6178f8246c4f3a:

  net: prefer socket bound to interface when not in VRF (2021-10-07 07:27:55 -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 d9aaaf223297f6146d9d7f36caca927c92ab855a:

  netfilter: ebtables: allocate chainstack on CPU local nodes (2021-10-18 00:23:57 +0200)

----------------------------------------------------------------
Antoine Tenart (1):
      netfilter: ipvs: make global sysctl readonly in non-init netns

Davidlohr Bueso (1):
      netfilter: ebtables: allocate chainstack on CPU local nodes

Florian Westphal (3):
      netfilter: nf_tables: skip netdev events generated on netns removal
      selftests: nft_nat: add udp hole punch test case
      selftests: netfilter: remove stray bash debug line

Juhee Kang (1):
      netfilter: xt_IDLETIMER: fix panic that occurs when timer_type has garbage value

Vegard Nossum (1):
      netfilter: Kconfig: use 'default y' instead of 'm' for bool config option

Xin Long (1):
      netfilter: ip6t_rt: fix rt0_hdr parsing in rt_mt6

 net/bridge/netfilter/ebtables.c                    |   4 +-
 net/ipv6/netfilter/ip6t_rt.c                       |  48 +------
 net/netfilter/Kconfig                              |   2 +-
 net/netfilter/ipvs/ip_vs_ctl.c                     |   5 +
 net/netfilter/nft_chain_filter.c                   |   9 +-
 net/netfilter/xt_IDLETIMER.c                       |   2 +-
 tools/testing/selftests/netfilter/nft_flowtable.sh |   1 -
 tools/testing/selftests/netfilter/nft_nat.sh       | 145 +++++++++++++++++++++
 8 files changed, 164 insertions(+), 52 deletions(-)

[PATCH net 1/8] netfilter: xt_IDLETIMER: fix panic that occurs when timer_type has garbage value

[Pablo Neira Ayuso]

From: Juhee Kang <claudiajkang@gmail.com>

Currently, when the rule related to IDLETIMER is added, idletimer_tg timer
structure is initialized by kmalloc on executing idletimer_tg_create
function. However, in this process timer->timer_type is not defined to
a specific value. Thus, timer->timer_type has garbage value and it occurs
kernel panic. So, this commit fixes the panic by initializing
timer->timer_type using kzalloc instead of kmalloc.

Test commands:
    # iptables -A OUTPUT -j IDLETIMER --timeout 1 --label test
    $ cat /sys/class/xt_idletimer/timers/test
      Killed

Splat looks like:
    BUG: KASAN: user-memory-access in alarm_expires_remaining+0x49/0x70
    Read of size 8 at addr 0000002e8c7bc4c8 by task cat/917
    CPU: 12 PID: 917 Comm: cat Not tainted 5.14.0+ #3 79940a339f71eb14fc81aee1757a20d5bf13eb0e
    Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.13.0-1ubuntu1.1 04/01/2014
    Call Trace:
     dump_stack_lvl+0x6e/0x9c
     kasan_report.cold+0x112/0x117
     ? alarm_expires_remaining+0x49/0x70
     __asan_load8+0x86/0xb0
     alarm_expires_remaining+0x49/0x70
     idletimer_tg_show+0xe5/0x19b [xt_IDLETIMER 11219304af9316a21bee5ba9d58f76a6b9bccc6d]
     dev_attr_show+0x3c/0x60
     sysfs_kf_seq_show+0x11d/0x1f0
     ? device_remove_bin_file+0x20/0x20
     kernfs_seq_show+0xa4/0xb0
     seq_read_iter+0x29c/0x750
     kernfs_fop_read_iter+0x25a/0x2c0
     ? __fsnotify_parent+0x3d1/0x570
     ? iov_iter_init+0x70/0x90
     new_sync_read+0x2a7/0x3d0
     ? __x64_sys_llseek+0x230/0x230
     ? rw_verify_area+0x81/0x150
     vfs_read+0x17b/0x240
     ksys_read+0xd9/0x180
     ? vfs_write+0x460/0x460
     ? do_syscall_64+0x16/0xc0
     ? lockdep_hardirqs_on+0x79/0x120
     __x64_sys_read+0x43/0x50
     do_syscall_64+0x3b/0xc0
     entry_SYSCALL_64_after_hwframe+0x44/0xae
    RIP: 0033:0x7f0cdc819142
    Code: c0 e9 c2 fe ff ff 50 48 8d 3d 3a ca 0a 00 e8 f5 19 02 00 0f 1f 44 00 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 0f 05 <48> 3d 00 f0 ff ff 77 56 c3 0f 1f 44 00 00 48 83 ec 28 48 89 54 24
    RSP: 002b:00007fff28eee5b8 EFLAGS: 00000246 ORIG_RAX: 0000000000000000
    RAX: ffffffffffffffda RBX: 0000000000020000 RCX: 00007f0cdc819142
    RDX: 0000000000020000 RSI: 00007f0cdc032000 RDI: 0000000000000003
    RBP: 00007f0cdc032000 R08: 00007f0cdc031010 R09: 0000000000000000
    R10: 0000000000000022 R11: 0000000000000246 R12: 00005607e9ee31f0
    R13: 0000000000000003 R14: 0000000000020000 R15: 0000000000020000

Fixes: 68983a354a65 ("netfilter: xtables: Add snapshot of hardidletimer target")
Signed-off-by: Juhee Kang <claudiajkang@gmail.com>
Reviewed-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/xt_IDLETIMER.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/netfilter/xt_IDLETIMER.c b/net/netfilter/xt_IDLETIMER.c
index 7b2f359bfce4..2f7cf5ecebf4 100644
--- a/net/netfilter/xt_IDLETIMER.c
+++ b/net/netfilter/xt_IDLETIMER.c
@@ -137,7 +137,7 @@ static int idletimer_tg_create(struct idletimer_tg_info *info)
 {
 	int ret;
 
-	info->timer = kmalloc(sizeof(*info->timer), GFP_KERNEL);
+	info->timer = kzalloc(sizeof(*info->timer), GFP_KERNEL);
 	if (!info->timer) {
 		ret = -ENOMEM;
 		goto out;
-- 
2.30.2

Re: [PATCH net 1/8] netfilter: xt_IDLETIMER: fix panic that occurs when timer_type has garbage value

[patchwork-bot+netdevbpf]

Hello:

This series was applied to netdev/net.git (master)
by Pablo Neira Ayuso <pablo@netfilter.org>:

On Thu, 21 Oct 2021 12:08:14 +0200 you wrote:
> From: Juhee Kang <claudiajkang@gmail.com>
> 
> Currently, when the rule related to IDLETIMER is added, idletimer_tg timer
> structure is initialized by kmalloc on executing idletimer_tg_create
> function. However, in this process timer->timer_type is not defined to
> a specific value. Thus, timer->timer_type has garbage value and it occurs
> kernel panic. So, this commit fixes the panic by initializing
> timer->timer_type using kzalloc instead of kmalloc.
> 
> [...]

Here is the summary with links:
  - [net,1/8] netfilter: xt_IDLETIMER: fix panic that occurs when timer_type has garbage value
    https://git.kernel.org/netdev/net/c/902c0b188752
  - [net,2/8] netfilter: Kconfig: use 'default y' instead of 'm' for bool config option
    https://git.kernel.org/netdev/net/c/77076934afdc
  - [net,3/8] netfilter: nf_tables: skip netdev events generated on netns removal
    https://git.kernel.org/netdev/net/c/68a3765c659f
  - [net,4/8] selftests: nft_nat: add udp hole punch test case
    https://git.kernel.org/netdev/net/c/465f15a6d1a8
  - [net,5/8] netfilter: ip6t_rt: fix rt0_hdr parsing in rt_mt6
    https://git.kernel.org/netdev/net/c/a482c5e00a9b
  - [net,6/8] netfilter: ipvs: make global sysctl readonly in non-init netns
    https://git.kernel.org/netdev/net/c/174c37627894
  - [net,7/8] selftests: netfilter: remove stray bash debug line
    https://git.kernel.org/netdev/net/c/3e6ed7703dae
  - [net,8/8] netfilter: ebtables: allocate chainstack on CPU local nodes
    https://git.kernel.org/netdev/net/c/d9aaaf223297

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

[PATCH net 2/8] netfilter: Kconfig: use 'default y' instead of 'm' for bool config option

[Pablo Neira Ayuso]

From: Vegard Nossum <vegard.nossum@gmail.com>

This option, NF_CONNTRACK_SECMARK, is a bool, so it can never be 'm'.

Fixes: 33b8e77605620 ("[NETFILTER]: Add CONFIG_NETFILTER_ADVANCED option")
Signed-off-by: Vegard Nossum <vegard.nossum@oracle.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/netfilter/Kconfig b/net/netfilter/Kconfig
index 54395266339d..92a747896f80 100644
--- a/net/netfilter/Kconfig
+++ b/net/netfilter/Kconfig
@@ -109,7 +109,7 @@ config NF_CONNTRACK_MARK
 config NF_CONNTRACK_SECMARK
 	bool  'Connection tracking security mark support'
 	depends on NETWORK_SECMARK
-	default m if NETFILTER_ADVANCED=n
+	default y if NETFILTER_ADVANCED=n
 	help
 	  This option enables security markings to be applied to
 	  connections.  Typically they are copied to connections from
-- 
2.30.2

[PATCH net 3/8] netfilter: nf_tables: skip netdev events generated on netns removal

[Pablo Neira Ayuso]

From: Florian Westphal <fw@strlen.de>

syzbot reported following (harmless) WARN:

 WARNING: CPU: 1 PID: 2648 at net/netfilter/core.c:468
  nft_netdev_unregister_hooks net/netfilter/nf_tables_api.c:230 [inline]
  nf_tables_unregister_hook include/net/netfilter/nf_tables.h:1090 [inline]
  __nft_release_basechain+0x138/0x640 net/netfilter/nf_tables_api.c:9524
  nft_netdev_event net/netfilter/nft_chain_filter.c:351 [inline]
  nf_tables_netdev_event+0x521/0x8a0 net/netfilter/nft_chain_filter.c:382

reproducer:
unshare -n bash -c 'ip link add br0 type bridge; nft add table netdev t ; \
 nft add chain netdev t ingress \{ type filter hook ingress device "br0" \
 priority 0\; policy drop\; \}'

Problem is that when netns device exit hooks create the UNREGISTER
event, the .pre_exit hook for nf_tables core has already removed the
base hook.  Notifier attempts to do this again.

The need to do base hook unregister unconditionally was needed in the past,
because notifier was last stage where reg->dev dereference was safe.

Now that nf_tables does the hook removal in .pre_exit, this isn't
needed anymore.

Reported-and-tested-by: syzbot+154bd5be532a63aa778b@syzkaller.appspotmail.com
Fixes: 767d1216bff825 ("netfilter: nftables: fix possible UAF over chains from packet path in netns")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nft_chain_filter.c | 9 +++------
 1 file changed, 3 insertions(+), 6 deletions(-)

diff --git a/net/netfilter/nft_chain_filter.c b/net/netfilter/nft_chain_filter.c
index 5b02408a920b..3ced0eb6b7c3 100644
--- a/net/netfilter/nft_chain_filter.c
+++ b/net/netfilter/nft_chain_filter.c
@@ -342,12 +342,6 @@ static void nft_netdev_event(unsigned long event, struct net_device *dev,
 		return;
 	}
 
-	/* UNREGISTER events are also happening on netns exit.
-	 *
-	 * Although nf_tables core releases all tables/chains, only this event
-	 * handler provides guarantee that hook->ops.dev is still accessible,
-	 * so we cannot skip exiting net namespaces.
-	 */
 	__nft_release_basechain(ctx);
 }
 
@@ -366,6 +360,9 @@ static int nf_tables_netdev_event(struct notifier_block *this,
 	    event != NETDEV_CHANGENAME)
 		return NOTIFY_DONE;
 
+	if (!check_net(ctx.net))
+		return NOTIFY_DONE;
+
 	nft_net = nft_pernet(ctx.net);
 	mutex_lock(&nft_net->commit_mutex);
 	list_for_each_entry(table, &nft_net->tables, list) {
-- 
2.30.2

[PATCH net 4/8] selftests: nft_nat: add udp hole punch test case

[Pablo Neira Ayuso]

From: Florian Westphal <fw@strlen.de>

Add a test case that demonstrates port shadowing via UDP.

ns2 sends packet to ns1, from source port used by a udp service on the
router, ns0.  Then, ns1 sends packet to ns0:service, but that ends up getting
forwarded to ns2.

Also add three test cases that demonstrate mitigations:
1. disable use of $port as source from 'unstrusted' origin
2. make the service untracked.  This prevents masquerade entries
   from having any effects.
3. add forced PAT via 'random' mode to translate the "wrong" sport
   into an acceptable range.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 tools/testing/selftests/netfilter/nft_nat.sh | 145 +++++++++++++++++++
 1 file changed, 145 insertions(+)

diff --git a/tools/testing/selftests/netfilter/nft_nat.sh b/tools/testing/selftests/netfilter/nft_nat.sh
index d7e07f4c3d7f..da1c1e4b6c86 100755
--- a/tools/testing/selftests/netfilter/nft_nat.sh
+++ b/tools/testing/selftests/netfilter/nft_nat.sh
@@ -741,6 +741,149 @@ EOF
 	return $lret
 }
 
+# test port shadowing.
+# create two listening services, one on router (ns0), one
+# on client (ns2), which is masqueraded from ns1 point of view.
+# ns2 sends udp packet coming from service port to ns1, on a highport.
+# Later, if n1 uses same highport to connect to ns0:service, packet
+# might be port-forwarded to ns2 instead.
+
+# second argument tells if we expect the 'fake-entry' to take effect
+# (CLIENT) or not (ROUTER).
+test_port_shadow()
+{
+	local test=$1
+	local expect=$2
+	local daddrc="10.0.1.99"
+	local daddrs="10.0.1.1"
+	local result=""
+	local logmsg=""
+
+	echo ROUTER | ip netns exec "$ns0" nc -w 5 -u -l -p 1405 >/dev/null 2>&1 &
+	nc_r=$!
+
+	echo CLIENT | ip netns exec "$ns2" nc -w 5 -u -l -p 1405 >/dev/null 2>&1 &
+	nc_c=$!
+
+	# make shadow entry, from client (ns2), going to (ns1), port 41404, sport 1405.
+	echo "fake-entry" | ip netns exec "$ns2" nc -w 1 -p 1405 -u "$daddrc" 41404 > /dev/null
+
+	# ns1 tries to connect to ns0:1405.  With default settings this should connect
+	# to client, it matches the conntrack entry created above.
+
+	result=$(echo "" | ip netns exec "$ns1" nc -w 1 -p 41404 -u "$daddrs" 1405)
+
+	if [ "$result" = "$expect" ] ;then
+		echo "PASS: portshadow test $test: got reply from ${expect}${logmsg}"
+	else
+		echo "ERROR: portshadow test $test: got reply from \"$result\", not $expect as intended"
+		ret=1
+	fi
+
+	kill $nc_r $nc_c 2>/dev/null
+
+	# flush udp entries for next test round, if any
+	ip netns exec "$ns0" conntrack -F >/dev/null 2>&1
+}
+
+# This prevents port shadow of router service via packet filter,
+# packets claiming to originate from service port from internal
+# network are dropped.
+test_port_shadow_filter()
+{
+	local family=$1
+
+ip netns exec "$ns0" nft -f /dev/stdin <<EOF
+table $family filter {
+	chain forward {
+		type filter hook forward priority 0; policy accept;
+		meta iif veth1 udp sport 1405 drop
+	}
+}
+EOF
+	test_port_shadow "port-filter" "ROUTER"
+
+	ip netns exec "$ns0" nft delete table $family filter
+}
+
+# This prevents port shadow of router service via notrack.
+test_port_shadow_notrack()
+{
+	local family=$1
+
+ip netns exec "$ns0" nft -f /dev/stdin <<EOF
+table $family raw {
+	chain prerouting {
+		type filter hook prerouting priority -300; policy accept;
+		meta iif veth0 udp dport 1405 notrack
+		udp dport 1405 notrack
+	}
+	chain output {
+		type filter hook output priority -300; policy accept;
+		udp sport 1405 notrack
+	}
+}
+EOF
+	test_port_shadow "port-notrack" "ROUTER"
+
+	ip netns exec "$ns0" nft delete table $family raw
+}
+
+# This prevents port shadow of router service via sport remap.
+test_port_shadow_pat()
+{
+	local family=$1
+
+ip netns exec "$ns0" nft -f /dev/stdin <<EOF
+table $family pat {
+	chain postrouting {
+		type nat hook postrouting priority -1; policy accept;
+		meta iif veth1 udp sport <= 1405 masquerade to : 1406-65535 random
+	}
+}
+EOF
+	test_port_shadow "pat" "ROUTER"
+
+	ip netns exec "$ns0" nft delete table $family pat
+}
+
+test_port_shadowing()
+{
+	local family="ip"
+
+	ip netns exec "$ns0" sysctl net.ipv4.conf.veth0.forwarding=1 > /dev/null
+	ip netns exec "$ns0" sysctl net.ipv4.conf.veth1.forwarding=1 > /dev/null
+
+	ip netns exec "$ns0" nft -f /dev/stdin <<EOF
+table $family nat {
+	chain postrouting {
+		type nat hook postrouting priority 0; policy accept;
+		meta oif veth0 masquerade
+	}
+}
+EOF
+	if [ $? -ne 0 ]; then
+		echo "SKIP: Could not add add $family masquerade hook"
+		return $ksft_skip
+	fi
+
+	# test default behaviour. Packet from ns1 to ns0 is redirected to ns2.
+	test_port_shadow "default" "CLIENT"
+
+	# test packet filter based mitigation: prevent forwarding of
+	# packets claiming to come from the service port.
+	test_port_shadow_filter "$family"
+
+	# test conntrack based mitigation: connections going or coming
+	# from router:service bypass connection tracking.
+	test_port_shadow_notrack "$family"
+
+	# test nat based mitigation: fowarded packets coming from service port
+	# are masqueraded with random highport.
+	test_port_shadow_pat "$family"
+
+	ip netns exec "$ns0" nft delete table $family nat
+}
 
 # ip netns exec "$ns0" ping -c 1 -q 10.0.$i.99
 for i in 0 1 2; do
@@ -861,6 +1004,8 @@ reset_counters
 $test_inet_nat && test_redirect inet
 $test_inet_nat && test_redirect6 inet
 
+test_port_shadowing
+
 if [ $ret -ne 0 ];then
 	echo -n "FAIL: "
 	nft --version
-- 
2.30.2

[PATCH net 5/8] netfilter: ip6t_rt: fix rt0_hdr parsing in rt_mt6

[Pablo Neira Ayuso]

From: Xin Long <lucien.xin@gmail.com>

In rt_mt6(), when it's a nonlinear skb, the 1st skb_header_pointer()
only copies sizeof(struct ipv6_rt_hdr) to _route that rh points to.
The access by ((const struct rt0_hdr *)rh)->reserved will overflow
the buffer. So this access should be moved below the 2nd call to
skb_header_pointer().

Besides, after the 2nd skb_header_pointer(), its return value should
also be checked, othersize, *rp may cause null-pointer-ref.

v1->v2:
  - clean up some old debugging log.

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/ipv6/netfilter/ip6t_rt.c | 48 +++++-------------------------------
 1 file changed, 6 insertions(+), 42 deletions(-)

diff --git a/net/ipv6/netfilter/ip6t_rt.c b/net/ipv6/netfilter/ip6t_rt.c
index 733c83d38b30..4ad8b2032f1f 100644
--- a/net/ipv6/netfilter/ip6t_rt.c
+++ b/net/ipv6/netfilter/ip6t_rt.c
@@ -25,12 +25,7 @@ MODULE_AUTHOR("Andras Kis-Szabo <kisza@sch.bme.hu>");
 static inline bool
 segsleft_match(u_int32_t min, u_int32_t max, u_int32_t id, bool invert)
 {
-	bool r;
-	pr_debug("segsleft_match:%c 0x%x <= 0x%x <= 0x%x\n",
-		 invert ? '!' : ' ', min, id, max);
-	r = (id >= min && id <= max) ^ invert;
-	pr_debug(" result %s\n", r ? "PASS" : "FAILED");
-	return r;
+	return (id >= min && id <= max) ^ invert;
 }
 
 static bool rt_mt6(const struct sk_buff *skb, struct xt_action_param *par)
@@ -65,30 +60,6 @@ static bool rt_mt6(const struct sk_buff *skb, struct xt_action_param *par)
 		return false;
 	}
 
-	pr_debug("IPv6 RT LEN %u %u ", hdrlen, rh->hdrlen);
-	pr_debug("TYPE %04X ", rh->type);
-	pr_debug("SGS_LEFT %u %02X\n", rh->segments_left, rh->segments_left);
-
-	pr_debug("IPv6 RT segsleft %02X ",
-		 segsleft_match(rtinfo->segsleft[0], rtinfo->segsleft[1],
-				rh->segments_left,
-				!!(rtinfo->invflags & IP6T_RT_INV_SGS)));
-	pr_debug("type %02X %02X %02X ",
-		 rtinfo->rt_type, rh->type,
-		 (!(rtinfo->flags & IP6T_RT_TYP) ||
-		  ((rtinfo->rt_type == rh->type) ^
-		   !!(rtinfo->invflags & IP6T_RT_INV_TYP))));
-	pr_debug("len %02X %04X %02X ",
-		 rtinfo->hdrlen, hdrlen,
-		 !(rtinfo->flags & IP6T_RT_LEN) ||
-		  ((rtinfo->hdrlen == hdrlen) ^
-		   !!(rtinfo->invflags & IP6T_RT_INV_LEN)));
-	pr_debug("res %02X %02X %02X ",
-		 rtinfo->flags & IP6T_RT_RES,
-		 ((const struct rt0_hdr *)rh)->reserved,
-		 !((rtinfo->flags & IP6T_RT_RES) &&
-		   (((const struct rt0_hdr *)rh)->reserved)));
-
 	ret = (segsleft_match(rtinfo->segsleft[0], rtinfo->segsleft[1],
 			      rh->segments_left,
 			      !!(rtinfo->invflags & IP6T_RT_INV_SGS))) &&
@@ -107,22 +78,22 @@ static bool rt_mt6(const struct sk_buff *skb, struct xt_action_param *par)
 						       reserved),
 					sizeof(_reserved),
 					&_reserved);
+		if (!rp) {
+			par->hotdrop = true;
+			return false;
+		}
 
 		ret = (*rp == 0);
 	}
 
-	pr_debug("#%d ", rtinfo->addrnr);
 	if (!(rtinfo->flags & IP6T_RT_FST)) {
 		return ret;
 	} else if (rtinfo->flags & IP6T_RT_FST_NSTRICT) {
-		pr_debug("Not strict ");
 		if (rtinfo->addrnr > (unsigned int)((hdrlen - 8) / 16)) {
-			pr_debug("There isn't enough space\n");
 			return false;
 		} else {
 			unsigned int i = 0;
 
-			pr_debug("#%d ", rtinfo->addrnr);
 			for (temp = 0;
 			     temp < (unsigned int)((hdrlen - 8) / 16);
 			     temp++) {
@@ -138,26 +109,20 @@ static bool rt_mt6(const struct sk_buff *skb, struct xt_action_param *par)
 					return false;
 				}
 
-				if (ipv6_addr_equal(ap, &rtinfo->addrs[i])) {
-					pr_debug("i=%d temp=%d;\n", i, temp);
+				if (ipv6_addr_equal(ap, &rtinfo->addrs[i]))
 					i++;
-				}
 				if (i == rtinfo->addrnr)
 					break;
 			}
-			pr_debug("i=%d #%d\n", i, rtinfo->addrnr);
 			if (i == rtinfo->addrnr)
 				return ret;
 			else
 				return false;
 		}
 	} else {
-		pr_debug("Strict ");
 		if (rtinfo->addrnr > (unsigned int)((hdrlen - 8) / 16)) {
-			pr_debug("There isn't enough space\n");
 			return false;
 		} else {
-			pr_debug("#%d ", rtinfo->addrnr);
 			for (temp = 0; temp < rtinfo->addrnr; temp++) {
 				ap = skb_header_pointer(skb,
 							ptr
@@ -173,7 +138,6 @@ static bool rt_mt6(const struct sk_buff *skb, struct xt_action_param *par)
 				if (!ipv6_addr_equal(ap, &rtinfo->addrs[temp]))
 					break;
 			}
-			pr_debug("temp=%d #%d\n", temp, rtinfo->addrnr);
 			if (temp == rtinfo->addrnr &&
 			    temp == (unsigned int)((hdrlen - 8) / 16))
 				return ret;
-- 
2.30.2

[PATCH net 6/8] netfilter: ipvs: make global sysctl readonly in non-init netns

[Pablo Neira Ayuso]

From: Antoine Tenart <atenart@kernel.org>

Because the data pointer of net/ipv4/vs/debug_level is not updated per
netns, it must be marked as read-only in non-init netns.

Fixes: c6d2d445d8de ("IPVS: netns, final patch enabling network name space.")
Signed-off-by: Antoine Tenart <atenart@kernel.org>
Acked-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/ipvs/ip_vs_ctl.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c
index c25097092a06..29ec3ef63edc 100644
--- a/net/netfilter/ipvs/ip_vs_ctl.c
+++ b/net/netfilter/ipvs/ip_vs_ctl.c
@@ -4090,6 +4090,11 @@ static int __net_init ip_vs_control_net_init_sysctl(struct netns_ipvs *ipvs)
 	tbl[idx++].data = &ipvs->sysctl_conn_reuse_mode;
 	tbl[idx++].data = &ipvs->sysctl_schedule_icmp;
 	tbl[idx++].data = &ipvs->sysctl_ignore_tunneled;
+#ifdef CONFIG_IP_VS_DEBUG
+	/* Global sysctls must be ro in non-init netns */
+	if (!net_eq(net, &init_net))
+		tbl[idx++].mode = 0444;
+#endif
 
 	ipvs->sysctl_hdr = register_net_sysctl(net, "net/ipv4/vs", tbl);
 	if (ipvs->sysctl_hdr == NULL) {
-- 
2.30.2

[PATCH net 7/8] selftests: netfilter: remove stray bash debug line

[Pablo Neira Ayuso]

From: Florian Westphal <fw@strlen.de>

This should not be there.

Fixes: 2de03b45236f ("selftests: netfilter: add flowtable test script")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 tools/testing/selftests/netfilter/nft_flowtable.sh | 1 -
 1 file changed, 1 deletion(-)

diff --git a/tools/testing/selftests/netfilter/nft_flowtable.sh b/tools/testing/selftests/netfilter/nft_flowtable.sh
index 427d94816f2d..d4ffebb989f8 100755
--- a/tools/testing/selftests/netfilter/nft_flowtable.sh
+++ b/tools/testing/selftests/netfilter/nft_flowtable.sh
@@ -199,7 +199,6 @@ fi
 # test basic connectivity
 if ! ip netns exec ns1 ping -c 1 -q 10.0.2.99 > /dev/null; then
   echo "ERROR: ns1 cannot reach ns2" 1>&2
-  bash
   exit 1
 fi
 
-- 
2.30.2

[PATCH net 8/8] netfilter: ebtables: allocate chainstack on CPU local nodes

[Pablo Neira Ayuso]

From: Davidlohr Bueso <dave@stgolabs.net>

Keep the per-CPU memory allocated for chainstacks local.

Signed-off-by: Davidlohr Bueso <dbueso@suse.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/bridge/netfilter/ebtables.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/net/bridge/netfilter/ebtables.c b/net/bridge/netfilter/ebtables.c
index 83d1798dfbb4..ba045f35114d 100644
--- a/net/bridge/netfilter/ebtables.c
+++ b/net/bridge/netfilter/ebtables.c
@@ -926,7 +926,9 @@ static int translate_table(struct net *net, const char *name,
 			return -ENOMEM;
 		for_each_possible_cpu(i) {
 			newinfo->chainstack[i] =
-			  vmalloc(array_size(udc_cnt, sizeof(*(newinfo->chainstack[0]))));
+			  vmalloc_node(array_size(udc_cnt,
+					  sizeof(*(newinfo->chainstack[0]))),
+				       cpu_to_node(i));
 			if (!newinfo->chainstack[i]) {
 				while (i)
 					vfree(newinfo->chainstack[--i]);
-- 
2.30.2