[PATCH net 0/3] Netfilter fixes for net

[PATCH net 0/3] Netfilter fixes for net

[Pablo Neira Ayuso]

Hi,

The following patchset contains Netfilter fixes for net:

1) Fix bogus error report in selftests/netfilter/nft_nat.sh,
   from Hangbin Liu.

2) Initialize last and quota expressions from template when
   expr_ops::clone is called, otherwise, states are not restored
   accordingly when loading a dynamic set with elements using
   these two expressions.

Please, pull these changes from:

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

Thanks.

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

The following changes since commit 8f9850dd8d23c1290cb642ce9548a440da5771ec:

  net: phy: unlock on error in phy_probe() (2023-02-28 12:40:12 +0100)

are available in the Git repository at:

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

for you to fetch changes up to aabef97a35160461e9c576848ded737558d89055:

  netfilter: nft_quota: copy content when cloning expression (2023-03-01 17:23:23 +0100)

----------------------------------------------------------------
Hangbin Liu (1):
      selftests: nft_nat: ensuring the listening side is up before starting the client

Pablo Neira Ayuso (2):
      netfilter: nft_last: copy content when cloning expression
      netfilter: nft_quota: copy content when cloning expression

 net/netfilter/nft_last.c                     | 4 ++++
 net/netfilter/nft_quota.c                    | 6 +++++-
 tools/testing/selftests/netfilter/nft_nat.sh | 2 ++
 3 files changed, 11 insertions(+), 1 deletion(-)

[PATCH net 1/3] selftests: nft_nat: ensuring the listening side is up before starting the client

[Pablo Neira Ayuso]

From: Hangbin Liu <liuhangbin@gmail.com>

The test_local_dnat_portonly() function initiates the client-side as
soon as it sets the listening side to the background. This could lead to
a race condition where the server may not be ready to listen. To ensure
that the server-side is up and running before initiating the
client-side, a delay is introduced to the test_local_dnat_portonly()
function.

Before the fix:
  # ./nft_nat.sh
  PASS: netns routing/connectivity: ns0-rthlYrBU can reach ns1-rthlYrBU and ns2-rthlYrBU
  PASS: ping to ns1-rthlYrBU was ip NATted to ns2-rthlYrBU
  PASS: ping to ns1-rthlYrBU OK after ip nat output chain flush
  PASS: ipv6 ping to ns1-rthlYrBU was ip6 NATted to ns2-rthlYrBU
  2023/02/27 04:11:03 socat[6055] E connect(5, AF=2 10.0.1.99:2000, 16): Connection refused
  ERROR: inet port rewrite

After the fix:
  # ./nft_nat.sh
  PASS: netns routing/connectivity: ns0-9sPJV6JJ can reach ns1-9sPJV6JJ and ns2-9sPJV6JJ
  PASS: ping to ns1-9sPJV6JJ was ip NATted to ns2-9sPJV6JJ
  PASS: ping to ns1-9sPJV6JJ OK after ip nat output chain flush
  PASS: ipv6 ping to ns1-9sPJV6JJ was ip6 NATted to ns2-9sPJV6JJ
  PASS: inet port rewrite without l3 address

Fixes: 282e5f8fe907 ("netfilter: nat: really support inet nat without l3 address")
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 tools/testing/selftests/netfilter/nft_nat.sh | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/tools/testing/selftests/netfilter/nft_nat.sh b/tools/testing/selftests/netfilter/nft_nat.sh
index 924ecb3f1f73..dd40d9f6f259 100755
--- a/tools/testing/selftests/netfilter/nft_nat.sh
+++ b/tools/testing/selftests/netfilter/nft_nat.sh
@@ -404,6 +404,8 @@ EOF
 	echo SERVER-$family | ip netns exec "$ns1" timeout 5 socat -u STDIN TCP-LISTEN:2000 &
 	sc_s=$!
 
+	sleep 1
+
 	result=$(ip netns exec "$ns0" timeout 1 socat TCP:$daddr:2000 STDOUT)
 
 	if [ "$result" = "SERVER-inet" ];then
-- 
2.30.2

Re: [PATCH net 1/3] selftests: nft_nat: ensuring the listening side is up before starting the client

[patchwork-bot+netdevbpf]

Hello:

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

On Wed,  1 Mar 2023 23:20:19 +0100 you wrote:
> From: Hangbin Liu <liuhangbin@gmail.com>
> 
> The test_local_dnat_portonly() function initiates the client-side as
> soon as it sets the listening side to the background. This could lead to
> a race condition where the server may not be ready to listen. To ensure
> that the server-side is up and running before initiating the
> client-side, a delay is introduced to the test_local_dnat_portonly()
> function.
> 
> [...]

Here is the summary with links:
  - [net,1/3] selftests: nft_nat: ensuring the listening side is up before starting the client
    https://git.kernel.org/netdev/net/c/2067e7a00aa6
  - [net,2/3] netfilter: nft_last: copy content when cloning expression
    https://git.kernel.org/netdev/net/c/860e874290fb
  - [net,3/3] netfilter: nft_quota: copy content when cloning expression
    https://git.kernel.org/netdev/net/c/aabef97a3516

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

[PATCH net 2/3] netfilter: nft_last: copy content when cloning expression

[Pablo Neira Ayuso]

If the ruleset contains last timestamps, restore them accordingly.
Otherwise, listing after restoration shows never used items.

Fixes: 33a24de37e81 ("netfilter: nft_last: move stateful fields out of expression data")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nft_last.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/net/netfilter/nft_last.c b/net/netfilter/nft_last.c
index 7f2bda6641bd..8e6d7eaf9dc8 100644
--- a/net/netfilter/nft_last.c
+++ b/net/netfilter/nft_last.c
@@ -105,11 +105,15 @@ static void nft_last_destroy(const struct nft_ctx *ctx,
 static int nft_last_clone(struct nft_expr *dst, const struct nft_expr *src)
 {
 	struct nft_last_priv *priv_dst = nft_expr_priv(dst);
+	struct nft_last_priv *priv_src = nft_expr_priv(src);
 
 	priv_dst->last = kzalloc(sizeof(*priv_dst->last), GFP_ATOMIC);
 	if (!priv_dst->last)
 		return -ENOMEM;
 
+	priv_dst->last->set = priv_src->last->set;
+	priv_dst->last->jiffies = priv_src->last->jiffies;
+
 	return 0;
 }
 
-- 
2.30.2

[PATCH net 3/3] netfilter: nft_quota: copy content when cloning expression

[Pablo Neira Ayuso]

If the ruleset contains consumed quota, restore them accordingly.
Otherwise, listing after restoration shows never used items.

Restore the user-defined quota and flags too.

Fixes: ed0a0c60f0e5 ("netfilter: nft_quota: move stateful fields out of expression data")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nft_quota.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/net/netfilter/nft_quota.c b/net/netfilter/nft_quota.c
index 123578e28917..3ba12a7471b0 100644
--- a/net/netfilter/nft_quota.c
+++ b/net/netfilter/nft_quota.c
@@ -236,12 +236,16 @@ static void nft_quota_destroy(const struct nft_ctx *ctx,
 static int nft_quota_clone(struct nft_expr *dst, const struct nft_expr *src)
 {
 	struct nft_quota *priv_dst = nft_expr_priv(dst);
+	struct nft_quota *priv_src = nft_expr_priv(src);
+
+	priv_dst->quota = priv_src->quota;
+	priv_dst->flags = priv_src->flags;
 
 	priv_dst->consumed = kmalloc(sizeof(*priv_dst->consumed), GFP_ATOMIC);
 	if (!priv_dst->consumed)
 		return -ENOMEM;
 
-	atomic64_set(priv_dst->consumed, 0);
+	*priv_dst->consumed = *priv_src->consumed;
 
 	return 0;
 }
-- 
2.30.2