Re: [PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

Re: [PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

[Florian Westphal]

Madhu Koriginja <madhu.koriginja@nxp.com> wrote:
> Keep the conntrack reference until policy checks have been performed for
> IPsec V6 NAT support. The reference needs to be dropped before a packet is
> queued to avoid having the conntrack module unloadable.

Subject Line should be:

[PATCH net] net: netfilter: Keep conntrack reference until IPsecv6 policy checks are done
or
[PATCH net-next] net: netfilter: Keep ..

see below why net-next makes more sense to me.

> Signed-off-by: Madhu Koriginja <madhu.koriginja@nxp.com>
> 	V1-V2: added missing () in ip6_input.c in below condition
> 	if (!(ipprot->flags & INET6_PROTO_NOPOLICY))

This should appear before your signed-off-by, or 
> ---
>  net/dccp/ipv6.c      |  1 +

... here.

I think its fine to place it here because in this case
the mini-changelog doesn't provide any additional context
worth keeping in git.

Paolo, Jakub, David: This is a bug, but its not a regression
either.  I would suggest that Madhu resubmits this AFTER
net-next re-opens.

Madhu, if thats the agreed-upon procedure, you may include

Reviewed-by: Florian Westphal <fw@strlen.de>

when you resend this patch as-is.

Re: [PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

[Florian Westphal]

Madhu Koriginja <madhu.koriginja@nxp.com> wrote:
> Keep the conntrack reference until policy checks have been performed for
> IPsec V6 NAT support. The reference needs to be dropped before a packet is
> queued to avoid having the conntrack module unloadable.

Subject Line should be:

[PATCH net] net: netfilter: Keep conntrack reference until IPsecv6 policy checks are done
or
[PATCH net-next] net: netfilter: Keep ..

see below why net-next makes more sense to me.

> Signed-off-by: Madhu Koriginja <madhu.koriginja@nxp.com>
> 	V1-V2: added missing () in ip6_input.c in below condition
> 	if (!(ipprot->flags & INET6_PROTO_NOPOLICY))

This should appear before your signed-off-by, or 
> ---
>  net/dccp/ipv6.c      |  1 +

... here.

I think its fine to place it here because in this case
the mini-changelog doesn't provide any additional context
worth keeping in git.

Paolo, Jakub, David: This is a bug, but its not a regression
either.  I would suggest that Madhu resubmits this AFTER
net-next re-opens.

Madhu, if thats the agreed-upon procedure, you may include

Reviewed-by: Florian Westphal <fw@strlen.de>

when you resend this patch as-is.

Re: [PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

[Paolo Abeni]

On Thu, 2023-03-02 at 16:53 +0530, Madhu Koriginja wrote:
> Keep the conntrack reference until policy checks have been performed for
> IPsec V6 NAT support. The reference needs to be dropped before a packet is
> queued to avoid having the conntrack module unloadable.
> 
> Signed-off-by: Madhu Koriginja <madhu.koriginja@nxp.com>
> 	V1-V2: added missing () in ip6_input.c in below condition
> 	if (!(ipprot->flags & INET6_PROTO_NOPOLICY))
> ---
>  net/dccp/ipv6.c      |  1 +
>  net/ipv6/ip6_input.c | 14 +++++++-------
>  net/ipv6/raw.c       |  2 +-
>  net/ipv6/tcp_ipv6.c  |  2 ++
>  net/ipv6/udp.c       |  2 ++
>  5 files changed, 13 insertions(+), 8 deletions(-)
> 
> diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
> index 58a401e9cf09..eb503096db6c 100644
> --- a/net/dccp/ipv6.c
> +++ b/net/dccp/ipv6.c
> @@ -771,6 +771,7 @@ static int dccp_v6_rcv(struct sk_buff *skb)
>  
>  	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
>  		goto discard_and_relse;
> +	nf_reset(skb);

nf_reset() is gone since commit 895b5c9f206e ("netfilter: drop bridge
nf reset from nf_reset"), you should use instead nf_reset_ct(): in the
current form the patch does not apply cleanly (nor build after manual
edit).


Cheers,

Paolo

Re: [PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

[Paolo Abeni]

On Thu, 2023-03-02 at 16:53 +0530, Madhu Koriginja wrote:
> Keep the conntrack reference until policy checks have been performed for
> IPsec V6 NAT support. The reference needs to be dropped before a packet is
> queued to avoid having the conntrack module unloadable.
> 
> Signed-off-by: Madhu Koriginja <madhu.koriginja@nxp.com>
> 	V1-V2: added missing () in ip6_input.c in below condition
> 	if (!(ipprot->flags & INET6_PROTO_NOPOLICY))
> ---
>  net/dccp/ipv6.c      |  1 +
>  net/ipv6/ip6_input.c | 14 +++++++-------
>  net/ipv6/raw.c       |  2 +-
>  net/ipv6/tcp_ipv6.c  |  2 ++
>  net/ipv6/udp.c       |  2 ++
>  5 files changed, 13 insertions(+), 8 deletions(-)
> 
> diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
> index 58a401e9cf09..eb503096db6c 100644
> --- a/net/dccp/ipv6.c
> +++ b/net/dccp/ipv6.c
> @@ -771,6 +771,7 @@ static int dccp_v6_rcv(struct sk_buff *skb)
>  
>  	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
>  		goto discard_and_relse;
> +	nf_reset(skb);

nf_reset() is gone since commit 895b5c9f206e ("netfilter: drop bridge
nf reset from nf_reset"), you should use instead nf_reset_ct(): in the
current form the patch does not apply cleanly (nor build after manual
edit).


Cheers,

Paolo

[PATCH] net : netfilter: Keep conntrack reference until IPsecv6 policy checks are done

[Madhu Koriginja]

Keep the conntrack reference until policy checks have been performed for
IPsec V6 NAT support. The reference needs to be dropped before a packet is
queued to avoid having the conntrack module unloadable.

Signed-off-by: Madhu Koriginja <madhu.koriginja@nxp.com>
---
 net/dccp/ipv6.c      |  1 +
 net/ipv6/ip6_input.c | 15 ++++++---------
 net/ipv6/raw.c       |  2 +-
 net/ipv6/tcp_ipv6.c  |  2 ++
 net/ipv6/udp.c       |  2 ++
 5 files changed, 12 insertions(+), 10 deletions(-)

diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
index b9d7c3dd1cb3..c0fd8f5f3b94 100644
--- a/net/dccp/ipv6.c
+++ b/net/dccp/ipv6.c
@@ -783,6 +783,7 @@ static int dccp_v6_rcv(struct sk_buff *skb)
 
 	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
 		goto discard_and_relse;
+	nf_reset_ct(skb);
 
 	return __sk_receive_skb(sk, skb, 1, dh->dccph_doff * 4,
 				refcounted) ? -1 : 0;
diff --git a/net/ipv6/ip6_input.c b/net/ipv6/ip6_input.c
index e1ebf5e42ebe..25249cd3b639 100644
--- a/net/ipv6/ip6_input.c
+++ b/net/ipv6/ip6_input.c
@@ -404,10 +404,6 @@ void ip6_protocol_deliver_rcu(struct net *net, struct sk_buff *skb, int nexthdr,
 			/* Only do this once for first final protocol */
 			have_final = true;
 
-			/* Free reference early: we don't need it any more,
-			   and it may hold ip_conntrack module loaded
-			   indefinitely. */
-			nf_reset_ct(skb);
 
 			skb_postpull_rcsum(skb, skb_network_header(skb),
 					   skb_network_header_len(skb));
@@ -430,12 +426,13 @@ void ip6_protocol_deliver_rcu(struct net *net, struct sk_buff *skb, int nexthdr,
 				goto discard;
 			}
 		}
-		if (!(ipprot->flags & INET6_PROTO_NOPOLICY) &&
-		    !xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb)) {
-			SKB_DR_SET(reason, XFRM_POLICY);
-			goto discard;
+		if (!(ipprot->flags & INET6_PROTO_NOPOLICY)) {
+			if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb)) {
+				SKB_DR_SET(reason, XFRM_POLICY);
+				goto discard;
+			}
+			nf_reset_ct(skb);
 		}
-
 		ret = INDIRECT_CALL_2(ipprot->handler, tcp_v6_rcv, udpv6_rcv,
 				      skb);
 		if (ret > 0) {
diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c
index bac9ba747bde..b035e049fb65 100644
--- a/net/ipv6/raw.c
+++ b/net/ipv6/raw.c
@@ -195,7 +195,6 @@ static bool ipv6_raw_deliver(struct sk_buff *skb, int nexthdr)
 
 			/* Not releasing hash table! */
 			if (clone) {
-				nf_reset_ct(clone);
 				rawv6_rcv(sk, clone);
 			}
 		}
@@ -391,6 +390,7 @@ int rawv6_rcv(struct sock *sk, struct sk_buff *skb)
 		kfree_skb_reason(skb, SKB_DROP_REASON_XFRM_POLICY);
 		return NET_RX_DROP;
 	}
+	nf_reset_ct(skb);
 
 	if (!rp->checksum)
 		skb->ip_summed = CHECKSUM_UNNECESSARY;
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 1bf93b61aa06..1e747241c771 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1722,6 +1722,8 @@ INDIRECT_CALLABLE_SCOPE int tcp_v6_rcv(struct sk_buff *skb)
 	if (drop_reason)
 		goto discard_and_relse;
 
+	nf_reset_ct(skb);
+
 	if (tcp_filter(sk, skb)) {
 		drop_reason = SKB_DROP_REASON_SOCKET_FILTER;
 		goto discard_and_relse;
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 9fb2f33ee3a7..75900564f42a 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -704,6 +704,7 @@ static int udpv6_queue_rcv_one_skb(struct sock *sk, struct sk_buff *skb)
 		drop_reason = SKB_DROP_REASON_XFRM_POLICY;
 		goto drop;
 	}
+	nf_reset_ct(skb);
 
 	if (static_branch_unlikely(&udpv6_encap_needed_key) && up->encap_type) {
 		int (*encap_rcv)(struct sock *sk, struct sk_buff *skb);
@@ -1027,6 +1028,7 @@ int __udp6_lib_rcv(struct sk_buff *skb, struct udp_table *udptable,
 
 	if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
 		goto discard;
+	nf_reset_ct(skb);
 
 	if (udp_lib_checksum_complete(skb))
 		goto csum_error;
-- 
2.25.1

RE: [EXT] Re: [PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

[Madhu Koriginja]
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RE: [EXT] Re: [PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

[Madhu Koriginja]

Hi Paolo,
 I updated the patch with nf_reset_ct change (nf_reset() replaced with nf_reset_ct) to align with the latest kernel code. 

Hi All,
I also want to push this change (with nf_reset) in 4.19 LTS kernel, please can you tell me how can I push to that version? Thanks in advance.

 Hi Florian,
In patchv3 updated the subject line as you suggested.

Thanks & Regards,
Madhu K  

-----Original Message-----
From: Paolo Abeni <pabeni@redhat.com> 
Sent: Thursday, March 2, 2023 7:58 PM
To: Madhu Koriginja <madhu.koriginja@nxp.com>; gerrit@erg.abdn.ac.uk; davem@davemloft.net; kuznet@ms2.inr.ac.ru; yoshfuji@linux-ipv6.org; edumazet@google.com; dccp@vger.kernel.org; netdev@vger.kernel.org; linux-kernel@vger.kernel.org
Cc: Vani Namala <vani.namala@nxp.com>
Subject: [EXT] Re: [PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

Caution: EXT Email

On Thu, 2023-03-02 at 16:53 +0530, Madhu Koriginja wrote:
> Keep the conntrack reference until policy checks have been performed 
> for IPsec V6 NAT support. The reference needs to be dropped before a 
> packet is queued to avoid having the conntrack module unloadable.
>
> Signed-off-by: Madhu Koriginja <madhu.koriginja@nxp.com>
>       V1-V2: added missing () in ip6_input.c in below condition
>       if (!(ipprot->flags & INET6_PROTO_NOPOLICY))
> ---
>  net/dccp/ipv6.c      |  1 +
>  net/ipv6/ip6_input.c | 14 +++++++-------
>  net/ipv6/raw.c       |  2 +-
>  net/ipv6/tcp_ipv6.c  |  2 ++
>  net/ipv6/udp.c       |  2 ++
>  5 files changed, 13 insertions(+), 8 deletions(-)
>
> diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c index 
> 58a401e9cf09..eb503096db6c 100644
> --- a/net/dccp/ipv6.c
> +++ b/net/dccp/ipv6.c
> @@ -771,6 +771,7 @@ static int dccp_v6_rcv(struct sk_buff *skb)
>
>       if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
>               goto discard_and_relse;
> +     nf_reset(skb);

nf_reset() is gone since commit 895b5c9f206e ("netfilter: drop bridge nf reset from nf_reset"), you should use instead nf_reset_ct(): in the current form the patch does not apply cleanly (nor build after manual edit).


Cheers,

Paolo

RE: [EXT] Re: [PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

[Madhu Koriginja]
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RE: [EXT] Re: [PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

[Madhu Koriginja]

Hi Florian/Paolo,
As Florian suggested I changed the subject title, due to this patch is pushed in different link. Please find the nf_reset_ct changes patch in the below link
https://lore.kernel.org/lkml/20230303094221.1501961-1-madhu.koriginja@nxp.com/T/

Please can you also suggest me how to push patch on 4.19 LTS kernel.

Thanks & Regards,
Madhu K

-----Original Message-----
From: Madhu Koriginja 
Sent: Friday, March 3, 2023 3:23 PM
To: Paolo Abeni <pabeni@redhat.com>; gerrit@erg.abdn.ac.uk; davem@davemloft.net; kuznet@ms2.inr.ac.ru; yoshfuji@linux-ipv6.org; edumazet@google.com; dccp@vger.kernel.org; netdev@vger.kernel.org; linux-kernel@vger.kernel.org
Cc: Vani Namala <vani.namala@nxp.com>
Subject: RE: [EXT] Re: [PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

Hi Paolo,
 I updated the patch with nf_reset_ct change (nf_reset() replaced with nf_reset_ct) to align with the latest kernel code. 

Hi All,
I also want to push this change (with nf_reset) in 4.19 LTS kernel, please can you tell me how can I push to that version? Thanks in advance.

 Hi Florian,
In patchv3 updated the subject line as you suggested.

Thanks & Regards,
Madhu K  

-----Original Message-----
From: Paolo Abeni <pabeni@redhat.com>
Sent: Thursday, March 2, 2023 7:58 PM
To: Madhu Koriginja <madhu.koriginja@nxp.com>; gerrit@erg.abdn.ac.uk; davem@davemloft.net; kuznet@ms2.inr.ac.ru; yoshfuji@linux-ipv6.org; edumazet@google.com; dccp@vger.kernel.org; netdev@vger.kernel.org; linux-kernel@vger.kernel.org
Cc: Vani Namala <vani.namala@nxp.com>
Subject: [EXT] Re: [PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

Caution: EXT Email

On Thu, 2023-03-02 at 16:53 +0530, Madhu Koriginja wrote:
> Keep the conntrack reference until policy checks have been performed 
> for IPsec V6 NAT support. The reference needs to be dropped before a 
> packet is queued to avoid having the conntrack module unloadable.
>
> Signed-off-by: Madhu Koriginja <madhu.koriginja@nxp.com>
>       V1-V2: added missing () in ip6_input.c in below condition
>       if (!(ipprot->flags & INET6_PROTO_NOPOLICY))
> ---
>  net/dccp/ipv6.c      |  1 +
>  net/ipv6/ip6_input.c | 14 +++++++-------
>  net/ipv6/raw.c       |  2 +-
>  net/ipv6/tcp_ipv6.c  |  2 ++
>  net/ipv6/udp.c       |  2 ++
>  5 files changed, 13 insertions(+), 8 deletions(-)
>
> diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c index 
> 58a401e9cf09..eb503096db6c 100644
> --- a/net/dccp/ipv6.c
> +++ b/net/dccp/ipv6.c
> @@ -771,6 +771,7 @@ static int dccp_v6_rcv(struct sk_buff *skb)
>
>       if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
>               goto discard_and_relse;
> +     nf_reset(skb);

nf_reset() is gone since commit 895b5c9f206e ("netfilter: drop bridge nf reset from nf_reset"), you should use instead nf_reset_ct(): in the current form the patch does not apply cleanly (nor build after manual edit).


Cheers,

Paolo

[PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

[Madhu Koriginja]

Keep the conntrack reference until policy checks have been performed for
IPsec V6 NAT support. The reference needs to be dropped before a packet is
queued to avoid having the conntrack module unloadable.

Signed-off-by: Madhu Koriginja <madhu.koriginja@nxp.com>
	V1-V2: added missing () in ip6_input.c in below condition
	if (!(ipprot->flags & INET6_PROTO_NOPOLICY))
---
 net/dccp/ipv6.c      |  1 +
 net/ipv6/ip6_input.c | 14 +++++++-------
 net/ipv6/raw.c       |  2 +-
 net/ipv6/tcp_ipv6.c  |  2 ++
 net/ipv6/udp.c       |  2 ++
 5 files changed, 13 insertions(+), 8 deletions(-)

diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
index 58a401e9cf09..eb503096db6c 100644
--- a/net/dccp/ipv6.c
+++ b/net/dccp/ipv6.c
@@ -771,6 +771,7 @@ static int dccp_v6_rcv(struct sk_buff *skb)
 
 	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
 		goto discard_and_relse;
+	nf_reset(skb);
 
 	return __sk_receive_skb(sk, skb, 1, dh->dccph_doff * 4,
 				refcounted) ? -1 : 0;
diff --git a/net/ipv6/ip6_input.c b/net/ipv6/ip6_input.c
index acf0749ee5bb..10d9c33cbdda 100644
--- a/net/ipv6/ip6_input.c
+++ b/net/ipv6/ip6_input.c
@@ -374,10 +374,6 @@ static int ip6_input_finish(struct net *net, struct sock *sk, struct sk_buff *sk
 			/* Only do this once for first final protocol */
 			have_final = true;
 
-			/* Free reference early: we don't need it any more,
-			   and it may hold ip_conntrack module loaded
-			   indefinitely. */
-			nf_reset(skb);
 
 			skb_postpull_rcsum(skb, skb_network_header(skb),
 					   skb_network_header_len(skb));
@@ -388,9 +384,13 @@ static int ip6_input_finish(struct net *net, struct sock *sk, struct sk_buff *sk
 			    !ipv6_is_mld(skb, nexthdr, skb_network_header_len(skb)))
 				goto discard;
 		}
-		if (!(ipprot->flags & INET6_PROTO_NOPOLICY) &&
-		    !xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
-			goto discard;
+
+		if (!(ipprot->flags & INET6_PROTO_NOPOLICY)) {
+			if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
+				goto discard;
+
+			nf_reset(skb);
+		}
 
 		ret = ipprot->handler(skb);
 		if (ret > 0) {
diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c
index 4856d9320b28..cf68c9418897 100644
--- a/net/ipv6/raw.c
+++ b/net/ipv6/raw.c
@@ -220,7 +220,6 @@ static bool ipv6_raw_deliver(struct sk_buff *skb, int nexthdr)
 
 			/* Not releasing hash table! */
 			if (clone) {
-				nf_reset(clone);
 				rawv6_rcv(sk, clone);
 			}
 		}
@@ -428,6 +427,7 @@ int rawv6_rcv(struct sock *sk, struct sk_buff *skb)
 		kfree_skb(skb);
 		return NET_RX_DROP;
 	}
+	nf_reset(skb);
 
 	if (!rp->checksum)
 		skb->ip_summed = CHECKSUM_UNNECESSARY;
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 9a117a79af65..0bc959cfbea4 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1534,6 +1534,8 @@ static int tcp_v6_rcv(struct sk_buff *skb)
 	if (tcp_v6_inbound_md5_hash(sk, skb))
 		goto discard_and_relse;
 
+	nf_reset(skb);
+
 	if (tcp_filter(sk, skb))
 		goto discard_and_relse;
 	th = (const struct tcphdr *)skb->data;
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 72b2e7809af6..aacb48e977cb 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -567,6 +567,7 @@ static int udpv6_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
 
 	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
 		goto drop;
+	nf_reset(skb);
 
 	if (static_branch_unlikely(&udpv6_encap_needed_key) && up->encap_type) {
 		int (*encap_rcv)(struct sock *sk, struct sk_buff *skb);
@@ -860,6 +861,7 @@ int __udp6_lib_rcv(struct sk_buff *skb, struct udp_table *udptable,
 
 	if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
 		goto discard;
+	nf_reset(skb);
 
 	if (udp_lib_checksum_complete(skb))
 		goto csum_error;
-- 
2.25.1

[PATCH] [net:netfilter]: Keep conntrack reference until IPsecv6 policy checks are done

[Madhu Koriginja]

Keep the conntrack reference until policy checks have been performed for
IPsec V6 NAT support. The reference needs to be dropped before a packet is
queued to avoid having the conntrack module unloadable.

Signed-off-by: Madhu Koriginja <madhu.koriginja@nxp.com>
	V1-V2: added missing () in ip6_input.c in below condition
	if (!(ipprot->flags & INET6_PROTO_NOPOLICY))
---
 net/dccp/ipv6.c      |  1 +
 net/ipv6/ip6_input.c | 14 +++++++-------
 net/ipv6/raw.c       |  2 +-
 net/ipv6/tcp_ipv6.c  |  2 ++
 net/ipv6/udp.c       |  2 ++
 5 files changed, 13 insertions(+), 8 deletions(-)

diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
index 58a401e9cf09..eb503096db6c 100644
--- a/net/dccp/ipv6.c
+++ b/net/dccp/ipv6.c
@@ -771,6 +771,7 @@ static int dccp_v6_rcv(struct sk_buff *skb)
 
 	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
 		goto discard_and_relse;
+	nf_reset(skb);
 
 	return __sk_receive_skb(sk, skb, 1, dh->dccph_doff * 4,
 				refcounted) ? -1 : 0;
diff --git a/net/ipv6/ip6_input.c b/net/ipv6/ip6_input.c
index acf0749ee5bb..10d9c33cbdda 100644
--- a/net/ipv6/ip6_input.c
+++ b/net/ipv6/ip6_input.c
@@ -374,10 +374,6 @@ static int ip6_input_finish(struct net *net, struct sock *sk, struct sk_buff *sk
 			/* Only do this once for first final protocol */
 			have_final = true;
 
-			/* Free reference early: we don't need it any more,
-			   and it may hold ip_conntrack module loaded
-			   indefinitely. */
-			nf_reset(skb);
 
 			skb_postpull_rcsum(skb, skb_network_header(skb),
 					   skb_network_header_len(skb));
@@ -388,9 +384,13 @@ static int ip6_input_finish(struct net *net, struct sock *sk, struct sk_buff *sk
 			    !ipv6_is_mld(skb, nexthdr, skb_network_header_len(skb)))
 				goto discard;
 		}
-		if (!(ipprot->flags & INET6_PROTO_NOPOLICY) &&
-		    !xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
-			goto discard;
+
+		if (!(ipprot->flags & INET6_PROTO_NOPOLICY)) {
+			if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
+				goto discard;
+
+			nf_reset(skb);
+		}
 
 		ret = ipprot->handler(skb);
 		if (ret > 0) {
diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c
index 4856d9320b28..cf68c9418897 100644
--- a/net/ipv6/raw.c
+++ b/net/ipv6/raw.c
@@ -220,7 +220,6 @@ static bool ipv6_raw_deliver(struct sk_buff *skb, int nexthdr)
 
 			/* Not releasing hash table! */
 			if (clone) {
-				nf_reset(clone);
 				rawv6_rcv(sk, clone);
 			}
 		}
@@ -428,6 +427,7 @@ int rawv6_rcv(struct sock *sk, struct sk_buff *skb)
 		kfree_skb(skb);
 		return NET_RX_DROP;
 	}
+	nf_reset(skb);
 
 	if (!rp->checksum)
 		skb->ip_summed = CHECKSUM_UNNECESSARY;
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 9a117a79af65..0bc959cfbea4 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1534,6 +1534,8 @@ static int tcp_v6_rcv(struct sk_buff *skb)
 	if (tcp_v6_inbound_md5_hash(sk, skb))
 		goto discard_and_relse;
 
+	nf_reset(skb);
+
 	if (tcp_filter(sk, skb))
 		goto discard_and_relse;
 	th = (const struct tcphdr *)skb->data;
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 72b2e7809af6..aacb48e977cb 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -567,6 +567,7 @@ static int udpv6_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
 
 	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
 		goto drop;
+	nf_reset(skb);
 
 	if (static_branch_unlikely(&udpv6_encap_needed_key) && up->encap_type) {
 		int (*encap_rcv)(struct sock *sk, struct sk_buff *skb);
@@ -860,6 +861,7 @@ int __udp6_lib_rcv(struct sk_buff *skb, struct udp_table *udptable,
 
 	if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
 		goto discard;
+	nf_reset(skb);
 
 	if (udp_lib_checksum_complete(skb))
 		goto csum_error;
-- 
2.25.1