netfilter-devel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Pablo Neira Ayuso <pablo@netfilter.org>
To: netfilter-devel@vger.kernel.org
Cc: fw@strlen.de, sven.auhagen@voleatech.de
Subject: Re: [PATCH nf-next] netfilter: nf_conntrack_tcp: preserve liberal flag in tcp options
Date: Mon, 21 Mar 2022 10:44:36 +0100	[thread overview]
Message-ID: <YjhJBGffCTEtOuB9@salvia> (raw)
In-Reply-To: <20220321094205.63121-1-pablo@netfilter.org>

On Mon, Mar 21, 2022 at 10:42:05AM +0100, Pablo Neira Ayuso wrote:
> When tcp_options is called all flags are cleared.
> When the IP_CT_TCP_FLAG_BE_LIBERAL is set it should be preserved
> otherwise such connections will fail in the window check.


This patch completes 8437a6209f76 ("netfilter: nft_flow_offload: set liberal tracking mode for tcp")

I'm going to send v2 to add a wrapper function to update these flags,
so we do not accidentally reset IP_CT_TCP_FLAG_BE_LIBERAL.

> Signed-off-by: Sven Auhagen <sven.auhagen@voleatech.de>
> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
> ---
> Supersedes: https://patchwork.ozlabs.org/project/netfilter-devel/patch/20220318144939.69465-1-pablo@netfilter.org/ which is not correct, it breaks TCP FIN handling.
> 
>  net/netfilter/nf_conntrack_proto_tcp.c | 6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c
> index d1582b888c0d..d8599ab5f767 100644
> --- a/net/netfilter/nf_conntrack_proto_tcp.c
> +++ b/net/netfilter/nf_conntrack_proto_tcp.c
> @@ -342,7 +342,7 @@ static void tcp_options(const struct sk_buff *skb,
>  		return;
>  
>  	state->td_scale =
> -	state->flags = 0;
> +	state->flags &= IP_CT_TCP_FLAG_BE_LIBERAL;
>  
>  	while (length > 0) {
>  		int opcode=*ptr++;
> @@ -873,7 +873,7 @@ int nf_conntrack_tcp_packet(struct nf_conn *ct,
>  	struct nf_tcp_net *tn = nf_tcp_pernet(net);
>  	struct nf_conntrack_tuple *tuple;
>  	enum tcp_conntrack new_state, old_state;
> -	unsigned int index, *timeouts;
> +	unsigned int index, flags, *timeouts;
>  	enum ip_conntrack_dir dir;
>  	const struct tcphdr *th;
>  	struct tcphdr _tcph;
> @@ -968,8 +968,10 @@ int nf_conntrack_tcp_packet(struct nf_conn *ct,
>  			ct->proto.tcp.last_flags &= ~IP_CT_EXP_CHALLENGE_ACK;
>  			ct->proto.tcp.seen[ct->proto.tcp.last_dir].flags =
>  				ct->proto.tcp.last_flags;
> +			flags = ct->proto.tcp.seen[dir].flags;
>  			memset(&ct->proto.tcp.seen[dir], 0,
>  			       sizeof(struct ip_ct_tcp_state));
> +			ct->proto.tcp.seen[dir].flags = flags & IP_CT_TCP_FLAG_BE_LIBERAL;
>  			break;
>  		}
>  		ct->proto.tcp.last_index = index;
> -- 
> 2.30.2
> 

  reply	other threads:[~2022-03-21  9:44 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-03-21  9:42 [PATCH nf-next] netfilter: nf_conntrack_tcp: preserve liberal flag in tcp options Pablo Neira Ayuso
2022-03-21  9:44 ` Pablo Neira Ayuso [this message]
2022-03-21  9:46   ` Pablo Neira Ayuso
2022-03-21 10:05     ` Pablo Neira Ayuso

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=YjhJBGffCTEtOuB9@salvia \
    --to=pablo@netfilter.org \
    --cc=fw@strlen.de \
    --cc=netfilter-devel@vger.kernel.org \
    --cc=sven.auhagen@voleatech.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).