netdev - Re: [PATCH net 1/1] netfilter: conntrack: Check offload bit on table dump

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20210130120114.GA7846@salvia>
Date:   Sat, 30 Jan 2021 13:01:14 +0100
From:   Pablo Neira Ayuso <pablo@...filter.org>
To:     Roi Dayan <roid@...dia.com>
Cc:     netdev@...r.kernel.org, Paul Blakey <paulb@...dia.com>,
        Oz Shlomo <ozsh@...dia.com>
Subject: Re: [PATCH net 1/1] netfilter: conntrack: Check offload bit on table
 dump

Hi Roi,

On Thu, Jan 28, 2021 at 09:40:52AM +0200, Roi Dayan wrote:
> Currently, offloaded flows might be deleted when executing conntrack -L
> or cat /proc/net/nf_conntrack while rules being offloaded.
> Ct timeout is not maintained for offloaded flows as aging
> of offloaded flows are managed by the flow table offload infrastructure.
> 
> Don't do garbage collection for offloaded flows when dumping the
> entries.
> 
> Fixes: 90964016e5d3 ("netfilter: nf_conntrack: add IPS_OFFLOAD status bit")
> Signed-off-by: Roi Dayan <roid@...dia.com>
> Reviewed-by: Oz Shlomo <ozsh@...dia.com>
> ---
>  include/net/netfilter/nf_conntrack.h | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/include/net/netfilter/nf_conntrack.h b/include/net/netfilter/nf_conntrack.h
> index 439379ca9ffa..87c85109946a 100644
> --- a/include/net/netfilter/nf_conntrack.h
> +++ b/include/net/netfilter/nf_conntrack.h
> @@ -276,7 +276,7 @@ static inline bool nf_ct_is_expired(const struct nf_conn *ct)
>  static inline bool nf_ct_should_gc(const struct nf_conn *ct)
>  {
>  	return nf_ct_is_expired(ct) && nf_ct_is_confirmed(ct) &&
> -	       !nf_ct_is_dying(ct);
> +	       !nf_ct_is_dying(ct) && !test_bit(IPS_OFFLOAD_BIT, &ct->status);

The gc_worker() calls nf_ct_offload_timeout() if the flow if
offloaded, so it extends the timeout to skip the garbage collection.

Could you update ctnetlink_dump_table() and ct_seq_show() to extend
the timeout if the flow is offloaded?

Thanks.