From: Florian Westphal <fw@strlen.de>
To: <netfilter-devel@vger.kernel.org>
Cc: Florian Westphal <fw@strlen.de>
Subject: [PATCH nf 1/4] netfilter: conntrack: remove two args from resolve_clash
Date: Mon, 3 Feb 2020 17:37:04 +0100 [thread overview]
Message-ID: <20200203163707.27254-2-fw@strlen.de> (raw)
In-Reply-To: <20200203163707.27254-1-fw@strlen.de>
ctinfo is whats taken from the skb, i.e.
ct = nf_ct_get(skb, &ctinfo).
We do not pass 'ct' and instead re-fetch it from the skb.
Just do the same for both netns and ctinfo.
Also add a comment on what clash resolution is supposed to do.
While at it, one indent level can be removed.
Signed-off-by: Florian Westphal <fw@strlen.de>
---
net/netfilter/nf_conntrack_core.c | 69 +++++++++++++++++++++++--------
1 file changed, 51 insertions(+), 18 deletions(-)
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index d1305423640f..5e332b01f3c0 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -894,31 +894,64 @@ static void nf_ct_acct_merge(struct nf_conn *ct, enum ip_conntrack_info ctinfo,
}
}
-/* Resolve race on insertion if this protocol allows this. */
+/**
+ * nf_ct_resolve_clash - attempt to handle clash without packet drop
+ *
+ * @skb: skb that causes the clash
+ * @h: tuplehash of the clashing entry already in table
+ *
+ * A conntrack entry can be inserted to the connection tracking table
+ * if there is no existing entry with an identical tuple.
+ *
+ * If there is one, @skb (and the assocated, unconfirmed conntrack) has
+ * to be dropped. In case @skb is retransmitted, next conntrack lookup
+ * will find the already-existing entry.
+ *
+ * The major problem with such packet drop is the extra delay added by
+ * the packet loss -- it will take some time for a retransmit to occur
+ * (or the sender to time out when waiting for a reply).
+ *
+ * This function attempts to handle the situation without packet drop.
+ *
+ * If @skb has no NAT transformation or if the colliding entries are
+ * exactly the same, only the to-be-confirmed conntrack entry is discarded
+ * and @skb is associated with the conntrack entry already in the table.
+ *
+ * Returns NF_DROP if the clash could not be resolved.
+ */
static __cold noinline int
-nf_ct_resolve_clash(struct net *net, struct sk_buff *skb,
- enum ip_conntrack_info ctinfo,
- struct nf_conntrack_tuple_hash *h)
+nf_ct_resolve_clash(struct sk_buff *skb, struct nf_conntrack_tuple_hash *h)
{
/* This is the conntrack entry already in hashes that won race. */
struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h);
const struct nf_conntrack_l4proto *l4proto;
- enum ip_conntrack_info oldinfo;
- struct nf_conn *loser_ct = nf_ct_get(skb, &oldinfo);
+ enum ip_conntrack_info ctinfo;
+ struct nf_conn *loser_ct;
+ struct net *net;
+
+ loser_ct = nf_ct_get(skb, &ctinfo);
l4proto = nf_ct_l4proto_find(nf_ct_protonum(ct));
- if (l4proto->allow_clash &&
- !nf_ct_is_dying(ct) &&
- atomic_inc_not_zero(&ct->ct_general.use)) {
- if (((ct->status & IPS_NAT_DONE_MASK) == 0) ||
- nf_ct_match(ct, loser_ct)) {
- nf_ct_acct_merge(ct, ctinfo, loser_ct);
- nf_conntrack_put(&loser_ct->ct_general);
- nf_ct_set(skb, ct, oldinfo);
- return NF_ACCEPT;
- }
- nf_ct_put(ct);
+ if (!l4proto->allow_clash)
+ goto drop;
+
+ if (nf_ct_is_dying(ct))
+ goto drop;
+
+ if (!atomic_inc_not_zero(&ct->ct_general.use))
+ goto drop;
+
+ if (((ct->status & IPS_NAT_DONE_MASK) == 0) ||
+ nf_ct_match(ct, loser_ct)) {
+ nf_ct_acct_merge(ct, ctinfo, loser_ct);
+ nf_conntrack_put(&loser_ct->ct_general);
+ nf_ct_set(skb, ct, ctinfo);
+ return NF_ACCEPT;
}
+
+ nf_ct_put(ct);
+drop:
+ net = nf_ct_net(loser_ct);
NF_CT_STAT_INC(net, drop);
return NF_DROP;
}
@@ -1036,7 +1069,7 @@ __nf_conntrack_confirm(struct sk_buff *skb)
out:
nf_ct_add_to_dying_list(ct);
- ret = nf_ct_resolve_clash(net, skb, ctinfo, h);
+ ret = nf_ct_resolve_clash(skb, h);
dying:
nf_conntrack_double_unlock(hash, reply_hash);
NF_CT_STAT_INC(net, insert_failed);
--
2.24.1
next prev parent reply other threads:[~2020-02-03 16:39 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-02-03 16:37 [PATCH nf 0/4] netfilter: conntrack: allow insertion of clashing entries Florian Westphal
2020-02-03 16:37 ` Florian Westphal [this message]
2020-02-03 16:37 ` [PATCH nf 2/4] netfilter: conntrack: place confirm-bit setting in a helper Florian Westphal
2020-02-03 16:37 ` [PATCH nf 3/4] netfilter: conntrack: split resolve_clash function Florian Westphal
2020-02-03 16:37 ` [PATCH nf 4/4] netfilter: conntrack: allow insertion of clashing entries Florian Westphal
2020-02-17 19:25 ` [PATCH nf 0/4] " Pablo Neira Ayuso
2020-02-17 20:12 ` Florian Westphal
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=20200203163707.27254-2-fw@strlen.de \
--to=fw@strlen.de \
--cc=netfilter-devel@vger.kernel.org \
/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).