1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
|
From: Felix Fietkau <nbd@nbd.name>
Date: Wed, 13 Jun 2018 12:33:39 +0200
Subject: [PATCH] netfilter: nf_flow_table: fix offloaded connection timeout
corner case
The full teardown of offloaded flows is deferred to a gc work item,
however processing of packets by netfilter needs to happen immediately
after a teardown is requested, because the conntrack state needs to be
fixed up.
Since the IPS_OFFLOAD_BIT is still kept until the teardown is complete,
the netfilter conntrack gc can accidentally bump the timeout of a
connection where offload was just stopped, causing a conntrack entry
leak.
Fix this by moving the conntrack timeout bumping from conntrack core to
the nf_flow_offload and add a check to prevent bogus timeout bumps.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -1040,18 +1040,6 @@ static bool gc_worker_can_early_drop(con
return false;
}
-#define DAY (86400 * HZ)
-
-/* Set an arbitrary timeout large enough not to ever expire, this save
- * us a check for the IPS_OFFLOAD_BIT from the packet path via
- * nf_ct_is_expired().
- */
-static void nf_ct_offload_timeout(struct nf_conn *ct)
-{
- if (nf_ct_expires(ct) < DAY / 2)
- ct->timeout = nfct_time_stamp + DAY;
-}
-
static void gc_worker(struct work_struct *work)
{
unsigned int min_interval = max(HZ / GC_MAX_BUCKETS_DIV, 1u);
@@ -1088,10 +1076,8 @@ static void gc_worker(struct work_struct
tmp = nf_ct_tuplehash_to_ctrack(h);
scanned++;
- if (test_bit(IPS_OFFLOAD_BIT, &tmp->status)) {
- nf_ct_offload_timeout(tmp);
+ if (test_bit(IPS_OFFLOAD_BIT, &tmp->status))
continue;
- }
if (nf_ct_is_expired(tmp)) {
nf_ct_gc_expired(tmp);
--- a/net/netfilter/nf_flow_table_core.c
+++ b/net/netfilter/nf_flow_table_core.c
@@ -185,8 +185,27 @@ static const struct rhashtable_params nf
.automatic_shrinking = true,
};
+#define DAY (86400 * HZ)
+
+/* Set an arbitrary timeout large enough not to ever expire, this save
+ * us a check for the IPS_OFFLOAD_BIT from the packet path via
+ * nf_ct_is_expired().
+ */
+static void nf_ct_offload_timeout(struct flow_offload *flow)
+{
+ struct flow_offload_entry *entry;
+ struct nf_conn *ct;
+
+ entry = container_of(flow, struct flow_offload_entry, flow);
+ ct = entry->ct;
+
+ if (nf_ct_expires(ct) < DAY / 2)
+ ct->timeout = nfct_time_stamp + DAY;
+}
+
int flow_offload_add(struct nf_flowtable *flow_table, struct flow_offload *flow)
{
+ nf_ct_offload_timeout(flow);
flow->timeout = (u32)jiffies;
rhashtable_insert_fast(&flow_table->rhashtable,
@@ -307,6 +326,8 @@ static int nf_flow_offload_gc_step(struc
rhashtable_walk_start(&hti);
while ((tuplehash = rhashtable_walk_next(&hti))) {
+ bool teardown;
+
if (IS_ERR(tuplehash)) {
err = PTR_ERR(tuplehash);
if (err != -EAGAIN)
@@ -319,9 +340,13 @@ static int nf_flow_offload_gc_step(struc
flow = container_of(tuplehash, struct flow_offload, tuplehash[0]);
- if (nf_flow_has_expired(flow) ||
- (flow->flags & (FLOW_OFFLOAD_DYING |
- FLOW_OFFLOAD_TEARDOWN)))
+ teardown = flow->flags & (FLOW_OFFLOAD_DYING |
+ FLOW_OFFLOAD_TEARDOWN);
+
+ if (!teardown)
+ nf_ct_offload_timeout(flow);
+
+ if (nf_flow_has_expired(flow) || teardown)
flow_offload_del(flow_table, flow);
}
out:
|
