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
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
|
From: Pablo Neira Ayuso <pablo@netfilter.org>
Date: Tue, 23 Mar 2021 00:56:22 +0100
Subject: [PATCH] netfilter: flowtable: consolidate
skb_try_make_writable() call
Fetch the layer 4 header size to be mangled by NAT when building the
tuple, then use it to make writable the network and the transport
headers. After this update, the NAT routines now assumes that the skbuff
area is writable. Do the pointer refetch only after the single
skb_try_make_writable() call.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
--- a/net/netfilter/nf_flow_table_core.c
+++ b/net/netfilter/nf_flow_table_core.c
@@ -394,9 +394,6 @@ static int nf_flow_nat_port_tcp(struct s
{
struct tcphdr *tcph;
- if (skb_try_make_writable(skb, thoff + sizeof(*tcph)))
- return -1;
-
tcph = (void *)(skb_network_header(skb) + thoff);
inet_proto_csum_replace2(&tcph->check, skb, port, new_port, false);
@@ -408,9 +405,6 @@ static int nf_flow_nat_port_udp(struct s
{
struct udphdr *udph;
- if (skb_try_make_writable(skb, thoff + sizeof(*udph)))
- return -1;
-
udph = (void *)(skb_network_header(skb) + thoff);
if (udph->check || skb->ip_summed == CHECKSUM_PARTIAL) {
inet_proto_csum_replace2(&udph->check, skb, port,
@@ -446,9 +440,6 @@ int nf_flow_snat_port(const struct flow_
struct flow_ports *hdr;
__be16 port, new_port;
- if (skb_try_make_writable(skb, thoff + sizeof(*hdr)))
- return -1;
-
hdr = (void *)(skb_network_header(skb) + thoff);
switch (dir) {
@@ -477,9 +468,6 @@ int nf_flow_dnat_port(const struct flow_
struct flow_ports *hdr;
__be16 port, new_port;
- if (skb_try_make_writable(skb, thoff + sizeof(*hdr)))
- return -1;
-
hdr = (void *)(skb_network_header(skb) + thoff);
switch (dir) {
--- a/net/netfilter/nf_flow_table_ip.c
+++ b/net/netfilter/nf_flow_table_ip.c
@@ -39,9 +39,6 @@ static int nf_flow_nat_ip_tcp(struct sk_
{
struct tcphdr *tcph;
- if (skb_try_make_writable(skb, thoff + sizeof(*tcph)))
- return -1;
-
tcph = (void *)(skb_network_header(skb) + thoff);
inet_proto_csum_replace4(&tcph->check, skb, addr, new_addr, true);
@@ -53,9 +50,6 @@ static int nf_flow_nat_ip_udp(struct sk_
{
struct udphdr *udph;
- if (skb_try_make_writable(skb, thoff + sizeof(*udph)))
- return -1;
-
udph = (void *)(skb_network_header(skb) + thoff);
if (udph->check || skb->ip_summed == CHECKSUM_PARTIAL) {
inet_proto_csum_replace4(&udph->check, skb, addr,
@@ -136,19 +130,17 @@ static int nf_flow_dnat_ip(const struct
}
static int nf_flow_nat_ip(const struct flow_offload *flow, struct sk_buff *skb,
- unsigned int thoff, enum flow_offload_tuple_dir dir)
+ unsigned int thoff, enum flow_offload_tuple_dir dir,
+ struct iphdr *iph)
{
- struct iphdr *iph = ip_hdr(skb);
-
if (test_bit(NF_FLOW_SNAT, &flow->flags) &&
(nf_flow_snat_port(flow, skb, thoff, iph->protocol, dir) < 0 ||
- nf_flow_snat_ip(flow, skb, ip_hdr(skb), thoff, dir) < 0))
+ nf_flow_snat_ip(flow, skb, iph, thoff, dir) < 0))
return -1;
- iph = ip_hdr(skb);
if (test_bit(NF_FLOW_DNAT, &flow->flags) &&
(nf_flow_dnat_port(flow, skb, thoff, iph->protocol, dir) < 0 ||
- nf_flow_dnat_ip(flow, skb, ip_hdr(skb), thoff, dir) < 0))
+ nf_flow_dnat_ip(flow, skb, iph, thoff, dir) < 0))
return -1;
return 0;
@@ -160,10 +152,10 @@ static bool ip_has_options(unsigned int
}
static int nf_flow_tuple_ip(struct sk_buff *skb, const struct net_device *dev,
- struct flow_offload_tuple *tuple)
+ struct flow_offload_tuple *tuple, u32 *hdrsize)
{
- unsigned int thoff, hdrsize;
struct flow_ports *ports;
+ unsigned int thoff;
struct iphdr *iph;
if (!pskb_may_pull(skb, sizeof(*iph)))
@@ -178,10 +170,10 @@ static int nf_flow_tuple_ip(struct sk_bu
switch (iph->protocol) {
case IPPROTO_TCP:
- hdrsize = sizeof(struct tcphdr);
+ *hdrsize = sizeof(struct tcphdr);
break;
case IPPROTO_UDP:
- hdrsize = sizeof(struct udphdr);
+ *hdrsize = sizeof(struct udphdr);
break;
default:
return -1;
@@ -191,7 +183,7 @@ static int nf_flow_tuple_ip(struct sk_bu
return -1;
thoff = iph->ihl * 4;
- if (!pskb_may_pull(skb, thoff + hdrsize))
+ if (!pskb_may_pull(skb, thoff + *hdrsize))
return -1;
iph = ip_hdr(skb);
@@ -252,11 +244,12 @@ nf_flow_offload_ip_hook(void *priv, stru
unsigned int thoff;
struct iphdr *iph;
__be32 nexthop;
+ u32 hdrsize;
if (skb->protocol != htons(ETH_P_IP))
return NF_ACCEPT;
- if (nf_flow_tuple_ip(skb, state->in, &tuple) < 0)
+ if (nf_flow_tuple_ip(skb, state->in, &tuple, &hdrsize) < 0)
return NF_ACCEPT;
tuplehash = flow_offload_lookup(flow_table, &tuple);
@@ -271,11 +264,13 @@ nf_flow_offload_ip_hook(void *priv, stru
if (unlikely(nf_flow_exceeds_mtu(skb, flow->tuplehash[dir].tuple.mtu)))
return NF_ACCEPT;
- if (skb_try_make_writable(skb, sizeof(*iph)))
+ iph = ip_hdr(skb);
+ thoff = iph->ihl * 4;
+ if (skb_try_make_writable(skb, thoff + hdrsize))
return NF_DROP;
- thoff = ip_hdr(skb)->ihl * 4;
- if (nf_flow_state_check(flow, ip_hdr(skb)->protocol, skb, thoff))
+ iph = ip_hdr(skb);
+ if (nf_flow_state_check(flow, iph->protocol, skb, thoff))
return NF_ACCEPT;
flow_offload_refresh(flow_table, flow);
@@ -285,10 +280,9 @@ nf_flow_offload_ip_hook(void *priv, stru
return NF_ACCEPT;
}
- if (nf_flow_nat_ip(flow, skb, thoff, dir) < 0)
+ if (nf_flow_nat_ip(flow, skb, thoff, dir, iph) < 0)
return NF_DROP;
- iph = ip_hdr(skb);
ip_decrease_ttl(iph);
skb->tstamp = 0;
@@ -317,9 +311,6 @@ static int nf_flow_nat_ipv6_tcp(struct s
{
struct tcphdr *tcph;
- if (skb_try_make_writable(skb, thoff + sizeof(*tcph)))
- return -1;
-
tcph = (void *)(skb_network_header(skb) + thoff);
inet_proto_csum_replace16(&tcph->check, skb, addr->s6_addr32,
new_addr->s6_addr32, true);
@@ -333,9 +324,6 @@ static int nf_flow_nat_ipv6_udp(struct s
{
struct udphdr *udph;
- if (skb_try_make_writable(skb, thoff + sizeof(*udph)))
- return -1;
-
udph = (void *)(skb_network_header(skb) + thoff);
if (udph->check || skb->ip_summed == CHECKSUM_PARTIAL) {
inet_proto_csum_replace16(&udph->check, skb, addr->s6_addr32,
@@ -417,31 +405,30 @@ static int nf_flow_dnat_ipv6(const struc
static int nf_flow_nat_ipv6(const struct flow_offload *flow,
struct sk_buff *skb,
- enum flow_offload_tuple_dir dir)
+ enum flow_offload_tuple_dir dir,
+ struct ipv6hdr *ip6h)
{
- struct ipv6hdr *ip6h = ipv6_hdr(skb);
unsigned int thoff = sizeof(*ip6h);
if (test_bit(NF_FLOW_SNAT, &flow->flags) &&
(nf_flow_snat_port(flow, skb, thoff, ip6h->nexthdr, dir) < 0 ||
- nf_flow_snat_ipv6(flow, skb, ipv6_hdr(skb), thoff, dir) < 0))
+ nf_flow_snat_ipv6(flow, skb, ip6h, thoff, dir) < 0))
return -1;
- ip6h = ipv6_hdr(skb);
if (test_bit(NF_FLOW_DNAT, &flow->flags) &&
(nf_flow_dnat_port(flow, skb, thoff, ip6h->nexthdr, dir) < 0 ||
- nf_flow_dnat_ipv6(flow, skb, ipv6_hdr(skb), thoff, dir) < 0))
+ nf_flow_dnat_ipv6(flow, skb, ip6h, thoff, dir) < 0))
return -1;
return 0;
}
static int nf_flow_tuple_ipv6(struct sk_buff *skb, const struct net_device *dev,
- struct flow_offload_tuple *tuple)
+ struct flow_offload_tuple *tuple, u32 *hdrsize)
{
- unsigned int thoff, hdrsize;
struct flow_ports *ports;
struct ipv6hdr *ip6h;
+ unsigned int thoff;
if (!pskb_may_pull(skb, sizeof(*ip6h)))
return -1;
@@ -450,10 +437,10 @@ static int nf_flow_tuple_ipv6(struct sk_
switch (ip6h->nexthdr) {
case IPPROTO_TCP:
- hdrsize = sizeof(struct tcphdr);
+ *hdrsize = sizeof(struct tcphdr);
break;
case IPPROTO_UDP:
- hdrsize = sizeof(struct udphdr);
+ *hdrsize = sizeof(struct udphdr);
break;
default:
return -1;
@@ -463,7 +450,7 @@ static int nf_flow_tuple_ipv6(struct sk_
return -1;
thoff = sizeof(*ip6h);
- if (!pskb_may_pull(skb, thoff + hdrsize))
+ if (!pskb_may_pull(skb, thoff + *hdrsize))
return -1;
ip6h = ipv6_hdr(skb);
@@ -493,11 +480,12 @@ nf_flow_offload_ipv6_hook(void *priv, st
struct net_device *outdev;
struct ipv6hdr *ip6h;
struct rt6_info *rt;
+ u32 hdrsize;
if (skb->protocol != htons(ETH_P_IPV6))
return NF_ACCEPT;
- if (nf_flow_tuple_ipv6(skb, state->in, &tuple) < 0)
+ if (nf_flow_tuple_ipv6(skb, state->in, &tuple, &hdrsize) < 0)
return NF_ACCEPT;
tuplehash = flow_offload_lookup(flow_table, &tuple);
@@ -523,13 +511,13 @@ nf_flow_offload_ipv6_hook(void *priv, st
return NF_ACCEPT;
}
- if (skb_try_make_writable(skb, sizeof(*ip6h)))
+ if (skb_try_make_writable(skb, sizeof(*ip6h) + hdrsize))
return NF_DROP;
- if (nf_flow_nat_ipv6(flow, skb, dir) < 0)
+ ip6h = ipv6_hdr(skb);
+ if (nf_flow_nat_ipv6(flow, skb, dir, ip6h) < 0)
return NF_DROP;
- ip6h = ipv6_hdr(skb);
ip6h->hop_limit--;
skb->tstamp = 0;
|
