Commit a10fa0b4 authored by Pablo Neira Ayuso's avatar Pablo Neira Ayuso Committed by Florian Westphal

netfilter: flowtable: split IPv4 datapath in helper functions

Add context structure and helper functions to look up for a matching
IPv4 entry in the flowtable and to forward packets.

No functional changes are intended.
Signed-off-by: default avatarPablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: default avatarFlorian Westphal <fw@strlen.de>
parent fa502c86
...@@ -163,38 +163,43 @@ static void nf_flow_tuple_encap(struct sk_buff *skb, ...@@ -163,38 +163,43 @@ static void nf_flow_tuple_encap(struct sk_buff *skb,
} }
} }
static int nf_flow_tuple_ip(struct sk_buff *skb, const struct net_device *dev, struct nf_flowtable_ctx {
struct flow_offload_tuple *tuple, u32 *hdrsize, const struct net_device *in;
u32 offset) u32 offset;
u32 hdrsize;
};
static int nf_flow_tuple_ip(struct nf_flowtable_ctx *ctx, struct sk_buff *skb,
struct flow_offload_tuple *tuple)
{ {
struct flow_ports *ports; struct flow_ports *ports;
unsigned int thoff; unsigned int thoff;
struct iphdr *iph; struct iphdr *iph;
u8 ipproto; u8 ipproto;
if (!pskb_may_pull(skb, sizeof(*iph) + offset)) if (!pskb_may_pull(skb, sizeof(*iph) + ctx->offset))
return -1; return -1;
iph = (struct iphdr *)(skb_network_header(skb) + offset); iph = (struct iphdr *)(skb_network_header(skb) + ctx->offset);
thoff = (iph->ihl * 4); thoff = (iph->ihl * 4);
if (ip_is_fragment(iph) || if (ip_is_fragment(iph) ||
unlikely(ip_has_options(thoff))) unlikely(ip_has_options(thoff)))
return -1; return -1;
thoff += offset; thoff += ctx->offset;
ipproto = iph->protocol; ipproto = iph->protocol;
switch (ipproto) { switch (ipproto) {
case IPPROTO_TCP: case IPPROTO_TCP:
*hdrsize = sizeof(struct tcphdr); ctx->hdrsize = sizeof(struct tcphdr);
break; break;
case IPPROTO_UDP: case IPPROTO_UDP:
*hdrsize = sizeof(struct udphdr); ctx->hdrsize = sizeof(struct udphdr);
break; break;
#ifdef CONFIG_NF_CT_PROTO_GRE #ifdef CONFIG_NF_CT_PROTO_GRE
case IPPROTO_GRE: case IPPROTO_GRE:
*hdrsize = sizeof(struct gre_base_hdr); ctx->hdrsize = sizeof(struct gre_base_hdr);
break; break;
#endif #endif
default: default:
...@@ -204,7 +209,7 @@ static int nf_flow_tuple_ip(struct sk_buff *skb, const struct net_device *dev, ...@@ -204,7 +209,7 @@ static int nf_flow_tuple_ip(struct sk_buff *skb, const struct net_device *dev,
if (iph->ttl <= 1) if (iph->ttl <= 1)
return -1; return -1;
if (!pskb_may_pull(skb, thoff + *hdrsize)) if (!pskb_may_pull(skb, thoff + ctx->hdrsize))
return -1; return -1;
switch (ipproto) { switch (ipproto) {
...@@ -224,13 +229,13 @@ static int nf_flow_tuple_ip(struct sk_buff *skb, const struct net_device *dev, ...@@ -224,13 +229,13 @@ static int nf_flow_tuple_ip(struct sk_buff *skb, const struct net_device *dev,
} }
} }
iph = (struct iphdr *)(skb_network_header(skb) + offset); iph = (struct iphdr *)(skb_network_header(skb) + ctx->offset);
tuple->src_v4.s_addr = iph->saddr; tuple->src_v4.s_addr = iph->saddr;
tuple->dst_v4.s_addr = iph->daddr; tuple->dst_v4.s_addr = iph->daddr;
tuple->l3proto = AF_INET; tuple->l3proto = AF_INET;
tuple->l4proto = ipproto; tuple->l4proto = ipproto;
tuple->iifidx = dev->ifindex; tuple->iifidx = ctx->in->ifindex;
nf_flow_tuple_encap(skb, tuple); nf_flow_tuple_encap(skb, tuple);
return 0; return 0;
...@@ -336,58 +341,56 @@ static unsigned int nf_flow_queue_xmit(struct net *net, struct sk_buff *skb, ...@@ -336,58 +341,56 @@ static unsigned int nf_flow_queue_xmit(struct net *net, struct sk_buff *skb,
return NF_STOLEN; return NF_STOLEN;
} }
unsigned int static struct flow_offload_tuple_rhash *
nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, nf_flow_offload_lookup(struct nf_flowtable_ctx *ctx,
const struct nf_hook_state *state) struct nf_flowtable *flow_table, struct sk_buff *skb)
{ {
struct flow_offload_tuple_rhash *tuplehash;
struct nf_flowtable *flow_table = priv;
struct flow_offload_tuple tuple = {}; struct flow_offload_tuple tuple = {};
enum flow_offload_tuple_dir dir;
struct flow_offload *flow;
struct net_device *outdev;
u32 hdrsize, offset = 0;
unsigned int thoff, mtu;
struct rtable *rt;
struct iphdr *iph;
__be32 nexthop;
int ret;
if (skb->protocol != htons(ETH_P_IP) && if (skb->protocol != htons(ETH_P_IP) &&
!nf_flow_skb_encap_protocol(skb, htons(ETH_P_IP), &offset)) !nf_flow_skb_encap_protocol(skb, htons(ETH_P_IP), &ctx->offset))
return NF_ACCEPT; return NULL;
if (nf_flow_tuple_ip(skb, state->in, &tuple, &hdrsize, offset) < 0) if (nf_flow_tuple_ip(ctx, skb, &tuple) < 0)
return NF_ACCEPT; return NULL;
tuplehash = flow_offload_lookup(flow_table, &tuple); return flow_offload_lookup(flow_table, &tuple);
if (tuplehash == NULL) }
return NF_ACCEPT;
static int nf_flow_offload_forward(struct nf_flowtable_ctx *ctx,
struct nf_flowtable *flow_table,
struct flow_offload_tuple_rhash *tuplehash,
struct sk_buff *skb)
{
enum flow_offload_tuple_dir dir;
struct flow_offload *flow;
unsigned int thoff, mtu;
struct iphdr *iph;
dir = tuplehash->tuple.dir; dir = tuplehash->tuple.dir;
flow = container_of(tuplehash, struct flow_offload, tuplehash[dir]); flow = container_of(tuplehash, struct flow_offload, tuplehash[dir]);
mtu = flow->tuplehash[dir].tuple.mtu + offset; mtu = flow->tuplehash[dir].tuple.mtu + ctx->offset;
if (unlikely(nf_flow_exceeds_mtu(skb, mtu))) if (unlikely(nf_flow_exceeds_mtu(skb, mtu)))
return NF_ACCEPT; return 0;
iph = (struct iphdr *)(skb_network_header(skb) + offset); iph = (struct iphdr *)(skb_network_header(skb) + ctx->offset);
thoff = (iph->ihl * 4) + offset; thoff = (iph->ihl * 4) + ctx->offset;
if (nf_flow_state_check(flow, iph->protocol, skb, thoff)) if (nf_flow_state_check(flow, iph->protocol, skb, thoff))
return NF_ACCEPT; return 0;
if (!nf_flow_dst_check(&tuplehash->tuple)) { if (!nf_flow_dst_check(&tuplehash->tuple)) {
flow_offload_teardown(flow); flow_offload_teardown(flow);
return NF_ACCEPT; return 0;
} }
if (skb_try_make_writable(skb, thoff + hdrsize)) if (skb_try_make_writable(skb, thoff + ctx->hdrsize))
return NF_DROP; return -1;
flow_offload_refresh(flow_table, flow); flow_offload_refresh(flow_table, flow);
nf_flow_encap_pop(skb, tuplehash); nf_flow_encap_pop(skb, tuplehash);
thoff -= offset; thoff -= ctx->offset;
iph = ip_hdr(skb); iph = ip_hdr(skb);
nf_flow_nat_ip(flow, skb, thoff, dir, iph); nf_flow_nat_ip(flow, skb, thoff, dir, iph);
...@@ -398,6 +401,35 @@ nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, ...@@ -398,6 +401,35 @@ nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb,
if (flow_table->flags & NF_FLOWTABLE_COUNTER) if (flow_table->flags & NF_FLOWTABLE_COUNTER)
nf_ct_acct_update(flow->ct, tuplehash->tuple.dir, skb->len); nf_ct_acct_update(flow->ct, tuplehash->tuple.dir, skb->len);
return 1;
}
unsigned int
nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb,
const struct nf_hook_state *state)
{
struct flow_offload_tuple_rhash *tuplehash;
struct nf_flowtable *flow_table = priv;
enum flow_offload_tuple_dir dir;
struct nf_flowtable_ctx ctx = {
.in = state->in,
};
struct flow_offload *flow;
struct net_device *outdev;
struct rtable *rt;
__be32 nexthop;
int ret;
tuplehash = nf_flow_offload_lookup(&ctx, flow_table, skb);
if (!tuplehash)
return NF_ACCEPT;
ret = nf_flow_offload_forward(&ctx, flow_table, tuplehash, skb);
if (ret < 0)
return NF_DROP;
else if (ret == 0)
return NF_ACCEPT;
if (unlikely(tuplehash->tuple.xmit_type == FLOW_OFFLOAD_XMIT_XFRM)) { if (unlikely(tuplehash->tuple.xmit_type == FLOW_OFFLOAD_XMIT_XFRM)) {
rt = (struct rtable *)tuplehash->tuple.dst_cache; rt = (struct rtable *)tuplehash->tuple.dst_cache;
memset(skb->cb, 0, sizeof(struct inet_skb_parm)); memset(skb->cb, 0, sizeof(struct inet_skb_parm));
...@@ -406,6 +438,9 @@ nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, ...@@ -406,6 +438,9 @@ nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb,
return nf_flow_xmit_xfrm(skb, state, &rt->dst); return nf_flow_xmit_xfrm(skb, state, &rt->dst);
} }
dir = tuplehash->tuple.dir;
flow = container_of(tuplehash, struct flow_offload, tuplehash[dir]);
switch (tuplehash->tuple.xmit_type) { switch (tuplehash->tuple.xmit_type) {
case FLOW_OFFLOAD_XMIT_NEIGH: case FLOW_OFFLOAD_XMIT_NEIGH:
rt = (struct rtable *)tuplehash->tuple.dst_cache; rt = (struct rtable *)tuplehash->tuple.dst_cache;
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment