[PATCH 21/24] net, diet: Make packet tpacket/mmap/fanout/rings optional

[Andi Kleen]

From: Andi Kleen <ak@xxxxxxxxxxxxxxx>

Many DHCP clients need basic packet sockets, but they don't need
the fancy zero copy packet capture code, like tpacket, mmap, rings,
fanouts. This is quite substantial code, so it's worthwhile to
make it optional

Worth nearly 10k code.

   text	   data	    bss	    dec	    hex	filename
 952827	  71874	  25352	1050053	 1005c5	net/built-in.o-with-packet-mmap
 943211	  71810	  25352	1040373	  fdff5	net/built-in.o-wo-packet-mmap

Signed-off-by: Andi Kleen <ak@xxxxxxxxxxxxxxx>
---
 net/packet/Kconfig     |  8 +++++
 net/packet/af_packet.c | 82 +++++++++++++++++++++++++++++++++++++++++---------
 2 files changed, 75 insertions(+), 15 deletions(-)

diff --git a/net/packet/Kconfig b/net/packet/Kconfig
index cc55b35..c215d31 100644
--- a/net/packet/Kconfig
+++ b/net/packet/Kconfig
@@ -22,3 +22,11 @@ config PACKET_DIAG
 	---help---
 	  Support for PF_PACKET sockets monitoring interface used by the ss tool.
 	  If unsure, say Y.
+
+config PACKET_MMAP
+	bool "Enable packet mmap/ring support"
+	depends on PACKET
+	default y
+	---help---
+	  Enable support to mmap the packet data zero copy. This is useful for
+	  highspeed packet interceptors.
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index b85c67c..723f57f 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -158,6 +158,8 @@ struct packet_mreq_max {
 	unsigned char	mr_address[MAX_ADDR_LEN];
 };
 
+#ifdef CONFIG_PACKET_MMAP
+
 union tpacket_uhdr {
 	struct tpacket_hdr  *h1;
 	struct tpacket2_hdr *h2;
@@ -165,8 +167,6 @@ union tpacket_uhdr {
 	void *raw;
 };
 
-static int packet_set_ring(struct sock *sk, union tpacket_req_u *req_u,
-		int closing, int tx_ring);
 
 #define V3_ALIGNMENT	(8)
 
@@ -213,6 +213,9 @@ static void prb_clear_rxhash(struct tpacket_kbdq_core *,
 		struct tpacket3_hdr *);
 static void prb_fill_vlan_info(struct tpacket_kbdq_core *,
 		struct tpacket3_hdr *);
+
+#endif
+
 static void packet_flush_mclist(struct sock *sk);
 
 struct packet_skb_cb {
@@ -384,6 +387,8 @@ static void unregister_prot_hook(struct sock *sk, bool sync)
 		__unregister_prot_hook(sk, sync);
 }
 
+#ifdef CONFIG_PACKET_MMAP
+
 static inline __pure struct page *pgv_to_page(void *addr)
 {
 	if (is_vmalloc_addr(addr))
@@ -1210,6 +1215,8 @@ static unsigned int packet_read_pending(const struct packet_ring_buffer *rb)
 	return refcnt;
 }
 
+#endif
+
 static int packet_alloc_pending(struct packet_sock *po)
 {
 	po->rx_ring.pending_refcnt = NULL;
@@ -1226,6 +1233,7 @@ static void packet_free_pending(struct packet_sock *po)
 	free_percpu(po->tx_ring.pending_refcnt);
 }
 
+#ifdef CONFIG_PACKET_MMAP
 static bool packet_rcv_has_room(struct packet_sock *po, struct sk_buff *skb)
 {
 	struct sock *sk = &po->sk;
@@ -1249,6 +1257,8 @@ static bool packet_rcv_has_room(struct packet_sock *po, struct sk_buff *skb)
 	return has_room;
 }
 
+#endif
+
 static void packet_sock_destruct(struct sock *sk)
 {
 	skb_queue_purge(&sk->sk_error_queue);
@@ -1264,6 +1274,8 @@ static void packet_sock_destruct(struct sock *sk)
 	sk_refcnt_debug_dec(sk);
 }
 
+#ifdef CONFIG_PACKET_MMAP
+
 static int fanout_rr_next(struct packet_fanout *f, unsigned int num)
 {
 	int x = atomic_read(&f->rr_cur) + 1;
@@ -1530,6 +1542,12 @@ static void fanout_release(struct sock *sk)
 	mutex_unlock(&fanout_mutex);
 }
 
+#else
+static void __fanout_unlink(struct sock *sk, struct packet_sock *po) {}
+static void __fanout_link(struct sock *sk, struct packet_sock *po) {}
+static void fanout_release(struct sock *sk) {}
+#endif
+
 static const struct proto_ops packet_ops;
 
 static const struct proto_ops packet_ops_spkt;
@@ -1867,6 +1885,11 @@ drop:
 	return 0;
 }
 
+#ifdef CONFIG_PACKET_MMAP
+
+static int packet_set_ring(struct sock *sk, union tpacket_req_u *req_u,
+		int closing, int tx_ring);
+
 static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev,
 		       struct packet_type *pt, struct net_device *orig_dev)
 {
@@ -2357,6 +2380,35 @@ out:
 	return err;
 }
 
+static inline bool use_tpacket(struct packet_sock *po)
+{
+	return po->tx_ring.pg_vec;
+}
+
+static void tpacket_release(struct sock *sk, struct packet_sock *po)
+{
+	union tpacket_req_u req_u;
+
+	if (po->rx_ring.pg_vec) {
+		memset(&req_u, 0, sizeof(req_u));
+		packet_set_ring(sk, &req_u, 1, 0);
+	}
+
+	if (po->tx_ring.pg_vec) {
+		memset(&req_u, 0, sizeof(req_u));
+		packet_set_ring(sk, &req_u, 1, 1);
+	}
+}
+
+#else
+static inline bool use_tpacket(struct packet_sock *po) { return false; }
+static inline void tpacket_release(struct sock *sk, struct packet_sock *po) {}
+static inline int tpacket_snd(struct packet_sock *po, struct msghdr *msg) { return 0; }
+static inline int tpacket_rcv(struct sk_buff *skb, struct net_device *dev,
+			      struct packet_type *pt, struct net_device *orig_dev)
+{ return 0; }
+#endif
+
 static struct sk_buff *packet_alloc_skb(struct sock *sk, size_t prepad,
 				        size_t reserve, size_t len,
 				        size_t linear, int noblock,
@@ -2576,7 +2628,7 @@ static int packet_sendmsg(struct kiocb *iocb, struct socket *sock,
 	struct sock *sk = sock->sk;
 	struct packet_sock *po = pkt_sk(sk);
 
-	if (po->tx_ring.pg_vec)
+	if (use_tpacket(po))
 		return tpacket_snd(po, msg);
 	else
 		return packet_snd(sock, msg, len);
@@ -2592,7 +2644,6 @@ static int packet_release(struct socket *sock)
 	struct sock *sk = sock->sk;
 	struct packet_sock *po;
 	struct net *net;
-	union tpacket_req_u req_u;
 
 	if (!sk)
 		return 0;
@@ -2620,15 +2671,7 @@ static int packet_release(struct socket *sock)
 
 	packet_flush_mclist(sk);
 
-	if (po->rx_ring.pg_vec) {
-		memset(&req_u, 0, sizeof(req_u));
-		packet_set_ring(sk, &req_u, 1, 0);
-	}
-
-	if (po->tx_ring.pg_vec) {
-		memset(&req_u, 0, sizeof(req_u));
-		packet_set_ring(sk, &req_u, 1, 1);
-	}
+	tpacket_release(sk, po);
 
 	fanout_release(sk);
 
@@ -3203,7 +3246,7 @@ static int
 packet_setsockopt(struct socket *sock, int level, int optname, char __user *optval, unsigned int optlen)
 {
 	struct sock *sk = sock->sk;
-	struct packet_sock *po = pkt_sk(sk);
+	struct packet_sock *po __maybe_unused = pkt_sk(sk);
 	int ret;
 
 	if (level != SOL_PACKET)
@@ -3231,6 +3274,7 @@ packet_setsockopt(struct socket *sock, int level, int optname, char __user *optv
 		return ret;
 	}
 
+#ifdef CONFIG_PACKET_MMAP
 	case PACKET_RX_RING:
 	case PACKET_TX_RING:
 	{
@@ -3314,6 +3358,7 @@ packet_setsockopt(struct socket *sock, int level, int optname, char __user *optv
 		po->tp_loss = !!val;
 		return 0;
 	}
+#endif
 	case PACKET_AUXDATA:
 	{
 		int val;
@@ -3366,6 +3411,7 @@ packet_setsockopt(struct socket *sock, int level, int optname, char __user *optv
 		po->tp_tstamp = val;
 		return 0;
 	}
+#ifdef CONFIG_PACKET_MMAP
 	case PACKET_FANOUT:
 	{
 		int val;
@@ -3390,6 +3436,7 @@ packet_setsockopt(struct socket *sock, int level, int optname, char __user *optv
 		po->tp_tx_has_off = !!val;
 		return 0;
 	}
+#endif
 	case PACKET_QDISC_BYPASS:
 	{
 		int val;
@@ -3615,6 +3662,7 @@ static int packet_ioctl(struct socket *sock, unsigned int cmd,
 	return 0;
 }
 
+#ifdef CONFIG_PACKET_MMAP
 static unsigned int packet_poll(struct file *file, struct socket *sock,
 				poll_table *wait)
 {
@@ -3855,7 +3903,7 @@ static int packet_set_ring(struct sock *sk, union tpacket_req_u *req_u,
 		swap(rb->pg_vec_len, req->tp_block_nr);
 
 		rb->pg_vec_pages = req->tp_block_size/PAGE_SIZE;
-		po->prot_hook.func = (po->rx_ring.pg_vec) ?
+		po->prot_hook.func = use_tpacket(po) ?
 						tpacket_rcv : packet_rcv;
 		skb_queue_purge(rb_queue);
 		if (atomic_read(&po->mapped))
@@ -3944,6 +3992,10 @@ out:
 	mutex_unlock(&po->pg_vec_lock);
 	return err;
 }
+#else
+#define packet_mmap sock_no_mmap
+#define packet_poll datagram_poll
+#endif
 
 static const struct proto_ops packet_ops_spkt = {
 	.family =	PF_PACKET,
-- 
1.9.0

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/