Hi David,
please find enclosed for review the proposed changes to integrate UDP-Lite
code with UDP.
Please disregard all earlier patches, I have been putting in more hard work to
consolidate the code further, also with regard to expanding to UDP(-Lite) v6.
(Reductions are drastic: udplite.c is slimmed down to 186 lines.)
This is the v4-side and with a minimum of formatting diffs. Any suggestions
for v4 will automatically be incorporated into the v6 side in the subsequent
revision.
Enclosed patch applies on your 2.6.19 tree and on 2.6.18-rc6-mm2 and has been
tested (UDPv4/v6 as well) for quite a while.
R e s o l v e d I s s u e s
a) Common Naming scheme:
--stuff that works for both UDPv4/6 and UDP-Litev4/6 is called
udp_lib_xxx()
* restricted to v4: udp4_lib_xxx()
* restricted to v6: udp6_lib_xxx()
* when it should not be called directly: __udp_lib_xxx()
[resp.__udp{4,6}_lib_xxx()]
--some old functions (e.g. udp_v4_mcast_next()) have retained their name
although they would need to be renamed (here udp4_lib_mcast_next());
this is left for later -- a `name-change-only' patch
b) Correct MIB-counting of de facto received datagrams (resolving
http://bugzilla.kernel.org/show_bug.cgi?id=6660 for this module; some
follow-up
work may have to look at doing the same for sunrpc, nfs, and likewise)
c) udp_v{4,6}_get_port(): consolidated.
C h a n g e l o g
* inline functions which are shared between UDP and UDP-Lite and v4/v6 have
been put into header files rather than duplicated in source files
* made net/ipv4/udplite.c self-contained unit (#included at the end of udp.c)
* consolidated v4/v6 checksumming code (generic now for UDP and UDP-Lite)
--new: udp_csum_outgoing()
--contrasts with udplite_csum_outgoing()
--removed support for disabling UDPv6 checksums (RFC 2460 says it's illegal)
* udp_v{4,6}_hash() and udp_v{4,6}_unhash() identical: inlined into header file
* basic xfrm/netfilter support (thanks to Patrick McHardy)
* updated the protocol registration in udplite4_register to not do
/proc registration on error (thanks to comments by James Morris)
* detailed per-function breakdown as per earlier mail
Regards,
Gerrit
--
include/linux/udp.h | 11 +
include/net/udp.h | 89 +++++++++
net/ipv4/udp.c | 485 ++++++++++++++++++++++++++++++++--------------------
net/ipv6/udp.c | 50 -----
4 files changed, 406 insertions(+), 229 deletions(-)
diff --git a/include/linux/udp.h b/include/linux/udp.h
index 90223f0..1b7cf10 100644
--- a/include/linux/udp.h
+++ b/include/linux/udp.h
@@ -50,12 +50,23 @@ struct udp_sock {
* when the socket is uncorked.
*/
__u16 len; /* total length of pending frames */
+ /*
+ * Fields specific to UDP-Lite.
+ */
+ __u16 pcslen;
+ __u16 pcrlen;
+/* indicator bits used by pcflag: */
+#define UDPLITE_BIT 0x1 /* set by udplite proto init function */
+#define UDPLITE_SEND_CC 0x2 /* set via udplite setsockopt */
+#define UDPLITE_RECV_CC 0x4 /* set via udplite setsocktopt */
+ __u8 pcflag; /* marks socket as UDP-Lite if > 0 */
};
static inline struct udp_sock *udp_sk(const struct sock *sk)
{
return (struct udp_sock *)sk;
}
+#define IS_UDPLITE(__sk) (udp_sk(__sk)->pcflag)
#endif
diff --git a/include/net/udp.h b/include/net/udp.h
index db0c05f..b22abb3 100644
--- a/include/net/udp.h
+++ b/include/net/udp.h
@@ -26,9 +26,31 @@ #include <linux/list.h>
#include <net/inet_sock.h>
#include <net/sock.h>
#include <net/snmp.h>
+#include <net/ip.h>
+#include <linux/ipv6.h>
#include <linux/seq_file.h>
#define UDP_HTABLE_SIZE 128
+#include <net/udplite.h>
+
+/**
+ * struct udp_skb_cb - UDP(-Lite) private variables
+ *
+ * @header: private variables used by IPv4/IPv6
+ * @cscov: checksum coverage length (UDP-Lite only)
+ * @partial_cov: if set indicates partial csum coverage
+ */
+struct udp_skb_cb {
+ union {
+ struct inet_skb_parm h4;
+#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE)
+ struct inet6_skb_parm h6;
+#endif
+ } header;
+ __u16 cscov;
+ __u8 partial_cov;
+};
+#define UDP_SKB_CB(__skb) ((struct udp_skb_cb *)((__skb)->cb))
extern struct hlist_head udp_hash[UDP_HTABLE_SIZE];
extern rwlock_t udp_hash_lock;
@@ -47,6 +69,56 @@ extern struct proto udp_prot;
struct sk_buff;
+/*
+ * Generic checksumming routines for UDP(-Lite) v4 and v6
+ */
+static inline u16 __udp_lib_checksum_complete(struct sk_buff *skb)
+{
+ if (! UDP_SKB_CB(skb)->partial_cov)
+ return __skb_checksum_complete(skb);
+ return csum_fold(skb_checksum(skb, 0, UDP_SKB_CB(skb)->cscov,
+ skb->csum));
+}
+
+static __inline__ int udp_checksum_complete(struct sk_buff *skb)
+{
+ return skb->ip_summed != CHECKSUM_UNNECESSARY &&
+ __udp_lib_checksum_complete(skb);
+}
+
+/**
+ * udp_csum_outgoing - compute UDPv4/v6 checksum over fragments
+ * @sk: socket we are writing to
+ * @skb: sk_buff containing the filled-in UDP header
+ * (checksum field must be zeroed out)
+ */
+static inline u32 udp_csum_outgoing(struct sock *sk, struct sk_buff *skb)
+{
+ u32 csum = csum_partial(skb->h.raw, sizeof(struct udphdr), 0);
+
+ skb_queue_walk(&sk->sk_write_queue, skb) {
+ csum = csum_add(csum, skb->csum);
+ }
+ return csum;
+}
+
+/* hash routines shared between UDPv4/6 and UDP-Litev4/6 */
+static inline void udp_lib_hash(struct sock *sk)
+{
+ BUG();
+}
+
+static inline void udp_lib_unhash(struct sock *sk)
+{
+ write_lock_bh(&udp_hash_lock);
+ if (sk_del_node_init(sk)) {
+ inet_sk(sk)->num = 0;
+ sock_prot_dec_use(sk->sk_prot);
+ }
+ write_unlock_bh(&udp_hash_lock);
+}
+
+/* net/ipv4/udp.c */
extern int udp_get_port(struct sock *sk, unsigned short snum,
int (*saddr_cmp)(const struct sock *, const struct
sock *));
extern void udp_err(struct sk_buff *, u32);
@@ -61,21 +133,32 @@ extern unsigned int udp_poll(struct file
poll_table *wait);
DECLARE_SNMP_STAT(struct udp_mib, udp_statistics);
-#define UDP_INC_STATS(field) SNMP_INC_STATS(udp_statistics, field)
-#define UDP_INC_STATS_BH(field)
SNMP_INC_STATS_BH(udp_statistics, field)
-#define UDP_INC_STATS_USER(field) SNMP_INC_STATS_USER(udp_statistics,
field)
+/*
+ * SNMP statistics for UDP and UDP-Lite
+ */
+#define UDP_INC_STATS_USER(field, is_udplite) do { \
+ if (is_udplite) SNMP_INC_STATS_USER(udplite_statistics, field); \
+ else SNMP_INC_STATS_USER(udp_statistics, field); } while(0)
+#define UDP_INC_STATS_BH(field, is_udplite) do { \
+ if (is_udplite) SNMP_INC_STATS_BH(udplite_statistics, field); \
+ else SNMP_INC_STATS_BH(udp_statistics, field); } while(0)
+#define UDP_DEC_STATS_BH(field, is_udplite) do { \
+ if (is_udplite) SNMP_DEC_STATS_BH(udplite_statistics, field); \
+ else SNMP_DEC_STATS_BH(udp_statistics, field); } while(0)
/* /proc */
struct udp_seq_afinfo {
struct module *owner;
char *name;
sa_family_t family;
+ struct hlist_head *hashtable;
int (*seq_show) (struct seq_file *m, void *v);
struct file_operations *seq_fops;
};
struct udp_iter_state {
sa_family_t family;
+ struct hlist_head *hashtable;
int bucket;
struct seq_operations seq_ops;
};
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index 77e265d..9bbb2b2 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -92,10 +92,8 @@ #include <linux/errno.h>
#include <linux/timer.h>
#include <linux/mm.h>
#include <linux/inet.h>
-#include <linux/ipv6.h>
#include <linux/netdevice.h>
#include <net/snmp.h>
-#include <net/ip.h>
#include <net/tcp_states.h>
#include <net/protocol.h>
#include <linux/skbuff.h>
@@ -120,26 +118,29 @@ DEFINE_RWLOCK(udp_hash_lock);
static int udp_port_rover;
-static inline int udp_lport_inuse(u16 num)
+static inline int __udp_lib_lport_inuse(u16 num, struct hlist_head udptable[])
{
struct sock *sk;
struct hlist_node *node;
- sk_for_each(sk, node, &udp_hash[num & (UDP_HTABLE_SIZE - 1)])
+ sk_for_each(sk, node, &udptable[num & (UDP_HTABLE_SIZE - 1)])
if (inet_sk(sk)->num == num)
return 1;
return 0;
}
/**
- * udp_get_port - common port lookup for IPv4 and IPv6
+ * __udp_lib_get_port - UDP/-Lite port lookup for IPv4 and IPv6
*
* @sk: socket struct in question
* @snum: port number to look up
+ * @udptable: hash list table, must be of UDP_HTABLE_SIZE
+ * @port_rover: pointer to record of last unallocated port
* @saddr_comp: AF-dependent comparison of bound local IP addresses
*/
-int udp_get_port(struct sock *sk, unsigned short snum,
- int (*saddr_cmp)(const struct sock *sk1, const struct sock
*sk2))
+static int __udp_lib_get_port(struct sock *sk, unsigned short snum,
+ struct hlist_head udptable[], int *port_rover,
+ int (*saddr_cmp)(const struct sock *, const struct sock *))
{
struct hlist_node *node;
struct hlist_head *head;
@@ -150,15 +151,15 @@ int udp_get_port(struct sock *sk, unsign
if (snum == 0) {
int best_size_so_far, best, result, i;
- if (udp_port_rover > sysctl_local_port_range[1] ||
- udp_port_rover < sysctl_local_port_range[0])
- udp_port_rover = sysctl_local_port_range[0];
+ if (*port_rover > sysctl_local_port_range[1] ||
+ *port_rover < sysctl_local_port_range[0])
+ *port_rover = sysctl_local_port_range[0];
best_size_so_far = 32767;
- best = result = udp_port_rover;
+ best = result = *port_rover;
for (i = 0; i < UDP_HTABLE_SIZE; i++, result++) {
int size;
- head = &udp_hash[result & (UDP_HTABLE_SIZE - 1)];
+ head = &udptable[result & (UDP_HTABLE_SIZE - 1)];
if (hlist_empty(head)) {
if (result > sysctl_local_port_range[1])
result = sysctl_local_port_range[0] +
@@ -179,15 +180,15 @@ int udp_get_port(struct sock *sk, unsign
result = sysctl_local_port_range[0]
+ ((result -
sysctl_local_port_range[0]) &
(UDP_HTABLE_SIZE - 1));
- if (!udp_lport_inuse(result))
+ if (! __udp_lib_lport_inuse(result, udptable))
break;
}
if (i >= (1 << 16) / UDP_HTABLE_SIZE)
goto fail;
gotit:
- udp_port_rover = snum = result;
+ *port_rover = snum = result;
} else {
- head = &udp_hash[snum & (UDP_HTABLE_SIZE - 1)];
+ head = &udptable[snum & (UDP_HTABLE_SIZE - 1)];
sk_for_each(sk2, node, head)
if (inet_sk(sk2)->num == snum &&
@@ -200,7 +201,7 @@ gotit:
}
inet_sk(sk)->num = snum;
if (sk_unhashed(sk)) {
- head = &udp_hash[snum & (UDP_HTABLE_SIZE - 1)];
+ head = &udptable[snum & (UDP_HTABLE_SIZE - 1)];
sk_add_node(sk, head);
sock_prot_inc_use(sk->sk_prot);
}
@@ -210,6 +211,12 @@ fail:
return error;
}
+__inline__ int udp_get_port(struct sock *sk, unsigned short snum,
+ int (*scmp)(const struct sock *, const struct sock *))
+{
+ return __udp_lib_get_port(sk, snum, udp_hash, &udp_port_rover, scmp);
+}
+
static inline int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct
sock *sk2)
{
struct inet_sock *inet1 = inet_sk(sk1), *inet2 = inet_sk(sk2);
@@ -225,33 +232,20 @@ static inline int udp_v4_get_port(struct
}
-static void udp_v4_hash(struct sock *sk)
-{
- BUG();
-}
-
-static void udp_v4_unhash(struct sock *sk)
-{
- write_lock_bh(&udp_hash_lock);
- if (sk_del_node_init(sk)) {
- inet_sk(sk)->num = 0;
- sock_prot_dec_use(sk->sk_prot);
- }
- write_unlock_bh(&udp_hash_lock);
-}
-
/* UDP is nearly always wildcards out the wazoo, it makes no sense to try
* harder than this. -DaveM
*/
-static struct sock *udp_v4_lookup_longway(u32 saddr, u16 sport,
- u32 daddr, u16 dport, int dif)
+static struct sock *__udp4_lib_lookup(u32 saddr, u16 sport,
+ u32 daddr, u16 dport,
+ int dif, struct hlist_head udptable[])
{
struct sock *sk, *result = NULL;
struct hlist_node *node;
unsigned short hnum = ntohs(dport);
int badness = -1;
- sk_for_each(sk, node, &udp_hash[hnum & (UDP_HTABLE_SIZE - 1)]) {
+ read_lock(&udp_hash_lock);
+ sk_for_each(sk, node, &udptable[hnum & (UDP_HTABLE_SIZE - 1)]) {
struct inet_sock *inet = inet_sk(sk);
if (inet->num == hnum && !ipv6_only_sock(sk)) {
@@ -285,20 +279,17 @@ static struct sock *udp_v4_lookup_longwa
}
}
}
+ if (result)
+ sock_hold(result);
+ read_unlock(&udp_hash_lock);
+
return result;
}
static __inline__ struct sock *udp_v4_lookup(u32 saddr, u16 sport,
u32 daddr, u16 dport, int dif)
{
- struct sock *sk;
-
- read_lock(&udp_hash_lock);
- sk = udp_v4_lookup_longway(saddr, sport, daddr, dport, dif);
- if (sk)
- sock_hold(sk);
- read_unlock(&udp_hash_lock);
- return sk;
+ return __udp4_lib_lookup(saddr, sport, daddr, dport, dif, udp_hash);
}
static inline struct sock *udp_v4_mcast_next(struct sock *sk,
@@ -340,7 +331,8 @@ found:
* to find the appropriate port.
*/
-void udp_err(struct sk_buff *skb, u32 info)
+static void __udp4_lib_err(struct sk_buff *skb, u32 info,
+ struct hlist_head udptable[] )
{
struct inet_sock *inet;
struct iphdr *iph = (struct iphdr*)skb->data;
@@ -351,7 +343,8 @@ void udp_err(struct sk_buff *skb, u32 in
int harderr;
int err;
- sk = udp_v4_lookup(iph->daddr, uh->dest, iph->saddr, uh->source,
skb->dev->ifindex);
+ sk = __udp4_lib_lookup(iph->daddr, uh->dest, iph->saddr, uh->source,
+ skb->dev->ifindex, udptable );
if (sk == NULL) {
ICMP_INC_STATS_BH(ICMP_MIB_INERRORS);
return; /* No socket for error */
@@ -405,6 +398,11 @@ out:
sock_put(sk);
}
+__inline__ void udp_err(struct sk_buff *skb, u32 info)
+{
+ return __udp4_lib_err(skb, info, udp_hash);
+}
+
/*
* Throw away all pending data and cancel the corking. Socket is locked.
*/
@@ -419,6 +417,45 @@ static void udp_flush_pending_frames(str
}
}
+/**
+ * udp4_hwcsum_outgoing - handle outgoing HW checksumming
+ * @sk: socket we are sending on
+ * @skb: sk_buff containing the filled-in UDP header
+ * (checksum field must be zeroed out)
+ */
+static void udp4_hwcsum_outgoing(struct sock *sk, struct sk_buff *skb,
+ u32 src, u32 dst, int len )
+{
+ unsigned int csum = 0, offset;
+ struct udphdr *uh = skb->h.uh;
+
+ if (skb_queue_len(&sk->sk_write_queue) == 1) {
+ /*
+ * Only one fragment on the socket.
+ */
+ skb->csum = offsetof(struct udphdr, check);
+ uh->check = ~csum_tcpudp_magic(src, dst, len, IPPROTO_UDP, 0);
+ } else {
+ /*
+ * HW-checksum won't work as there are two or more
+ * fragments on the socket so that all csums of sk_buffs
+ * should be together
+ */
+ offset = skb->h.raw - skb->data;
+ skb->csum = skb_checksum(skb, offset, skb->len - offset, 0);
+
+ skb->ip_summed = CHECKSUM_NONE;
+
+ skb_queue_walk(&sk->sk_write_queue, skb) {
+ csum = csum_add(csum, skb->csum);
+ }
+
+ uh->check = csum_tcpudp_magic(src, dst, len, IPPROTO_UDP, csum);
+ if (uh->check == 0)
+ uh->check = -1;
+ }
+}
+
/*
* Push out all pending data as one UDP datagram. Socket is locked.
*/
@@ -429,6 +466,7 @@ static int udp_push_pending_frames(struc
struct sk_buff *skb;
struct udphdr *uh;
int err = 0;
+ u32 csum = 0;
/* Grab the skbuff where UDP header space exists. */
if ((skb = skb_peek(&sk->sk_write_queue)) == NULL)
@@ -443,52 +481,31 @@ static int udp_push_pending_frames(struc
uh->len = htons(up->len);
uh->check = 0;
- if (sk->sk_no_check == UDP_CSUM_NOXMIT) {
+ if (up->pcflag) { /* UDP-Lite */
+ int cscov = udplite_sender_cscov(up, uh);
+
+ csum = udplite_csum_outgoing(sk, cscov);
+ skb->ip_summed = CHECKSUM_NONE;
+
+ } else if (sk->sk_no_check == UDP_CSUM_NOXMIT) { /* UDP csum disabled */
+
skb->ip_summed = CHECKSUM_NONE;
goto send;
- }
- if (skb_queue_len(&sk->sk_write_queue) == 1) {
- /*
- * Only one fragment on the socket.
- */
- if (skb->ip_summed == CHECKSUM_PARTIAL) {
- skb->csum = offsetof(struct udphdr, check);
- uh->check = ~csum_tcpudp_magic(fl->fl4_src, fl->fl4_dst,
- up->len, IPPROTO_UDP, 0);
- } else {
- skb->csum = csum_partial((char *)uh,
- sizeof(struct udphdr), skb->csum);
- uh->check = csum_tcpudp_magic(fl->fl4_src, fl->fl4_dst,
- up->len, IPPROTO_UDP, skb->csum);
- if (uh->check == 0)
- uh->check = -1;
- }
- } else {
- unsigned int csum = 0;
- /*
- * HW-checksum won't work as there are two or more
- * fragments on the socket so that all csums of sk_buffs
- * should be together.
- */
- if (skb->ip_summed == CHECKSUM_PARTIAL) {
- int offset = (unsigned char *)uh - skb->data;
- skb->csum = skb_checksum(skb, offset, skb->len -
offset, 0);
+ } else if (skb->ip_summed == CHECKSUM_PARTIAL) { /* UDP hardware csum */
- skb->ip_summed = CHECKSUM_NONE;
- } else {
- skb->csum = csum_partial((char *)uh,
- sizeof(struct udphdr), skb->csum);
- }
+ udp4_hwcsum_outgoing(sk, skb, fl->fl4_src,fl->fl4_dst, up->len);
+ goto send;
+
+ } else /* `normal' UDP */
+ csum = udp_csum_outgoing(sk, skb);
+
+ /* add protocol-dependent pseudo-header */
+ uh->check = csum_tcpudp_magic(fl->fl4_src, fl->fl4_dst, up->len,
+ sk->sk_protocol, csum );
+ if (uh->check == 0)
+ uh->check = -1;
- skb_queue_walk(&sk->sk_write_queue, skb) {
- csum = csum_add(csum, skb->csum);
- }
- uh->check = csum_tcpudp_magic(fl->fl4_src, fl->fl4_dst,
- up->len, IPPROTO_UDP, csum);
- if (uh->check == 0)
- uh->check = -1;
- }
send:
err = ip_push_pending_frames(sk);
out:
@@ -497,12 +514,6 @@ out:
return err;
}
-
-static unsigned short udp_check(struct udphdr *uh, int len, unsigned long
saddr, unsigned long daddr, unsigned long base)
-{
- return(csum_tcpudp_magic(saddr, daddr, len, IPPROTO_UDP, base));
-}
-
int udp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
size_t len)
{
@@ -516,8 +527,9 @@ int udp_sendmsg(struct kiocb *iocb, stru
u32 daddr, faddr, saddr;
u16 dport;
u8 tos;
- int err;
+ int err, is_udplite = up->pcflag;
int corkreq = up->corkflag || msg->msg_flags&MSG_MORE;
+ int (*getfrag)(void *, char *, int, int, int, struct sk_buff *);
if (len > 0xFFFF)
return -EMSGSIZE;
@@ -622,7 +634,7 @@ int udp_sendmsg(struct kiocb *iocb, stru
{ .daddr = faddr,
.saddr = saddr,
.tos = tos } },
- .proto = IPPROTO_UDP,
+ .proto = sk->sk_protocol,
.uli_u = { .ports =
{ .sport = inet->sport,
.dport = dport } } };
@@ -668,8 +680,9 @@ back_from_confirm:
do_append_data:
up->len += ulen;
- err = ip_append_data(sk, ip_generic_getfrag, msg->msg_iov, ulen,
- sizeof(struct udphdr), &ipc, rt,
+ getfrag = is_udplite ? udplite_getfrag : ip_generic_getfrag;
+ err = ip_append_data(sk, getfrag, msg->msg_iov, ulen,
+ sizeof(struct udphdr), &ipc, rt,
corkreq ? msg->msg_flags|MSG_MORE : msg->msg_flags);
if (err)
udp_flush_pending_frames(sk);
@@ -682,7 +695,7 @@ out:
if (free)
kfree(ipc.opt);
if (!err) {
- UDP_INC_STATS_USER(UDP_MIB_OUTDATAGRAMS);
+ UDP_INC_STATS_USER(UDP_MIB_OUTDATAGRAMS, is_udplite);
return len;
}
/*
@@ -693,7 +706,7 @@ out:
* seems like overkill.
*/
if (err == -ENOBUFS || test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) {
- UDP_INC_STATS_USER(UDP_MIB_SNDBUFERRORS);
+ UDP_INC_STATS_USER(UDP_MIB_SNDBUFERRORS, is_udplite);
}
return err;
@@ -793,17 +806,6 @@ int udp_ioctl(struct sock *sk, int cmd,
return(0);
}
-static __inline__ int __udp_checksum_complete(struct sk_buff *skb)
-{
- return __skb_checksum_complete(skb);
-}
-
-static __inline__ int udp_checksum_complete(struct sk_buff *skb)
-{
- return skb->ip_summed != CHECKSUM_UNNECESSARY &&
- __udp_checksum_complete(skb);
-}
-
/*
* This should be easy, if there is something there we
* return it, otherwise we block.
@@ -815,7 +817,7 @@ static int udp_recvmsg(struct kiocb *ioc
struct inet_sock *inet = inet_sk(sk);
struct sockaddr_in *sin = (struct sockaddr_in *)msg->msg_name;
struct sk_buff *skb;
- int copied, err;
+ int copied, err, copy_only, is_udplite = IS_UDPLITE(sk);
/*
* Check any passed addresses
@@ -837,15 +839,25 @@ try_again:
msg->msg_flags |= MSG_TRUNC;
}
- if (skb->ip_summed==CHECKSUM_UNNECESSARY) {
- err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
msg->msg_iov,
- copied);
- } else if (msg->msg_flags&MSG_TRUNC) {
- if (__udp_checksum_complete(skb))
+ /*
+ * Decide whether to checksum and/or copy data.
+ *
+ * UDP: checksum may have been computed in HW,
+ * (re-)compute it if message is truncated.
+ * UDP-Lite: always needs to checksum, no HW support.
+ */
+ copy_only = (skb->ip_summed==CHECKSUM_UNNECESSARY);
+
+ if (is_udplite || (!copy_only && msg->msg_flags&MSG_TRUNC)) {
+ if (__udp_lib_checksum_complete(skb))
goto csum_copy_err;
- err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
msg->msg_iov,
- copied);
- } else {
+ copy_only = 1;
+ }
+
+ if (copy_only)
+ err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
+ msg->msg_iov, copied );
+ else {
err = skb_copy_and_csum_datagram_iovec(skb, sizeof(struct
udphdr), msg->msg_iov);
if (err == -EINVAL)
@@ -878,7 +890,8 @@ out:
return err;
csum_copy_err:
- UDP_INC_STATS_BH(UDP_MIB_INERRORS);
+ UDP_INC_STATS_BH(UDP_MIB_INERRORS, is_udplite);
+ UDP_DEC_STATS_BH(UDP_MIB_INDATAGRAMS, is_udplite);
skb_kill_datagram(sk, skb, flags);
@@ -1019,10 +1032,8 @@ static int udp_queue_rcv_skb(struct sock
/*
* Charge it to the socket, dropping if the queue is full.
*/
- if (!xfrm4_policy_check(sk, XFRM_POLICY_IN, skb)) {
- kfree_skb(skb);
- return -1;
- }
+ if (!xfrm4_policy_check(sk, XFRM_POLICY_IN, skb))
+ goto drop;
nf_reset(skb);
if (up->encap_type) {
@@ -1046,31 +1057,77 @@ static int udp_queue_rcv_skb(struct sock
if (ret < 0) {
/* process the ESP packet */
ret = xfrm4_rcv_encap(skb, up->encap_type);
- UDP_INC_STATS_BH(UDP_MIB_INDATAGRAMS);
+ UDP_INC_STATS_BH(UDP_MIB_INDATAGRAMS, up->pcflag);
return -ret;
}
/* FALLTHROUGH -- it's a UDP Packet */
}
- if (sk->sk_filter && skb->ip_summed != CHECKSUM_UNNECESSARY) {
- if (__udp_checksum_complete(skb)) {
- UDP_INC_STATS_BH(UDP_MIB_INERRORS);
- kfree_skb(skb);
- return -1;
+ /*
+ * UDP-Lite specific tests, ignored on UDP sockets
+ */
+ if ((up->pcflag & UDPLITE_RECV_CC) && UDP_SKB_CB(skb)->partial_cov) {
+
+ /*
+ * MIB statistics other than incrementing the error count are
+ * disabled for the following two types of errors: these depend
+ * on the application settings, not on the functioning of the
+ * protocol stack as such.
+ *
+ *
+ * RFC 3828 here recommends (sec 3.3): "There should also be a
+ * way ... to ... at least let the receiving application block
+ * delivery of packets with coverage values less than a value
+ * provided by the application."
+ */
+ if (up->pcrlen == 0) { /* full coverage was set */
+ LIMIT_NETDEBUG(KERN_WARNING "UDPLITE: partial coverage "
+ "%d while full coverage %d requested\n",
+ UDP_SKB_CB(skb)->cscov, skb->len);
+ goto drop;
}
+ /* The next case involves violating the min. coverage requested
+ * by the receiver. This is subtle: if receiver wants x and x is
+ * greater than the buffersize/MTU then receiver will complain
+ * that it wants x while sender emits packets of smaller size y.
+ * Therefore the above ...()->partial_cov statement is
essential.
+ */
+ if (UDP_SKB_CB(skb)->cscov < up->pcrlen) {
+ LIMIT_NETDEBUG(KERN_WARNING
+ "UDPLITE: coverage %d too small, need min %d\n",
+ UDP_SKB_CB(skb)->cscov, up->pcrlen);
+ goto drop;
+ }
+ }
+
+ if (sk->sk_filter && skb->ip_summed != CHECKSUM_UNNECESSARY) {
+ if (__udp_lib_checksum_complete(skb))
+ goto drop;
skb->ip_summed = CHECKSUM_UNNECESSARY;
}
if ((rc = sock_queue_rcv_skb(sk,skb)) < 0) {
/* Note that an ENOMEM error is charged twice */
if (rc == -ENOMEM)
- UDP_INC_STATS_BH(UDP_MIB_RCVBUFERRORS);
- UDP_INC_STATS_BH(UDP_MIB_INERRORS);
- kfree_skb(skb);
- return -1;
+ UDP_INC_STATS_BH(UDP_MIB_RCVBUFERRORS, up->pcflag);
+ goto drop;
}
- UDP_INC_STATS_BH(UDP_MIB_INDATAGRAMS);
+
+ /*
+ * XXX Incrementing this counter when the datagram is later taken
+ * off the queue due to receive failure is problematic, cf.
+ * http://bugzilla.kernel.org/show_bug.cgi?id=6660
+ * This module counts correctly by decrementing InDatagrams
+ * whenever the datagram is popped off a queue without being
+ * actually delivered; see also udp_recvmsg() and udp_poll().
+ */
+ UDP_INC_STATS_BH(UDP_MIB_INDATAGRAMS, up->pcflag);
return 0;
+
+drop:
+ UDP_INC_STATS_BH(UDP_MIB_INERRORS, up->pcflag);
+ kfree_skb(skb);
+ return -1;
}
/*
@@ -1079,14 +1136,14 @@ static int udp_queue_rcv_skb(struct sock
* Note: called only from the BH handler context,
* so we don't need to lock the hashes.
*/
-static int udp_v4_mcast_deliver(struct sk_buff *skb, struct udphdr *uh,
- u32 saddr, u32 daddr)
+static int __udp4_lib_mcast_deliver(struct sk_buff *skb, struct udphdr *uh,
+ u32 saddr, u32 daddr, struct hlist_head udptable[])
{
struct sock *sk;
int dif;
read_lock(&udp_hash_lock);
- sk = sk_head(&udp_hash[ntohs(uh->dest) & (UDP_HTABLE_SIZE - 1)]);
+ sk = sk_head(&udptable[ntohs(uh->dest) & (UDP_HTABLE_SIZE - 1)]);
dif = skb->dev->ifindex;
sk = udp_v4_mcast_next(sk, uh->dest, daddr, uh->source, saddr, dif);
if (sk) {
@@ -1115,6 +1172,12 @@ static int udp_v4_mcast_deliver(struct s
return 0;
}
+static __inline__ int udp_v4_mcast_deliver(struct sk_buff *skb,
+ struct udphdr *uh, u32 saddr, u32 daddr)
+{
+ return __udp4_lib_mcast_deliver(skb, uh, saddr, daddr, udp_hash);
+}
+
/* Initialize UDP checksum. If exited with zero value (success),
* CHECKSUM_UNNECESSARY means, that no more checks are required.
* Otherwise, csum completion requires chacksumming packet body,
@@ -1126,7 +1189,7 @@ static void udp_checksum_init(struct sk_
if (uh->check == 0) {
skb->ip_summed = CHECKSUM_UNNECESSARY;
} else if (skb->ip_summed == CHECKSUM_COMPLETE) {
- if (!udp_check(uh, ulen, saddr, daddr, skb->csum))
+ if (!csum_tcpudp_magic(saddr,daddr,ulen, IPPROTO_UDP, skb->csum))
skb->ip_summed = CHECKSUM_UNNECESSARY;
}
if (skb->ip_summed != CHECKSUM_UNNECESSARY)
@@ -1134,16 +1197,19 @@ static void udp_checksum_init(struct sk_
/* Probably, we should checksum udp header (it should be in cache
* in any case) and data in tiny packets (< rx copybreak).
*/
+
+ /* UDP = UDP-Lite with a non-partial checksum coverage */
+ UDP_SKB_CB(skb)->partial_cov = 0;
}
/*
- * All we need to do is get the socket, and then do a checksum.
+ * All we need to do is get the socket, and then do a checksum.
*/
-
-int udp_rcv(struct sk_buff *skb)
+static int __udp4_lib_rcv(struct sk_buff *skb,
+ struct hlist_head udptable[], int is_udplite)
{
struct sock *sk;
- struct udphdr *uh;
+ struct udphdr *uh = skb->h.uh;
unsigned short ulen;
struct rtable *rt = (struct rtable*)skb->dst;
u32 saddr = skb->nh.iph->saddr;
@@ -1151,34 +1217,39 @@ int udp_rcv(struct sk_buff *skb)
int len = skb->len;
/*
- * Validate the packet and the UDP length.
+ * Validate the packet.
*/
if (!pskb_may_pull(skb, sizeof(struct udphdr)))
- goto no_header;
-
- uh = skb->h.uh;
+ goto drop; /* No space for header. */
ulen = ntohs(uh->len);
-
- if (ulen > len || ulen < sizeof(*uh))
+ if (ulen > len)
goto short_packet;
- if (pskb_trim_rcsum(skb, ulen))
- goto short_packet;
+ if(! is_udplite ) { /* UDP validates ulen. */
+
+ if ( ulen < sizeof(*uh) || pskb_trim_rcsum(skb, ulen) )
+ goto short_packet;
+ /* note the difference: UDP uses ulen, UDP-Lite uses len */
+ udp_checksum_init(skb, uh, ulen, saddr, daddr);
- udp_checksum_init(skb, uh, ulen, saddr, daddr);
+ } else { /* UDP-Lite validates cscov. */
+ if (! udplite_checksum_init(skb, uh, len, saddr, daddr) )
+ goto csum_error;
+ }
if(rt->rt_flags & (RTCF_BROADCAST|RTCF_MULTICAST))
- return udp_v4_mcast_deliver(skb, uh, saddr, daddr);
+ return __udp4_lib_mcast_deliver(skb, uh, saddr, daddr,
udptable);
- sk = udp_v4_lookup(saddr, uh->source, daddr, uh->dest,
skb->dev->ifindex);
+ sk = __udp4_lib_lookup(saddr, uh->source, daddr, uh->dest,
+ skb->dev->ifindex, udptable );
if (sk != NULL) {
int ret = udp_queue_rcv_skb(sk, skb);
sock_put(sk);
/* a return value > 0 means to resubmit the input, but
- * it it wants the return to be -protocol, or 0
+ * it wants the return to be -protocol, or 0
*/
if (ret > 0)
return -ret;
@@ -1193,7 +1264,7 @@ int udp_rcv(struct sk_buff *skb)
if (udp_checksum_complete(skb))
goto csum_error;
- UDP_INC_STATS_BH(UDP_MIB_NOPORTS);
+ UDP_INC_STATS_BH(UDP_MIB_NOPORTS, is_udplite);
icmp_send(skb, ICMP_DEST_UNREACH, ICMP_PORT_UNREACH, 0);
/*
@@ -1204,35 +1275,39 @@ int udp_rcv(struct sk_buff *skb)
return(0);
short_packet:
- LIMIT_NETDEBUG(KERN_DEBUG "UDP: short packet: From %u.%u.%u.%u:%u %d/%d
to %u.%u.%u.%u:%u\n",
+ LIMIT_NETDEBUG(KERN_DEBUG "UDP%s: short packet: From %u.%u.%u.%u:%u
%d/%d to %u.%u.%u.%u:%u\n",
+ is_udplite? "-Lite" : "",
NIPQUAD(saddr),
ntohs(uh->source),
ulen,
len,
NIPQUAD(daddr),
ntohs(uh->dest));
-no_header:
- UDP_INC_STATS_BH(UDP_MIB_INERRORS);
- kfree_skb(skb);
- return(0);
+ goto drop;
csum_error:
- /*
- * RFC1122: OK. Discards the bad packet silently (as far as
- * the network is concerned, anyway) as per 4.1.3.4 (MUST).
+ /*
+ * RFC1122: OK. Discards the bad packet silently (as far as
+ * the network is concerned, anyway) as per 4.1.3.4 (MUST).
*/
- LIMIT_NETDEBUG(KERN_DEBUG "UDP: bad checksum. From %d.%d.%d.%d:%d to
%d.%d.%d.%d:%d ulen %d\n",
+ LIMIT_NETDEBUG(KERN_DEBUG "UDP%s: bad checksum. From %d.%d.%d.%d:%d to
%d.%d.%d.%d:%d ulen %d\n",
+ is_udplite? "-Lite" : "",
NIPQUAD(saddr),
ntohs(uh->source),
NIPQUAD(daddr),
ntohs(uh->dest),
ulen);
drop:
- UDP_INC_STATS_BH(UDP_MIB_INERRORS);
+ UDP_INC_STATS_BH(UDP_MIB_INERRORS, is_udplite);
kfree_skb(skb);
return(0);
}
+__inline__ int udp_rcv(struct sk_buff *skb)
+{
+ return __udp4_lib_rcv(skb, udp_hash, 0);
+}
+
static int udp_destroy_sock(struct sock *sk)
{
lock_sock(sk);
@@ -1282,6 +1357,32 @@ static int do_udp_setsockopt(struct sock
}
break;
+ /*
+ * UDP-Lite's partial checksum coverage (RFC 3828).
+ */
+ /* The sender sets actual checksum coverage length via this option.
+ * The case coverage > packet length is handled by send module. */
+ case UDPLITE_SEND_CSCOV:
+ if (!up->pcflag) /* Disable the option on UDP sockets */
+ return -ENOPROTOOPT;
+ if (val != 0 && val < 8) /* Illegal coverage: use default (8) */
+ val = 8;
+ up->pcslen = val;
+ up->pcflag |= UDPLITE_SEND_CC;
+ break;
+
+ /* The receiver specifies a minimum checksum coverage value. To make
+ * sense, this should be set to at least 8 (as done below). If zero is
+ * used, this again means full checksum coverage. */
+ case UDPLITE_RECV_CSCOV:
+ if (!up->pcflag) /* Disable the option on UDP sockets */
+ return -ENOPROTOOPT;
+ if (val != 0 && val < 8) /* Avoid silly minimal values. */
+ val = 8;
+ up->pcrlen = val;
+ up->pcflag |= UDPLITE_RECV_CC;
+ break;
+
default:
err = -ENOPROTOOPT;
break;
@@ -1293,18 +1394,18 @@ static int do_udp_setsockopt(struct sock
static int udp_setsockopt(struct sock *sk, int level, int optname,
char __user *optval, int optlen)
{
- if (level != SOL_UDP)
- return ip_setsockopt(sk, level, optname, optval, optlen);
- return do_udp_setsockopt(sk, level, optname, optval, optlen);
+ if (level == SOL_UDP || level == SOL_UDPLITE)
+ return do_udp_setsockopt(sk, level, optname, optval, optlen);
+ return ip_setsockopt(sk, level, optname, optval, optlen);
}
#ifdef CONFIG_COMPAT
static int compat_udp_setsockopt(struct sock *sk, int level, int optname,
char __user *optval, int optlen)
{
- if (level != SOL_UDP)
- return compat_ip_setsockopt(sk, level, optname, optval, optlen);
- return do_udp_setsockopt(sk, level, optname, optval, optlen);
+ if (level == SOL_UDP || level == SOL_UDPLITE)
+ return do_udp_setsockopt(sk, level, optname, optval, optlen);
+ return compat_ip_setsockopt(sk, level, optname, optval, optlen);
}
#endif
@@ -1331,6 +1432,15 @@ static int do_udp_getsockopt(struct sock
val = up->encap_type;
break;
+ /* the following two always return 0 on UDP sockets */
+ case UDPLITE_SEND_CSCOV:
+ val = up->pcslen;
+ break;
+
+ case UDPLITE_RECV_CSCOV:
+ val = up->pcrlen;
+ break;
+
default:
return -ENOPROTOOPT;
};
@@ -1345,18 +1455,18 @@ static int do_udp_getsockopt(struct sock
static int udp_getsockopt(struct sock *sk, int level, int optname,
char __user *optval, int __user *optlen)
{
- if (level != SOL_UDP)
- return ip_getsockopt(sk, level, optname, optval, optlen);
- return do_udp_getsockopt(sk, level, optname, optval, optlen);
+ if (level == SOL_UDP || level == SOL_UDPLITE)
+ return do_udp_getsockopt(sk, level, optname, optval, optlen);
+ return ip_getsockopt(sk, level, optname, optval, optlen);
}
#ifdef CONFIG_COMPAT
static int compat_udp_getsockopt(struct sock *sk, int level, int optname,
char __user *optval, int __user *optlen)
{
- if (level != SOL_UDP)
- return compat_ip_getsockopt(sk, level, optname, optval, optlen);
- return do_udp_getsockopt(sk, level, optname, optval, optlen);
+ if (level == SOL_UDP || level == SOL_UDPLITE)
+ return do_udp_getsockopt(sk, level, optname, optval, optlen);
+ return compat_ip_getsockopt(sk, level, optname, optval, optlen);
}
#endif
/**
@@ -1376,7 +1486,8 @@ unsigned int udp_poll(struct file *file,
{
unsigned int mask = datagram_poll(file, sock, wait);
struct sock *sk = sock->sk;
-
+ int is_lite = IS_UDPLITE(sk);
+
/* Check for false positives due to checksum errors */
if ( (mask & POLLRDNORM) &&
!(file->f_flags & O_NONBLOCK) &&
@@ -1387,7 +1498,11 @@ unsigned int udp_poll(struct file *file,
spin_lock_bh(&rcvq->lock);
while ((skb = skb_peek(rcvq)) != NULL) {
if (udp_checksum_complete(skb)) {
- UDP_INC_STATS_BH(UDP_MIB_INERRORS);
+ /* The datagram has already been counted as
+ * InDatagram when earlier it was enqueued.
+ * Update count of really received datagrams. */
+ UDP_DEC_STATS_BH(UDP_MIB_INDATAGRAMS, is_lite);
+ UDP_INC_STATS_BH(UDP_MIB_INERRORS, is_lite);
__skb_unlink(skb, rcvq);
kfree_skb(skb);
} else {
@@ -1420,8 +1535,8 @@ struct proto udp_prot = {
.recvmsg = udp_recvmsg,
.sendpage = udp_sendpage,
.backlog_rcv = udp_queue_rcv_skb,
- .hash = udp_v4_hash,
- .unhash = udp_v4_unhash,
+ .hash = udp_lib_hash,
+ .unhash = udp_lib_unhash,
.get_port = udp_v4_get_port,
.obj_size = sizeof(struct udp_sock),
#ifdef CONFIG_COMPAT
@@ -1440,7 +1555,7 @@ static struct sock *udp_get_first(struct
for (state->bucket = 0; state->bucket < UDP_HTABLE_SIZE;
++state->bucket) {
struct hlist_node *node;
- sk_for_each(sk, node, &udp_hash[state->bucket]) {
+ sk_for_each(sk, node, state->hashtable + state->bucket) {
if (sk->sk_family == state->family)
goto found;
}
@@ -1461,7 +1576,7 @@ try_again:
} while (sk && sk->sk_family != state->family);
if (!sk && ++state->bucket < UDP_HTABLE_SIZE) {
- sk = sk_head(&udp_hash[state->bucket]);
+ sk = sk_head(state->hashtable + state->bucket);
goto try_again;
}
return sk;
@@ -1511,6 +1626,7 @@ static int udp_seq_open(struct inode *in
if (!s)
goto out;
s->family = afinfo->family;
+ s->hashtable = afinfo->hashtable;
s->seq_ops.start = udp_seq_start;
s->seq_ops.next = udp_seq_next;
s->seq_ops.show = afinfo->seq_show;
@@ -1577,7 +1693,7 @@ static void udp4_format_sock(struct sock
atomic_read(&sp->sk_refcnt), sp);
}
-static int udp4_seq_show(struct seq_file *seq, void *v)
+int udp4_seq_show(struct seq_file *seq, void *v)
{
if (v == SEQ_START_TOKEN)
seq_printf(seq, "%-127s\n",
@@ -1600,6 +1716,7 @@ static struct udp_seq_afinfo udp4_seq_af
.owner = THIS_MODULE,
.name = "udp",
.family = AF_INET,
+ .hashtable = udp_hash,
.seq_show = udp4_seq_show,
.seq_fops = &udp4_seq_fops,
};
@@ -1628,3 +1745,5 @@ #ifdef CONFIG_PROC_FS
EXPORT_SYMBOL(udp_proc_register);
EXPORT_SYMBOL(udp_proc_unregister);
#endif
+/* the extensions for UDP-Lite (RFC 3828) */
+#include "udplite.c"
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 9662561..cf6cd7a 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -66,21 +66,6 @@ static inline int udp_v6_get_port(struct
return udp_get_port(sk, snum, ipv6_rcv_saddr_equal);
}
-static void udp_v6_hash(struct sock *sk)
-{
- BUG();
-}
-
-static void udp_v6_unhash(struct sock *sk)
-{
- write_lock_bh(&udp_hash_lock);
- if (sk_del_node_init(sk)) {
- inet_sk(sk)->num = 0;
- sock_prot_dec_use(sk->sk_prot);
- }
- write_unlock_bh(&udp_hash_lock);
-}
-
static struct sock *udp_v6_lookup(struct in6_addr *saddr, u16 sport,
struct in6_addr *daddr, u16 dport, int dif)
{
@@ -485,6 +470,7 @@ static int udp_v6_push_pending_frames(st
struct inet_sock *inet = inet_sk(sk);
struct flowi *fl = &inet->cork.fl;
int err = 0;
+ u32 csum;
/* Grab the skbuff where UDP header space exists. */
if ((skb = skb_peek(&sk->sk_write_queue)) == NULL)
@@ -499,35 +485,12 @@ static int udp_v6_push_pending_frames(st
uh->len = htons(up->len);
uh->check = 0;
- if (sk->sk_no_check == UDP_CSUM_NOXMIT) {
- skb->ip_summed = CHECKSUM_NONE;
- goto send;
- }
-
- if (skb_queue_len(&sk->sk_write_queue) == 1) {
- skb->csum = csum_partial((char *)uh,
- sizeof(struct udphdr), skb->csum);
- uh->check = csum_ipv6_magic(&fl->fl6_src,
- &fl->fl6_dst,
- up->len, fl->proto, skb->csum);
- } else {
- u32 tmp_csum = 0;
-
- skb_queue_walk(&sk->sk_write_queue, skb) {
- tmp_csum = csum_add(tmp_csum, skb->csum);
- }
- tmp_csum = csum_partial((char *)uh,
- sizeof(struct udphdr), tmp_csum);
- tmp_csum = csum_ipv6_magic(&fl->fl6_src,
- &fl->fl6_dst,
- up->len, fl->proto, tmp_csum);
- uh->check = tmp_csum;
-
- }
+ csum = udp_csum_outgoing(sk, skb);
+ uh->check = csum_ipv6_magic(&fl->fl6_src, &fl->fl6_dst,
+ up->len, fl->proto, csum );
if (uh->check == 0)
uh->check = -1;
-send:
err = ip6_push_pending_frames(sk);
out:
up->len = 0;
@@ -1001,6 +964,7 @@ static struct udp_seq_afinfo udp6_seq_af
.owner = THIS_MODULE,
.name = "udp6",
.family = AF_INET6,
+ .hashtable = udp_hash,
.seq_show = udp6_seq_show,
.seq_fops = &udp6_seq_fops,
};
@@ -1030,8 +994,8 @@ struct proto udpv6_prot = {
.sendmsg = udpv6_sendmsg,
.recvmsg = udpv6_recvmsg,
.backlog_rcv = udpv6_queue_rcv_skb,
- .hash = udp_v6_hash,
- .unhash = udp_v6_unhash,
+ .hash = udp_lib_hash,
+ .unhash = udp_lib_unhash,
.get_port = udp_v6_get_port,
.obj_size = sizeof(struct udp6_sock),
#ifdef CONFIG_COMPAT
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
