* [PATCH] new UDPCP Communication Protocol
From: stefani @ 2011-01-01 21:44 UTC (permalink / raw)
To: linux-kernel, akpm, davem, netdev, eric.dumazet, shemminger; +Cc: stefani
From: Stefani Seibold <stefani@seibold.net>
Changelog:
31.12.2010 first proposal
01.01.2011 code cleanup and fixes suggest by Eric Dumazet
UDPCP is a communication protocol specified by the Open Base Station
Architecture Initiative Special Interest Group (OBSAI SIG). The
protocol is based on UDP and is designed to meet the needs of "Mobile
Communcation Base Station" internal communications. It is widely used by
the major networks infrastructure supplier.
The UDPCP communication service supports the following features:
-Connectionless communication for serial mode data transfer
-Acknowledged and unacknowledged transfer modes
-Retransmissions Algorithm
-Checksum Algorithm using Adler32
-Fragmentation of long messages (disassembly/reassembly) to match to the MTU
during transport:
-Broadcasting and multicasting messages to multiple peers in unacknowledged
transfer mode
UDPCP supports application level messages up to 64 KBytes (limited by 16-bit
packet data length field). Messages that are longer than the MTU will be
fragmented to the MTU.
UDPCP provides a reliable transport service that will perform message
retransmissions in case transport failures occur.
The code is also a nice example how to implement a UDP based protocol as
a kernel socket modules.
Due the nature of UDPCP which has no sliding windows support, the latency has a
huge impact. The perfomance increase by implementing as a kernel module is
about the factor 10, because there are no context switches and data packets or
ACKs will be handled in the interrupt service.
There are no side effects to the network subsystems so i ask for merge it
into linux-next. Hope you like it.
The patch is against 2.6.37-rc8
- Stefani
Signed-off-by: Stefani Seibold <stefani@seibold.net>
---
include/linux/socket.h | 7 +-
include/net/udp.h | 1 +
include/net/udpcp.h | 47 +
net/Kconfig | 1 +
net/Makefile | 1 +
net/ipv4/ip_output.c | 2 +
net/ipv4/ip_sockglue.c | 2 +
net/udpcp/Kconfig | 34 +
net/udpcp/Makefile | 5 +
net/udpcp/udpcp.c | 2854 ++++++++++++++++++++++++++++++++++++++++++++++++
10 files changed, 2952 insertions(+), 2 deletions(-)
create mode 100644 include/net/udpcp.h
create mode 100644 net/udpcp/Kconfig
create mode 100644 net/udpcp/Makefile
create mode 100644 net/udpcp/udpcp.c
diff --git a/include/linux/socket.h b/include/linux/socket.h
index 86b652f..755eeb8 100644
--- a/include/linux/socket.h
+++ b/include/linux/socket.h
@@ -193,7 +193,8 @@ struct ucred {
#define AF_PHONET 35 /* Phonet sockets */
#define AF_IEEE802154 36 /* IEEE802154 sockets */
#define AF_CAIF 37 /* CAIF sockets */
-#define AF_MAX 38 /* For now.. */
+#define AF_UDPCP 38 /* UDPCP sockets */
+#define AF_MAX 39 /* For now.. */
/* Protocol families, same as address families. */
#define PF_UNSPEC AF_UNSPEC
@@ -203,7 +204,7 @@ struct ucred {
#define PF_AX25 AF_AX25
#define PF_IPX AF_IPX
#define PF_APPLETALK AF_APPLETALK
-#define PF_NETROM AF_NETROM
+#define PF_NETROM AF_NETROM
#define PF_BRIDGE AF_BRIDGE
#define PF_ATMPVC AF_ATMPVC
#define PF_X25 AF_X25
@@ -234,6 +235,7 @@ struct ucred {
#define PF_PHONET AF_PHONET
#define PF_IEEE802154 AF_IEEE802154
#define PF_CAIF AF_CAIF
+#define PF_UDPCP AF_UDPCP
#define PF_MAX AF_MAX
/* Maximum queue length specifiable by listen. */
@@ -307,6 +309,7 @@ struct ucred {
#define SOL_RDS 276
#define SOL_IUCV 277
#define SOL_CAIF 278
+#define SOL_UDPCP 279
/* IPX options */
#define IPX_TYPE 1
diff --git a/include/net/udp.h b/include/net/udp.h
index bb967dd..82c95a7 100644
--- a/include/net/udp.h
+++ b/include/net/udp.h
@@ -47,6 +47,7 @@ struct udp_skb_cb {
} header;
__u16 cscov;
__u8 partial_cov;
+ __u8 udpcp_flag;
};
#define UDP_SKB_CB(__skb) ((struct udp_skb_cb *)((__skb)->cb))
diff --git a/include/net/udpcp.h b/include/net/udpcp.h
new file mode 100644
index 0000000..45180a5
--- /dev/null
+++ b/include/net/udpcp.h
@@ -0,0 +1,47 @@
+/* Definitions for UDPCP sockets. */
+
+#ifndef __LINUX_IF_UDPCP
+#define __LINUX_IF_UDPCP
+
+#include "linux/ioctl.h"
+
+#define UDPCP_MAX_MSGSIZE 65487
+
+#define UDPCP_MAX_WAIT_SEC 60
+
+#define UDPCP_OPT_TRANSFER_MODE 0
+#define UDPCP_OPT_CHECKSUM_MODE 1
+#define UDPCP_OPT_TX_TIMEOUT 2
+#define UDPCP_OPT_RX_TIMEOUT 3
+#define UDPCP_OPT_MAXTRY 4
+#define UDPCP_OPT_OUTSTANDING_ACKS 5
+
+#define UDPCP_NOACK 0
+#define UDPCP_ACK 1
+#define UDPCP_SINGLE_ACK 2
+#define UDPCP_NOCHECKSUM 3
+#define UDPCP_CHECKSUM 4
+
+#define UDPCP_IOC_MAGIC 251
+
+#define UDPCP_IOCTL_GET_STATISTICS \
+ _IOR(UDPCP_IOC_MAGIC, 0x01, struct udpcp_statistics *)
+#define UDPCP_IOCTL_RESET_STATISTICS \
+ _IO(UDPCP_IOC_MAGIC, 0x02)
+#define UDPCP_IOCTL_SYNC \
+ _IOR(UDPCP_IOC_MAGIC, 0x03, unsigned long)
+
+struct udpcp_statistics {
+ unsigned int txMsgs; /* Num of transmitted messages */
+ unsigned int rxMsgs; /* Num of received messages */
+ unsigned int txNodes; /* Num of receiver nodes */
+ unsigned int rxNodes; /* Num of transmitter nodes */
+ unsigned int txTimeout; /* Num of unsuccessful transmissions */
+ unsigned int rxTimeout; /* Num of partial message receptions */
+ unsigned int txRetries; /* Num of resends */
+ unsigned int rxDiscardedFrags; /* Num of discarded fragments */
+ unsigned int crcErrors; /* Num of crc errors detected */
+};
+
+#endif
+
diff --git a/net/Kconfig b/net/Kconfig
index 55fd82e..4a206fc 100644
--- a/net/Kconfig
+++ b/net/Kconfig
@@ -294,6 +294,7 @@ source "net/rfkill/Kconfig"
source "net/9p/Kconfig"
source "net/caif/Kconfig"
source "net/ceph/Kconfig"
+source "net/udpcp/Kconfig"
endif # if NET
diff --git a/net/Makefile b/net/Makefile
index 6b7bfd7..a17ae27 100644
--- a/net/Makefile
+++ b/net/Makefile
@@ -69,3 +69,4 @@ endif
obj-$(CONFIG_WIMAX) += wimax/
obj-$(CONFIG_DNS_RESOLVER) += dns_resolver/
obj-$(CONFIG_CEPH_LIB) += ceph/
+obj-$(CONFIG_UDPCP) += udpcp/
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 439d2a3..55b2d0c 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -1085,6 +1085,7 @@ error:
IP_INC_STATS(sock_net(sk), IPSTATS_MIB_OUTDISCARDS);
return err;
}
+EXPORT_SYMBOL(ip_append_data);
ssize_t ip_append_page(struct sock *sk, struct page *page,
int offset, size_t size, int flags)
@@ -1341,6 +1342,7 @@ error:
IP_INC_STATS(net, IPSTATS_MIB_OUTDISCARDS);
goto out;
}
+EXPORT_SYMBOL(ip_push_pending_frames);
/*
* Throw away all pending data on the socket.
diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c
index 3948c86..310369c 100644
--- a/net/ipv4/ip_sockglue.c
+++ b/net/ipv4/ip_sockglue.c
@@ -226,6 +226,7 @@ int ip_cmsg_send(struct net *net, struct msghdr *msg, struct ipcm_cookie *ipc)
}
return 0;
}
+EXPORT_SYMBOL(ip_cmsg_send);
/* Special input handler for packets caught by router alert option.
@@ -369,6 +370,7 @@ void ip_local_error(struct sock *sk, int err, __be32 daddr, __be16 port, u32 inf
if (sock_queue_err_skb(sk, skb))
kfree_skb(skb);
}
+EXPORT_SYMBOL(ip_local_error);
/*
* Handle MSG_ERRQUEUE
diff --git a/net/udpcp/Kconfig b/net/udpcp/Kconfig
new file mode 100644
index 0000000..a58c1b0
--- /dev/null
+++ b/net/udpcp/Kconfig
@@ -0,0 +1,34 @@
+#
+# UDPCP protocol
+#
+
+config UDPCP
+ tristate "UDPCP Communication Protocol"
+ depends on INET
+ ---help---
+ UDPCP is a communication protocol specified by the Open Base Station
+ Architecture Initiative Special Interest Group (OBSAI SIG). The
+ protocol is based on UDP and is designed to meet the needs of "Mobile
+ Communcation Base Station" internal communications.
+
+ The UDPCP communication service supports the following features:
+
+ -Connectionless communication for serial mode data transfer
+ -Acknowledged and unacknowledged transfer modes
+ -Retransmissions Algorithm
+ -Checksum Algorithm using Adler32
+ -Fragmentation of long messages (disassembly/reassembly) to
+ match to the MTU during transport:
+ -Broadcasting and multicasting messages to multiple peers in
+ unacknowledged transfer mode
+
+ UDPCP supports application level messages up to 64 KBytes (limited
+ by 16-bit packet data length field). Messages that are longer than the
+ MTU will be fragmented to the MTU.
+
+ UDPCP provides a reliable transport service that will perform message
+ retransmissions in case transport failures occur.
+
+ To compile this driver as a module, choose M here: the module
+ will be called udpcp.
+
diff --git a/net/udpcp/Makefile b/net/udpcp/Makefile
new file mode 100644
index 0000000..37f87c5
--- /dev/null
+++ b/net/udpcp/Makefile
@@ -0,0 +1,5 @@
+#
+# Makefile for UDPCP support code.
+#
+
+obj-$(CONFIG_UDPCP) += udpcp.o
diff --git a/net/udpcp/udpcp.c b/net/udpcp/udpcp.c
new file mode 100644
index 0000000..c7990c2
--- /dev/null
+++ b/net/udpcp/udpcp.c
@@ -0,0 +1,2854 @@
+/*
+ * UDPCP communication protocol
+ *
+ * Copyright (C) 2010 Stefani Seibold <stefani@seibold.net>
+ * in order of NSN Ulm/Germany
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ */
+
+#include <net/xfrm.h>
+#include <net/protocol.h>
+#include <net/ip.h>
+#include <net/udp.h>
+#include <net/udplite.h>
+#include <net/inet_common.h>
+#include <linux/zutil.h>
+#include <linux/module.h>
+#include <linux/proc_fs.h>
+#include <linux/spinlock.h>
+#include <linux/errqueue.h>
+#include <linux/atomic.h>
+
+#include <net/udpcp.h>
+
+#define VERSION "0.70"
+
+/*
+ * UDPCP Protocol default parameters
+ */
+#define UDPCP_TX_TIMEOUT 100 /* milliseconds */
+#define UDPCP_RX_TIMEOUT 1000 /* milliseconds */
+#define UDPCP_TX_MAXTRY 5
+#define UDPCP_OUTSTANDING_ACKS 1
+
+/*
+ * UDPCP Protocol definitions
+ */
+#define UDPCP_MSG_TYPE_BIT 14
+#define UDPCP_PROTOCOL_VERSION_BIT 11
+#define UDPCP_NO_ACK_BIT 10
+#define UDPCP_CHECKSUM_BIT 9
+#define UDPCP_SINGLE_ACK_BIT 8
+#define UDPCP_DUPLICATE_BIT 7
+
+#define UDPCP_MSG_TYPE_MASK (3 << UDPCP_MSG_TYPE_BIT)
+#define UDPCP_PROTOCOL_MASK (7 << UDPCP_PROTOCOL_VERSION_BIT)
+
+#define UDPCP_MSG_TYPE_DATA (1 << UDPCP_MSG_TYPE_BIT)
+#define UDPCP_MSG_TYPE_ACK (2 << UDPCP_MSG_TYPE_BIT)
+#define UDPCP_PROTOCOL_VERSION_2 (2 << UDPCP_PROTOCOL_VERSION_BIT)
+
+#define UDPCP_NO_ACK_FLAG (1 << UDPCP_NO_ACK_BIT)
+#define UDPCP_CHECKSUM_FLAG (1 << UDPCP_CHECKSUM_BIT)
+#define UDPCP_SINGLE_ACK_FLAG (1 << UDPCP_SINGLE_ACK_BIT)
+#define UDPCP_DUPLICATE_FLAG (1 << UDPCP_DUPLICATE_BIT)
+
+/*
+ * helper macros
+ */
+#define list_to_udpcpdest(d) container_of(d, struct udpcp_dest, list)
+#define list_to_udpcpsock(d) container_of(d, struct udpcp_sock, udpcplist)
+
+#define UDPCP_HDRSIZE (sizeof(struct udpcphdr)-sizeof(struct udphdr))
+
+#define RX_NODE 1
+#define TX_NODE 2
+
+/*
+ * name of the /proc entry
+ */
+#define UDPCP_PROC "driver/udpcp"
+
+/*
+ * UDPCP message header
+ */
+struct udpcphdr {
+ struct udphdr udphdr;
+ __be32 chksum;
+ __be16 msginfo;
+ u8 fragamount;
+ u8 fragnum;
+ __be16 msgid;
+ __be16 length;
+};
+
+/*
+ * UDPCP destination descriptor
+ *
+ * For each communication address an individual destination descriptor will
+ * be create.
+ *
+ * The fields has the following meanings:
+ *
+ * list: link list: part of udpcp_sock.destlist
+ * xmit: messages fragments to be transmit
+ * tx_time: timestamp of the last transmitted message fragment
+ * rx_time: timestamp ot the last received message fragment
+ * txTimeout: statistic use only: number of transmit timeout
+ * rxTimeout: statistic use only: number of receive timeout
+ * txRetries: statistic use only: number of transmit retries
+ * rxDiscardedFrags: statistic use only: number of discarded messages
+ * xmit_wait: message fragment which is waiting for an ACK
+ * xmit_last: last fragment transmitted
+ * recv_msg: first fragment of the received message
+ * recv_last: last fragment of the received message
+ * lastmsg: last messages fragment header received
+ * ipc: linux internal ipc cookie
+ * fl: flow/routing information
+ * rt: routing entry currently used for this destination
+ * addr: ipv4 destination address
+ * port: destination port number
+ * msgid: current message id for outgoing data messages
+ * use_flag: statistic use only: flag for dest using TX and/or RX
+ * insync: flag for protocol synchronization
+ * ackmode; ack mode for the current assembled message
+ * chkmode; checksum mode for the current assembled message
+ * try: current number of retries xmit_wait message
+ * acks: number of outstandig ack's
+ */
+struct udpcp_dest {
+ struct list_head list;
+ struct sk_buff_head xmit;
+ unsigned long tx_time;
+ unsigned long rx_time;
+ u32 txTimeout;
+ u32 rxTimeout;
+ u32 txRetries;
+ u32 rxDiscardedFrags;
+ struct sk_buff *xmit_wait;
+ struct sk_buff *xmit_last;
+ struct sk_buff *recv_msg;
+ struct sk_buff *recv_last;
+ struct udpcphdr lastmsg;
+ struct ipcm_cookie ipc;
+ struct flowi fl;
+ struct rtable *rt;
+ __be32 addr;
+ __be16 port;
+ u16 msgid;
+ u8 use_flag;
+ u8 insync;
+ u8 ackmode;
+ u8 chkmode;
+ u8 try;
+ u8 acks;
+};
+
+/*
+ * UDPCP socket descriptor
+ *
+ * For each opened socket individual socket descriptor will
+ * be created
+ *
+ * The fields has the following meanings:
+ *
+ * udpsock: UDP socket has to be the first member of udpcp_sock
+ * assembly: messages fragments currently assembled
+ * assembly_len: current length of the assembled message
+ * assembly_dest: current destination assembled
+ * wq: wait queue for UDPCP_IOCTL_SYNC
+ * destlist: head of destination descriptors link list
+ * udpcplist: link list: part of udpcp_list
+ * timer: timeout handler
+ * stat: statistics for this socket
+ * pending: number of pending messages fragment in the queues
+ * tx_timeout: transmit timeout in jiffies
+ * rx_timeout: receive timeout in jiffies
+ * udp_data_ready: original data_ready handler for this socket
+ * ackmode: default ack mode
+ * chkmode: default checksum mode
+ * maxtry: max. number of resends
+ * acks: max. number of outstandig ack's
+ * timeout: flag for unhandled timeout
+ */
+struct udpcp_sock {
+ struct udp_sock udpsock;
+ struct sk_buff_head assembly;
+ u32 assembly_len;
+ struct udpcp_dest *assembly_dest;
+ wait_queue_head_t wq;
+ struct list_head destlist;
+ struct list_head udpcplist;
+ struct timer_list timer;
+ struct udpcp_statistics stat;
+ u32 pending;
+ unsigned long tx_timeout;
+ unsigned long rx_timeout;
+ void (*udp_data_ready) (struct sock *sk, int bytes);
+ u8 ackmode;
+ u8 chkmode;
+ u8 maxtry;
+ u8 acks;
+ u8 timeout;
+};
+
+/* head of struct udpcp_sock.udpcplist link list */
+static struct list_head udpcp_list;
+
+/* spinlock for race free access to the static variables */
+static spinlock_t udpcp_lock;
+
+/* debug flag, set != 0 to enable debug */
+static int debug;
+
+/* overall UDPCP statistics */
+static atomic_t udpcp_txMsgs;
+static atomic_t udpcp_rxMsgs;
+static atomic_t udpcp_txNodes;
+static atomic_t udpcp_rxNodes;
+static atomic_t udpcp_txTimeout;
+static atomic_t udpcp_rxTimeout;
+static atomic_t udpcp_txRetries;
+static atomic_t udpcp_rxDiscardedFrags;
+static atomic_t udpcp_crcErrors;
+
+module_param(debug, int, 0);
+MODULE_PARM_DESC(debug, "Debug enabled or not");
+
+#ifdef CONFIG_PROC_FS
+/*
+ * Handle /proc/driver/udpcp
+ *
+ * Show the statistics information
+ */
+static int udpcp_proc(char *page, char **start, off_t off, int count, int *eof,
+ void *data)
+{
+ int len;
+
+ len = snprintf(page, count,
+ "txMsgs: %u\n"
+ "rxMsgs: %u\n"
+ "txNodes: %u\n"
+ "rxNodes: %u\n"
+ "txTimeout: %u\n"
+ "rxTimeout: %u\n"
+ "txRetries: %u\n"
+ "rxDiscaredFrags: %u\n"
+ "crcErrors: %u\n",
+ atomic_read(&udpcp_txMsgs),
+ atomic_read(&udpcp_rxMsgs),
+ atomic_read(&udpcp_txNodes),
+ atomic_read(&udpcp_rxNodes),
+ atomic_read(&udpcp_txTimeout),
+ atomic_read(&udpcp_rxTimeout),
+ atomic_read(&udpcp_txRetries),
+ atomic_read(&udpcp_rxDiscardedFrags),
+ atomic_read(&udpcp_crcErrors)
+ );
+
+ if (len <= off)
+ return 0;
+
+ len -= off;
+
+ if (len > count)
+ return count;
+
+ return len;
+}
+#endif
+
+/*
+ * Helper for the UDPCP header from a socket buffer
+ */
+static inline struct udpcphdr *udpcp_hdr(const struct sk_buff *skb)
+{
+ return (struct udpcphdr *)skb_transport_header(skb);
+}
+
+/*
+ * Helper for conversion a basic socket into a UDPCP socket
+ */
+static inline struct udpcp_sock *udpcp_sk(const struct sock *sk)
+{
+ return (struct udpcp_sock *)sk;
+}
+
+/*
+ * Dump the transport data of a socket buffer
+ */
+static inline void dump_data(struct sk_buff *skb, unsigned int max)
+{
+ unsigned int i;
+ unsigned char *data;
+ int data_len;
+
+ data = skb_transport_header(skb) + sizeof(struct udpcphdr);
+ data_len = skb_tail_pointer(skb) - data;
+
+ pr_debug(" data: ");
+
+ if (!data_len) {
+ pr_cont("<none>\n");
+ return;
+ }
+
+ if (max > data_len)
+ max = data_len;
+
+ for (i = 0; i < max; i++)
+ pr_cont("%02x ", data[i]);
+
+ if (data_len > max)
+ pr_cont("...");
+ pr_cont("\n");
+}
+
+/*
+ * Dump and decode a msginfo value
+ */
+static inline void dump_msginfo(u16 msginfo)
+{
+ pr_debug(" msginfo:0x%04x (", msginfo);
+
+ pr_cont("PCKT:");
+ switch (msginfo & UDPCP_MSG_TYPE_MASK) {
+ case UDPCP_MSG_TYPE_DATA:
+ pr_cont("DATA");
+ break;
+ case UDPCP_MSG_TYPE_ACK:
+ pr_cont("ACK");
+ break;
+ default:
+ pr_cont("UNKNOWN");
+ break;
+ }
+ pr_cont(" VER:%d",
+ (msginfo & UDPCP_PROTOCOL_MASK) >> UDPCP_PROTOCOL_VERSION_BIT);
+
+ if (msginfo & UDPCP_NO_ACK_FLAG)
+ pr_cont(" NO_ACK");
+ if (msginfo & UDPCP_CHECKSUM_FLAG)
+ pr_cont(" CHECKSUM");
+ if (msginfo & UDPCP_SINGLE_ACK_FLAG)
+ pr_cont(" SINGLE_ACK");
+ if (msginfo & UDPCP_DUPLICATE_FLAG)
+ pr_cont(" DUPLICATE");
+ pr_cont(")\n");
+}
+
+/*
+ * Dump and decode a UDPCP message fragment
+ */
+static void dump_msg(const char *action, struct sk_buff *skb, __be32 saddr,
+ __be32 daddr)
+{
+ struct udpcphdr *uh = udpcp_hdr(skb);
+
+ pr_debug("udpcp: %s (%lu)\n", action, jiffies);
+
+ pr_debug(" src:0x%08x:%d dst:0x%08x:%d fraglen:%d\n",
+ saddr, uh->udphdr.source, daddr, uh->udphdr.dest, skb->len);
+
+ pr_debug(" fragamount:%u fragnum:%u msgid:%u%s"
+ " length:%u checksum:0x%08x\n",
+ uh->fragamount, uh->fragnum, ntohs(uh->msgid),
+ (!uh->msgid) ? "(Sync)" : "", ntohs(uh->length),
+ ntohl(uh->chksum)
+ );
+
+ dump_msginfo(ntohs(uh->msginfo));
+ dump_data(skb, 16);
+}
+
+/*
+ * Create a new destination descriptor for the given IPV4 address and port
+ */
+static struct udpcp_dest *new_dest(struct sock *sk, __be32 addr, __be16 port)
+{
+ struct udpcp_dest *dest;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ dest = kzalloc(sizeof(*dest), sk->sk_allocation);
+
+ if (dest) {
+ skb_queue_head_init(&dest->xmit);
+ dest->addr = addr;
+ dest->port = port;
+ dest->ackmode = UDPCP_ACK;
+ list_add_tail(&dest->list, &usk->destlist);
+ }
+
+ return dest;
+}
+
+/*
+ * Lookup for a destination descriptor for the given IPV4 address and port
+ */
+static struct udpcp_dest *__find_dest(struct sock *sk, __be32 addr, __be16 port)
+{
+ struct udpcp_dest *dest;
+ struct list_head *p;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ list_for_each(p, &usk->destlist) {
+ dest = list_to_udpcpdest(p);
+
+ if ((dest->addr == addr) && (dest->port == port))
+ return dest;
+ }
+ return NULL;
+}
+
+/*
+ * Lookup for a destination descriptor and create a new one if no
+ * descriptor was found.
+ */
+static struct udpcp_dest *find_dest(struct sock *sk, __be32 addr, __be16 port)
+{
+ struct udpcp_dest *dest;
+
+ dest = __find_dest(sk, addr, port);
+
+ if (!dest)
+ dest = new_dest(sk, addr, port);
+
+ return dest;
+}
+
+/*
+ * Calculate udp checksum, mostly stolen from udp stack
+ */
+static void udpcp_do_csum(struct sock *sk, struct sk_buff *skb,
+ struct udpcp_dest *dest)
+{
+ struct flowi *fl = &dest->fl;
+ struct udphdr *uh = udp_hdr(skb);
+ __wsum csum = 0;
+ unsigned short len = ntohs(uh->len);
+
+ if (IS_UDPLITE(sk)) {
+ int cscov = udplite_sender_cscov(udp_sk(sk), uh);
+ int off = skb_transport_offset(skb);
+ int n = skb->len - off;
+
+ skb->ip_summed = CHECKSUM_NONE;
+ csum = skb_checksum(skb, off, (cscov > n) ? n : cscov, csum);
+ } else {
+ if (sk->sk_no_check == UDP_CSUM_NOXMIT) {
+ skb->ip_summed = CHECKSUM_NONE;
+ return;
+ }
+ if (skb->ip_summed == CHECKSUM_PARTIAL) {
+ /* UDP hardware csum */
+ skb->csum_start = skb_transport_header(skb) - skb->head;
+ skb->csum_offset = offsetof(struct udphdr, check);
+ uh->check =
+ ~csum_tcpudp_magic(fl->fl4_src, fl->fl4_dst, len,
+ sk->sk_protocol, 0);
+ return;
+ }
+ csum = csum_partial(uh, sizeof(struct udpcphdr), 0);
+ csum = csum_add(csum, skb->csum);
+ }
+
+ /* add protocol-dependent pseudo-header */
+ uh->check =
+ csum_tcpudp_magic(fl->fl4_src, fl->fl4_dst, len, sk->sk_protocol,
+ csum);
+ if (uh->check == 0)
+ uh->check = CSUM_MANGLED_0;
+}
+
+/*
+ * Fetch data from kernel space and fill in checksum if needed.
+ */
+static int ip_reply_glue_bits(void *dptr, char *to, int offset,
+ int len, int odd, struct sk_buff *skb)
+{
+ __wsum csum;
+
+ csum = csum_partial_copy_nocheck(dptr+offset, to, len, 0);
+ skb->csum = csum_block_add(skb->csum, csum, odd);
+ return 0;
+}
+
+/*
+ * Send an ack for a received data message fragment
+ *
+ * If the argument duplicate is true a ACK with UDPCP_DUPLICATE_FLAG set will
+ * be send
+ */
+static void udpcp_send_ack(struct sock *sk, struct sk_buff *skb,
+ struct udpcp_dest *dest, int duplicate)
+{
+ struct inet_sock *inet = inet_sk(sk);
+ struct udpcphdr *uh = udpcp_hdr(skb);
+ struct rtable *rt = NULL;
+ __wsum csum;
+ struct ipcm_cookie ipc;
+ struct udpcphdr rep;
+
+ memset(&rep, 0, sizeof(rep));
+
+ /* Swap the send and the receive ports. */
+ rep.udphdr.source = uh->udphdr.dest;
+ rep.udphdr.dest = uh->udphdr.source;
+ rep.udphdr.len = htons(sizeof(struct udpcphdr));
+
+ rep.msginfo = htons(UDPCP_MSG_TYPE_ACK |
+ UDPCP_NO_ACK_FLAG |
+ UDPCP_SINGLE_ACK_FLAG | UDPCP_PROTOCOL_VERSION_2);
+ if (duplicate)
+ rep.msginfo |= htons(UDPCP_DUPLICATE_FLAG);
+ else
+ memcpy(&dest->lastmsg, uh, sizeof(dest->lastmsg));
+ rep.msgid = uh->msgid;
+ rep.fragamount = uh->fragamount;
+ rep.fragnum = uh->fragnum;
+ rep.length = 0;
+ rep.chksum = 0;
+ if (ntohs(uh->msginfo) & UDPCP_CHECKSUM_FLAG) {
+ u8 *data;
+ u32 data_len;
+
+ data = (u8 *) &rep + sizeof(struct udphdr);
+ data_len = sizeof(struct udpcphdr)-sizeof(struct udphdr);
+
+ rep.msginfo |= htons(UDPCP_CHECKSUM_FLAG);
+ rep.chksum = htonl(zlib_adler32(1, data, data_len));
+ }
+
+ if (unlikely(debug)) {
+ struct sk_buff tmp;
+
+ tmp.len = ntohs(rep.udphdr.len);
+ tmp.head = tmp.transport_header = tmp.data = (void *)&rep;
+ tmp.tail = tmp.head + tmp.len;
+
+ dump_msg("ack msg", &tmp, ip_hdr(skb)->daddr,
+ ip_hdr(skb)->saddr);
+ }
+
+ csum = csum_tcpudp_nofold(ip_hdr(skb)->daddr,
+ ip_hdr(skb)->saddr,
+ sizeof(rep), sk->sk_protocol, 0);
+
+ ipc.addr = dest->addr;
+ ipc.opt = NULL;
+ ipc.tx_flags = 0;
+
+ {
+ struct flowi fl = {
+ .nl_u = { .ip4_u = {
+ .daddr = ipc.addr,
+ .saddr = ip_hdr(skb)->daddr,
+ .tos = RT_TOS(ip_hdr(skb)->tos)
+ }
+ },
+ .uli_u = { .ports = {
+ .sport = udp_hdr(skb)->dest,
+ .dport = udp_hdr(skb)->source
+ }
+ },
+ .proto = sk->sk_protocol,
+ };
+ security_skb_classify_flow(skb, &fl);
+ if (ip_route_output_key(sock_net(sk), &rt, &fl))
+ return;
+ }
+
+ inet->tos = ip_hdr(skb)->tos;
+ sk->sk_priority = skb->priority;
+ sk->sk_protocol = ip_hdr(skb)->protocol;
+ sk->sk_bound_dev_if = 0;
+ ip_append_data(sk, ip_reply_glue_bits, &rep, sizeof(rep),
+ 0, &ipc, &rt, MSG_DONTWAIT);
+ skb = skb_peek(&sk->sk_write_queue);
+ if (skb) {
+ *((__sum16 *)skb_transport_header(skb) +
+ offsetof(struct udphdr, check) / 2) =
+ csum_fold(csum_add(skb->csum, csum));
+ skb->ip_summed = CHECKSUM_NONE;
+ ip_push_pending_frames(sk);
+ }
+
+ ip_rt_put(rt);
+
+ UDP_INC_STATS_BH(sock_net(sk), UDP_MIB_OUTDATAGRAMS, IS_UDPLITE(sk));
+}
+
+/*
+ * Pass a UDPCP skb buffer to the ip stack and send it
+ */
+static int udpcp_send_skb(struct sock *sk, struct sk_buff *skb,
+ struct udpcp_dest *dest, struct ip_options *opt)
+{
+ int err;
+
+ skb_dst_set(skb, dst_clone(&dest->rt->dst));
+
+ err = ip_build_and_send_pkt(skb, sk, dest->fl.fl4_src,
+ dest->fl.fl4_dst, opt);
+
+ if (!err)
+ UDP_INC_STATS_USER(sock_net(sk), UDP_MIB_OUTDATAGRAMS,
+ IS_UDPLITE(sk));
+ return err;
+}
+
+/*
+ * Release a routing table entry if no packed will be assembled
+ */
+static void udpcp_dst_release(struct udpcp_sock *usk, struct udpcp_dest *dest)
+{
+ if (usk->assembly_dest != dest) {
+ dst_release(&dest->rt->dst);
+ dest->rt = NULL;
+ }
+}
+
+/*
+ * Return true it the passed skb socket buffer is the last in the list
+ */
+static inline bool skb_is_eoq(const struct sk_buff_head *list,
+ const struct sk_buff *skb)
+{
+ return (skb->next == (struct sk_buff *)list);
+}
+
+/*
+ * Arm the timeout handler for the socket
+ */
+static void udpcp_timer(struct sock *sk, unsigned long timeout)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ mod_timer(&usk->timer, timeout);
+}
+
+/*
+ * Decrement the socket pending counter and wakeup a waiting UDPCP_IOCTL_SYNC
+ */
+static inline void udpcp_dec_pending(struct sock *sk)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ if (!--usk->pending) {
+ if (waitqueue_active(&usk->wq))
+ wake_up_interruptible(&usk->wq);
+ }
+}
+
+/*
+ * Returns true is the passed message fragment is the last fragment
+ */
+static inline int udpcp_is_last_frag(struct udpcphdr *uh)
+{
+ return uh->fragamount == uh->fragnum + 1;
+}
+
+/*
+ * Transmit data message fragments
+ */
+static int _udpcp_xmit(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct sk_buff *skb = NULL;
+ struct sk_buff *skbc;
+ struct udpcphdr *uh;
+ int err = 0;
+
+ if (dest->acks >= usk->acks)
+ goto out;
+
+ if (!dest->xmit_last) {
+ /*
+ * handle data message fragments without an ack
+ */
+ while ((skb = skb_peek(&dest->xmit))) {
+ uh = udpcp_hdr(skb);
+
+ if (!(ntohs(uh->msginfo) & UDPCP_NO_ACK_FLAG))
+ break;
+ if (udpcp_is_last_frag(uh)) {
+ usk->stat.txMsgs++;
+ atomic_inc(&udpcp_txMsgs);
+ }
+ skb_unlink(skb, &dest->xmit);
+ udpcp_dec_pending(sk);
+ if (unlikely(debug))
+ dump_msg("send msg", skb, dest->fl.fl4_src,
+ dest->fl.fl4_dst);
+ err = udpcp_send_skb(sk, skb, dest,
+ (struct ip_options *)skb->cb);
+ if (err) {
+ kfree_skb(skb);
+ skb = NULL;
+ break;
+ }
+ }
+ dest->xmit_wait = skb;
+ } else {
+ /*
+ * handle next data message fragment waiting for an ack
+ */
+ uh = udpcp_hdr(dest->xmit_last);
+
+ if (udpcp_is_last_frag(uh))
+ goto out;
+
+ /*
+ * get next data message fragment
+ */
+ skb = dest->xmit_last->next;
+ }
+
+ /*
+ * send all data message fragment till the first which must be acked
+ */
+ while (skb) {
+ skbc = skb_clone(skb, sk->sk_allocation);
+
+ if (!skbc)
+ break;
+
+ if (unlikely(debug))
+ dump_msg("send msg", skbc, dest->fl.fl4_src,
+ dest->fl.fl4_dst);
+ err = udpcp_send_skb(sk, skbc, dest,
+ (struct ip_options *)skb->cb);
+ if (err) {
+ kfree_skb(skbc);
+ break;
+ }
+
+ uh = udpcp_hdr(skb);
+
+ if (!(ntohs(uh->msginfo) & UDPCP_SINGLE_ACK_FLAG)
+ || udpcp_is_last_frag(uh)) {
+ dest->xmit_last = skb;
+
+ if (++dest->acks >= usk->acks || udpcp_is_last_frag(uh))
+ break;
+ }
+
+ skb = skb_is_eoq(&dest->xmit, skb) ? NULL : skb->next;
+ }
+
+out:
+ if (skb_queue_empty(&dest->xmit))
+ udpcp_dst_release(usk, dest);
+
+ return err;
+}
+
+/*
+ * Transmit data message fragments and rearm the timeout handler if necessary
+ */
+static int udpcp_xmit(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int ret;
+
+ ret = _udpcp_xmit(sk, dest);
+
+ if (dest->xmit_wait) {
+ dest->tx_time = jiffies;
+
+ if (!timer_pending(&usk->timer))
+ udpcp_timer(sk, dest->tx_time + usk->tx_timeout);
+ }
+ return ret;
+}
+
+/*
+ * Queue the assembled message fragment into the transmit queue
+ */
+static void udpcp_queue_xmit(struct sock *sk, struct udpcp_dest *dest,
+ u8 ackmode, u8 chkmode)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct udpcphdr *uh;
+ struct sk_buff *skb;
+ u8 fragamount;
+ u8 fragnum;
+ unsigned short msginfo;
+ struct flowi *fl = &dest->fl;
+
+ msginfo = UDPCP_MSG_TYPE_DATA | UDPCP_PROTOCOL_VERSION_2;
+ switch (ackmode) {
+ case UDPCP_NOACK:
+ msginfo |= UDPCP_NO_ACK_FLAG;
+ break;
+ case UDPCP_SINGLE_ACK:
+ msginfo |= UDPCP_SINGLE_ACK_FLAG;
+ break;
+ case UDPCP_ACK:
+ default:
+ break;
+ }
+ switch (chkmode) {
+ case UDPCP_NOCHECKSUM:
+ break;
+ case UDPCP_CHECKSUM:
+ default:
+ msginfo |= UDPCP_CHECKSUM_FLAG;
+ break;
+ }
+
+ fragamount = skb_queue_len(&usk->assembly);
+
+ udpcp_sk(sk)->pending += fragamount;
+
+ for (fragnum = 0; fragnum != fragamount; fragnum++) {
+ unsigned char *data;
+ int data_len;
+
+ skb = skb_dequeue(&usk->assembly);
+ uh = udpcp_hdr(skb);
+
+ /*
+ * setup a UDPCP header
+ */
+ uh->chksum = 0;
+ uh->msginfo = htons(msginfo);
+ uh->fragnum = fragnum;
+ uh->fragamount = fragamount;
+ uh->msgid = htons(dest->msgid);
+ uh->length = htons(usk->assembly_len);
+
+ data = skb_transport_header(skb) + sizeof(struct udphdr);
+ data_len = skb_tail_pointer(skb) - data;
+
+ if (chkmode == UDPCP_CHECKSUM)
+ uh->chksum = htonl(zlib_adler32(1, data, data_len));
+ /*
+ * create a UDP header
+ */
+ uh->udphdr.source = fl->fl_ip_sport;
+ uh->udphdr.dest = fl->fl_ip_dport;
+ uh->udphdr.len = htons(sizeof(struct udphdr) + data_len);
+ uh->udphdr.check = 0;
+
+ /*
+ * create UDP checksum
+ */
+ udpcp_do_csum(sk, skb, dest);
+
+ /*
+ * add to xmit queue
+ */
+ skb_queue_tail(&dest->xmit, skb);
+ }
+
+ dest->msgid++;
+ usk->assembly_len = 0;
+ usk->assembly_dest = NULL;
+}
+
+/*
+ * Remove all data message fragments of the first message from the transmit
+ * queue all fragments will be merged together
+ */
+static struct sk_buff *udpcp_dequeue_msg(struct sock *sk,
+ struct udpcp_dest *dest)
+{
+ struct sk_buff *msg;
+ struct sk_buff *skb;
+ struct sk_buff **next;
+ struct udpcphdr *uh;
+
+ msg = skb_dequeue(&dest->xmit);
+ if (!msg)
+ return NULL;
+ skb_orphan(msg);
+
+ uh = udpcp_hdr(msg);
+ if (!uh->msgid) {
+ /*
+ * sync message
+ */
+ kfree_skb(msg);
+ return NULL;
+ }
+
+ skb_pull(msg, sizeof(struct udpcphdr));
+ if (udpcp_is_last_frag(uh))
+ return msg;
+
+ next = &skb_shinfo(msg)->frag_list;
+ for (;;) {
+ skb = skb_dequeue(&dest->xmit);
+ if (!skb)
+ break;
+ skb_orphan(skb);
+ uh = udpcp_hdr(skb);
+ skb_pull(msg, sizeof(struct udpcphdr));
+ msg->len += skb->len;
+ msg->data_len += skb->len;
+ *next = skb;
+ if (udpcp_is_last_frag(uh))
+ break;
+ next = &skb->next;
+ }
+ return msg;
+}
+
+static void udpcp_flush_err(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct inet_sock *inet = inet_sk(sk);
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ if (!inet->recverr)
+ skb_queue_purge(&dest->xmit);
+ else {
+ struct sock_exterr_skb *serr;
+ struct iphdr *iph;
+ struct sk_buff *skb;
+
+ while (!skb_queue_empty(&dest->xmit)) {
+ skb = udpcp_dequeue_msg(sk, dest);
+ if (!skb)
+ continue;
+
+ if (unlikely(debug))
+ dump_msg("flush outgoing message", skb,
+ dest->fl.fl4_src, dest->fl.fl4_dst);
+
+ skb_push(skb, sizeof(struct iphdr));
+ skb_reset_network_header(skb);
+ iph = ip_hdr(skb);
+ iph->daddr = dest->rt->rt_dst;
+
+ serr = SKB_EXT_ERR(skb);
+ serr->ee.ee_errno = EPROTO;
+ serr->ee.ee_origin = SO_EE_ORIGIN_LOCAL;
+ serr->ee.ee_type = 0;
+ serr->ee.ee_code = 0;
+ serr->ee.ee_pad = 0;
+ serr->ee.ee_info = 0;
+ serr->ee.ee_data = 0;
+ serr->addr_offset = (u8 *) &iph->daddr -
+ skb_network_header(skb);
+ serr->port = dest->fl.fl_ip_dport;
+
+ skb_reset_transport_header(skb);
+ skb_pull(skb, sizeof(struct iphdr));
+
+ /*
+ * set a flag for UDPCP message
+ */
+ UDP_SKB_CB(skb)->udpcp_flag = 1;
+
+ /*
+ * pass the dequeued message to the error queue of the
+ * socket
+ */
+ skb_set_owner_r(skb, sk);
+ skb_queue_tail(&sk->sk_error_queue, skb);
+ if (!sock_flag(sk, SOCK_DEAD)) {
+ if (usk->udp_data_ready)
+ usk->udp_data_ready(sk, skb->len);
+ }
+ }
+ }
+
+ dest->xmit_wait = 0;
+ dest->xmit_last = 0;
+ dest->try = 0;
+ dest->acks = 0;
+
+ usk->pending = 0;
+ if (waitqueue_active(&usk->wq))
+ wake_up_interruptible(&usk->wq);
+}
+
+/*
+ * Purge the current incoming data message
+ */
+static void udpcp_purge_incoming(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ if (dest->recv_last) {
+ u32 fragnum = udpcp_hdr(dest->recv_last)->fragnum + 1;
+
+ dest->rxDiscardedFrags += fragnum;
+ usk->stat.rxDiscardedFrags += fragnum;
+ atomic_add(fragnum, &udpcp_rxDiscardedFrags);
+
+ dest->lastmsg.msgid = 0;
+
+ if (unlikely(debug))
+ dump_msg("purge incoming message", dest->recv_msg,
+ dest->fl.fl4_src, dest->fl.fl4_dst);
+ }
+
+ kfree_skb(dest->recv_msg);
+ dest->recv_msg = 0;
+ dest->recv_last = 0;
+}
+
+/*
+ * Resend all data message fragments to the one which is currently waiting for
+ * an ack
+ */
+static int udpcp_resend(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct sk_buff *skb;
+ struct sk_buff *skbc;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int err;
+
+ if (++dest->try >= usk->maxtry) {
+ dest->insync = 0;
+ udpcp_flush_err(sk, dest);
+ udpcp_purge_incoming(sk, dest);
+ udpcp_dst_release(usk, dest);
+ return 0;
+ }
+
+ dest->txRetries++;
+ usk->stat.txRetries++;
+ atomic_inc(&udpcp_txRetries);
+
+ if (!dest->xmit_last)
+ _udpcp_xmit(sk, dest);
+ else {
+ skb = dest->xmit_wait;
+
+ for (;;) {
+ skbc = skb_clone(skb, sk->sk_allocation);
+
+ if (skbc == NULL)
+ break;
+
+ if (unlikely(debug))
+ dump_msg("resend msg", skbc, dest->fl.fl4_src,
+ dest->fl.fl4_dst);
+ err = udpcp_send_skb(sk, skbc, dest,
+ (struct ip_options *)skb->cb);
+ if (err) {
+ kfree_skb(skbc);
+ break;
+ }
+
+ if (skb == dest->xmit_last) {
+ _udpcp_xmit(sk, dest);
+ break;
+ }
+
+ skb = skb->next;
+ }
+ }
+ dest->tx_time = jiffies;
+
+ return 1;
+}
+
+/*
+ * Handle udpcp timeout
+ */
+static void udpcp_handle_timeout(struct sock *sk)
+{
+ struct udpcp_dest *dest;
+ struct list_head *p;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int wflag = 0;
+ unsigned long t = jiffies + UDPCP_MAX_WAIT_SEC * HZ + 1;
+
+ usk->timeout = 0;
+
+ /*
+ * walk through all destinations
+ */
+ list_for_each(p, &usk->destlist) {
+ dest = list_to_udpcpdest(p);
+
+ if (dest->xmit_wait) {
+ if (time_is_before_eq_jiffies
+ (dest->tx_time + usk->tx_timeout)) {
+ /*
+ * transmit timeout expired
+ */
+ if (unlikely(debug))
+ dump_msg("send timeout",
+ dest->xmit_wait,
+ dest->fl.fl4_src,
+ dest->fl.fl4_dst);
+ if (udpcp_resend(sk, dest) == 0) {
+ dest->txTimeout++;
+ usk->stat.txTimeout++;
+ atomic_inc(&udpcp_txTimeout);
+ goto check_incoming;
+ }
+ wflag = 1;
+ }
+ if (time_before(dest->tx_time + usk->tx_timeout, t)) {
+ /*
+ * calculate new timeout timer value
+ */
+ t = dest->tx_time + usk->tx_timeout;
+ wflag = 1;
+ }
+ }
+check_incoming:
+ if (dest->recv_msg) {
+ if (time_is_before_eq_jiffies
+ (dest->rx_time + usk->rx_timeout)) {
+ /*
+ * receive timeout occurred
+ */
+ if (unlikely(debug))
+ dump_msg("receive timeout",
+ dest->recv_last,
+ dest->fl.fl4_src,
+ dest->fl.fl4_dst);
+ udpcp_purge_incoming(sk, dest);
+ dest->rxTimeout++;
+ usk->stat.rxTimeout++;
+ atomic_inc(&udpcp_rxTimeout);
+ } else
+ if (time_before(dest->rx_time + usk->rx_timeout, t)) {
+ /*
+ * calculate new timeout timer value
+ */
+ t = dest->rx_time + usk->rx_timeout;
+ wflag = 1;
+ }
+ }
+ }
+ /*
+ * restart timer if necessary
+ */
+ if (wflag)
+ udpcp_timer(sk, t);
+}
+
+/*
+ * Timeout function
+ */
+static void udpcp_timeout(unsigned long data)
+{
+ struct sock *sk = (struct sock *)data;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ bh_lock_sock(sk);
+ if (!sock_owned_by_user(sk))
+ udpcp_handle_timeout(sk);
+ else {
+ /*
+ * bad, cannot handle the timeout because the socket is in use
+ * set flag for unhandled timeout and rearm the timer
+ */
+ usk->timeout = 1;
+ udpcp_timer(sk, jiffies + 1);
+ }
+ bh_unlock_sock(sk);
+}
+
+/*
+ * Handle timeout if an the unhandled timeout flag is set
+ */
+static inline void check_timeout(struct sock *sk)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ while (usk->timeout) {
+ lock_sock(sk);
+ if (usk->timeout)
+ udpcp_handle_timeout(sk);
+ release_sock(sk);
+ }
+}
+
+/*
+ * Release the socket lock and test for unhandled timeouts
+ */
+static inline void udpcp_release_sock(struct sock *sk)
+{
+ release_sock(sk);
+ check_timeout(sk);
+}
+
+/*
+ * Parse sendmsg() control message
+ */
+static int udpcp_cmsg_send(struct msghdr *msg, u8 * ackmode, u8 * chkmode)
+{
+ struct cmsghdr *cmsg;
+
+ for (cmsg = CMSG_FIRSTHDR(msg); cmsg; cmsg = CMSG_NXTHDR(msg, cmsg)) {
+ if (!CMSG_OK(msg, cmsg))
+ return -EINVAL;
+ if (cmsg->cmsg_level != SOL_UDPCP)
+ continue;
+ switch (cmsg->cmsg_type) {
+ case UDPCP_NOACK:
+ case UDPCP_ACK:
+ case UDPCP_SINGLE_ACK:
+ *ackmode = cmsg->cmsg_type;
+ break;
+ case UDPCP_CHECKSUM:
+ case UDPCP_NOCHECKSUM:
+ *chkmode = cmsg->cmsg_type;
+ break;
+ default:
+ return -EINVAL;
+ }
+ }
+ return 0;
+}
+
+/*
+ * Validate a skb buffer
+ */
+static int udpcp_validate_skb(struct sk_buff *skb)
+{
+ if (skb->next) {
+ pr_err("udpcp: unexpected skb_buff->next != NULL\n");
+ BUG();
+ return 1;
+ }
+ if (skb_shinfo(skb)->frag_list) {
+ pr_err("udpcp: unexpected skb_shinfo(skb)->frag_list != NULL\n");
+ BUG();
+ return 1;
+ }
+ return 0;
+}
+
+/*
+ * Split a message into fragments and store it into the assemble queue
+ * mostly stolen from UDP stack
+ */
+static int udpcp_data(struct sock *sk, struct udpcp_dest *dest,
+ int getfrag(void *from, char *to, int offset, int len,
+ int odd, struct sk_buff *skb),
+ struct iovec *from, int length, unsigned int flags)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct inet_sock *inet = inet_sk(sk);
+ struct sk_buff *skb;
+ struct ipcm_cookie *ipc = &dest->ipc;
+ struct ip_options *opt = ipc->opt;
+ int hh_len;
+ int exthdrlen;
+ int mtu;
+ int copy;
+ int err;
+ int offset = 0;
+ unsigned int maxfraglen, fragheaderlen;
+ int csummode = CHECKSUM_NONE;
+ int transhdrlen = sizeof(struct udpcphdr);
+ struct rtable *rt = dest->rt;
+
+ if (opt && sizeof(skb->cb) < optlength(opt)) {
+ err = -EFAULT;
+ goto error;
+ }
+
+ usk->assembly_len += length;
+ usk->assembly_dest = dest;
+
+ if (usk->assembly_len > UDPCP_MAX_MSGSIZE) {
+ ip_local_error(sk, EMSGSIZE, rt->rt_dst, dest->fl.fl_ip_dport,
+ usk->assembly_len);
+ err = -EMSGSIZE;
+ goto error;
+ }
+
+ mtu = (inet->pmtudisc == IP_PMTUDISC_PROBE) ?
+ rt->dst.dev->mtu : dst_mtu(rt->dst.path);
+ sk->sk_sndmsg_page = NULL;
+ sk->sk_sndmsg_off = 0;
+ exthdrlen = rt->dst.header_len;
+ length += exthdrlen;
+ transhdrlen += exthdrlen;
+
+ hh_len = LL_RESERVED_SPACE(rt->dst.dev);
+
+ fragheaderlen = sizeof(struct iphdr) + (opt ? opt->optlen : 0);
+ maxfraglen = ((mtu - fragheaderlen) & ~7) + fragheaderlen;
+
+ if (rt->dst.dev->features & NETIF_F_V4_CSUM && !exthdrlen)
+ csummode = CHECKSUM_PARTIAL;
+
+ skb = skb_peek_tail(&usk->assembly);
+ if (skb) {
+ unsigned int off;
+
+ off = skb->len;
+
+ copy = mtu - skb->len;
+ if (copy > length)
+ copy = length;
+
+ if (copy > 0 &&
+ getfrag(from, skb_put(skb, copy), 0, copy, off, skb) < 0) {
+ __skb_trim(skb, off);
+ err = -EFAULT;
+ goto error;
+ }
+ length -= copy;
+ offset += copy;
+
+ if (!length)
+ return 0;
+ }
+
+ do {
+ char *data;
+ unsigned int datalen;
+ unsigned int fraglen;
+ unsigned int alloclen;
+
+ length += transhdrlen;
+ /*
+ * If remaining data exceeds the mtu,
+ * we know we need more fragment(s).
+ */
+ datalen = length;
+ if (datalen > mtu - fragheaderlen)
+ datalen = maxfraglen - fragheaderlen;
+ fraglen = datalen + fragheaderlen;
+
+ if ((flags & MSG_MORE)
+ && !(rt->dst.dev->features & NETIF_F_SG))
+ alloclen = mtu;
+ else
+ alloclen = fraglen;
+
+ alloclen += rt->dst.trailer_len + hh_len + 15;
+
+ udpcp_release_sock(sk);
+ skb = sock_alloc_send_skb(sk, alloclen,
+ (flags & MSG_DONTWAIT), &err);
+ lock_sock(sk);
+ if (skb == NULL)
+ goto error;
+
+ if (udpcp_validate_skb(skb)) {
+ kfree_skb(skb);
+
+ goto error;
+ }
+
+ /*
+ * Fill in the control structures
+ */
+ skb->ip_summed = csummode;
+ skb->csum = 0;
+ skb_reserve(skb, hh_len);
+
+ /*
+ * Find where to start putting bytes.
+ */
+ data = skb_put(skb, fraglen);
+ skb_set_network_header(skb, exthdrlen);
+ skb->transport_header = (skb->network_header + fragheaderlen);
+ data += fragheaderlen;
+
+ copy = datalen - transhdrlen;
+
+ if (copy > 0 &&
+ getfrag(from, data + transhdrlen, offset, copy, 0, skb) < 0) {
+ err = -EFAULT;
+ kfree_skb(skb);
+ goto error;
+ }
+
+ offset += copy;
+ length -= datalen;
+
+ if (ipc->opt)
+ memcpy(skb->cb, &ipc->opt, optlength(opt));
+
+ skb_pull(skb, fragheaderlen);
+ skb_queue_tail(&usk->assembly, skb);
+ } while (length > 0);
+
+ return 0;
+error:
+ skb_queue_purge(&usk->assembly);
+ usk->assembly_len = 0;
+
+ IP_INC_STATS(sock_net(sk), IPSTATS_MIB_OUTDISCARDS);
+ return err;
+}
+
+/*
+ * This function will be called by send(), sento() and sendmsg()
+ */
+static int udpcp_sendmsg(struct kiocb *iocb, struct sock *sk,
+ struct msghdr *msg, size_t len)
+{
+ struct inet_sock *inet = inet_sk(sk);
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct ipcm_cookie *ipc;
+ struct rtable *rt = NULL;
+ int free = 0;
+ int connected = 0;
+ __be32 daddr, faddr, saddr;
+ __be16 dport;
+ u8 tos;
+ int err = 0;
+ int corkreq = usk->udpsock.corkflag || msg->msg_flags & MSG_MORE;
+ int (*getfrag) (void *, char *, int, int, int, struct sk_buff *);
+ struct udpcp_dest *dest;
+
+ if (len > UDPCP_MAX_MSGSIZE)
+ return -EMSGSIZE;
+
+ /*
+ * Check the flags.
+ */
+ if (msg->msg_flags & MSG_OOB)
+ return -EOPNOTSUPP;
+
+ /*
+ * check if socket is binded to a port
+ */
+ if (!(sk->sk_userlocks & SOCK_BINDPORT_LOCK) || !inet->inet_num)
+ return -ENOTCONN;
+
+ /*
+ * Get and verify the address.
+ */
+ if (msg->msg_name) {
+ struct sockaddr_in *usin = (struct sockaddr_in *)msg->msg_name;
+ if (msg->msg_namelen < sizeof(*usin))
+ return -EINVAL;
+ if (usin->sin_family != AF_INET) {
+ if (usin->sin_family != AF_UNSPEC)
+ return -EAFNOSUPPORT;
+ }
+
+ daddr = usin->sin_addr.s_addr;
+ dport = usin->sin_port;
+ } else {
+ if (sk->sk_state != TCP_ESTABLISHED)
+ return -EDESTADDRREQ;
+ daddr = inet->inet_daddr;
+ dport = inet->inet_dport;
+ /* Open fast path for connected socket.
+ Route will not be used, if at least one option is set.
+ */
+ connected = 1;
+ }
+
+ if (dport == 0)
+ return -EINVAL;
+
+ dest = find_dest(sk, daddr, dport);
+
+ if (!(dest->use_flag & TX_NODE)) {
+ dest->use_flag |= TX_NODE;
+ usk->stat.txNodes++;
+ atomic_inc(&udpcp_txNodes);
+ }
+
+ ipc = &dest->ipc;
+
+ getfrag = IS_UDPLITE(sk) ? udplite_getfrag : ip_generic_getfrag;
+
+ if (!skb_queue_empty(&usk->assembly)) {
+ /*
+ * assembly is ongoing
+ */
+ lock_sock(sk);
+ if (likely(!skb_queue_empty(&usk->assembly))) {
+ if (usk->assembly_dest != dest) {
+ udpcp_release_sock(sk);
+ return -EUSERS;
+ }
+ ipc->opt =
+ (struct ip_options *)skb_peek(&usk->assembly)->cb;
+ goto queue_data;
+ }
+ udpcp_release_sock(sk);
+ }
+
+ ipc->addr = inet->inet_saddr;
+ ipc->oif = sk->sk_bound_dev_if;
+
+ dest->ackmode = usk->ackmode;
+ dest->chkmode = usk->chkmode;
+
+ if (msg->msg_controllen) {
+ /*
+ * handle control message
+ */
+ err = udpcp_cmsg_send(msg, &dest->ackmode, &dest->chkmode);
+ if (err)
+ return err;
+ err = ip_cmsg_send(sock_net(sk), msg, ipc);
+ if (err)
+ return err;
+ if (ipc->opt)
+ free = 1;
+ connected = 0;
+ }
+
+ if (!ipc->opt)
+ ipc->opt = inet->opt;
+
+ saddr = ipc->addr;
+ ipc->addr = faddr = daddr;
+
+ if (ipc->opt && ipc->opt->srr) {
+ if (!daddr)
+ return -EINVAL;
+ faddr = ipc->opt->faddr;
+ connected = 0;
+ }
+ tos = RT_TOS(inet->tos);
+ if (sock_flag(sk, SOCK_LOCALROUTE) ||
+ (msg->msg_flags & MSG_DONTROUTE) ||
+ (ipc->opt && ipc->opt->is_strictroute)) {
+ tos |= RTO_ONLINK;
+ connected = 0;
+ }
+
+ if (ipv4_is_multicast(daddr)) {
+ if (dest->ackmode != UDPCP_NOACK) {
+ err = EOPNOTSUPP;
+ goto out;
+ }
+ if (!ipc->oif)
+ ipc->oif = inet->mc_index;
+ if (!saddr)
+ saddr = inet->mc_addr;
+ connected = 0;
+ }
+
+ lock_sock(sk);
+ rt = dest->rt;
+ if (rt)
+ goto queue_data;
+ udpcp_release_sock(sk);
+
+ /*
+ * calculate routing
+ */
+ if (connected)
+ rt = (struct rtable *)sk_dst_check(sk, 0);
+
+ if (rt == NULL) {
+ struct flowi fl = {.oif = ipc->oif,
+ .nl_u = {.ip4_u = {.daddr = faddr,
+ .saddr = saddr,
+ .tos = tos} },
+ .proto = sk->sk_protocol,
+ .uli_u = {.ports = {.sport = inet->inet_sport,
+ .dport = dport} }
+ };
+ struct net *net = sock_net(sk);
+
+ security_sk_classify_flow(sk, &fl);
+ err = ip_route_output_flow(net, &rt, &fl, sk, 1);
+ if (err) {
+ if (err == -ENETUNREACH)
+ IP_INC_STATS_BH(net, IPSTATS_MIB_OUTNOROUTES);
+ goto out;
+ }
+
+ err = -EACCES;
+ if ((rt->rt_flags & RTCF_BROADCAST) &&
+ !sock_flag(sk, SOCK_BROADCAST))
+ goto out;
+ if (connected)
+ sk_dst_set(sk, dst_clone(&rt->dst));
+ }
+
+ if (msg->msg_flags & MSG_CONFIRM)
+ goto do_confirm;
+back_from_confirm:
+
+ saddr = rt->rt_src;
+ if (!ipc->addr)
+ daddr = ipc->addr = rt->rt_dst;
+
+ lock_sock(sk);
+
+ dest->fl.fl4_dst = daddr;
+ dest->fl.fl_ip_dport = dport;
+ dest->fl.fl4_src = saddr;
+ dest->fl.fl_ip_sport = inet->inet_sport;
+ dest->rt = rt;
+
+queue_data:
+ if (msg->msg_flags & MSG_PROBE)
+ goto release;
+
+ if (!dest->insync && skb_queue_empty(&dest->xmit)) {
+ /*
+ * if not synced, queue a SYNC message
+ */
+ err = udpcp_data(sk, dest, getfrag, NULL, 0, 0);
+ if (err)
+ goto release;
+ dest->msgid = 0;
+ udpcp_queue_xmit(sk, dest, UDPCP_ACK, UDPCP_CHECKSUM);
+ }
+
+ /*
+ * split message and store it to the assembly queue
+ */
+ err = udpcp_data(sk, dest, getfrag, msg->msg_iov, len,
+ corkreq ? msg->msg_flags | MSG_MORE : msg->msg_flags);
+ if (err)
+ goto release;
+
+ if (!dest->msgid)
+ dest->msgid = 1;
+
+ if (!corkreq) {
+ /*
+ * message is complete, transfer it from the assembly queue
+ * into the transmit queue
+ */
+ udpcp_queue_xmit(sk, dest, dest->ackmode, dest->chkmode);
+ /*
+ * start transmit if possible
+ */
+ err = udpcp_xmit(sk, dest);
+ }
+release:
+ udpcp_release_sock(sk);
+out:
+ if (free)
+ kfree(ipc->opt);
+
+ if (!err)
+ return len;
+ /*
+ * ENOBUFS = no kernel mem, SOCK_NOSPACE = no sndbuf space. Reporting
+ * ENOBUFS might not be good (it's not tunable per se), but otherwise
+ * we don't have a good statistic (IpOutDiscards but it can be too many
+ * things). We could add another new stat but at least for now that
+ * seems like overkill.
+ */
+ if (err == -ENOBUFS || test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) {
+ UDP_INC_STATS_USER(sock_net(sk),
+ UDP_MIB_SNDBUFERRORS, IS_UDPLITE(sk));
+ }
+ return err;
+
+do_confirm:
+ dst_confirm(&rt->dst);
+ if (!(msg->msg_flags & MSG_PROBE) || len)
+ goto back_from_confirm;
+
+ err = 0;
+ goto out;
+}
+
+/*
+ * Sendpage() is not really implemented
+ */
+static int udpcp_sendpage(struct sock *sk, struct page *page, int offset,
+ size_t size, int flags)
+{
+ return sock_no_sendpage(sk->sk_socket, page, offset, size, flags);
+}
+
+/*
+ * Release all message fragments of the first in the transmit queue
+ */
+static void udpcp_release_xmit(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct sk_buff *skb;
+ struct udpcphdr *uh;
+
+ for (;;) {
+ skb = skb_dequeue(&dest->xmit);
+
+ uh = udpcp_hdr(skb);
+
+ if (udpcp_is_last_frag(uh) && uh->msgid) {
+ usk->stat.txMsgs++;
+ atomic_inc(&udpcp_txMsgs);
+ }
+
+ udpcp_dec_pending(sk);
+
+ kfree_skb(skb);
+ if (skb == dest->xmit_last)
+ break;
+ }
+
+ dest->xmit_wait = 0;
+ dest->xmit_last = 0;
+ dest->try = 0;
+}
+
+/*
+ * Set the sync state
+ */
+static void udpcp_sync(struct sock *sk, struct udpcp_dest *dest)
+{
+ dest->xmit_wait = 0;
+ dest->xmit_last = 0;
+ dest->try = 0;
+ dest->acks = 0;
+ dest->insync = 1;
+}
+
+/*
+ * Returns true if the first message in the transmit queue is a sync message
+ */
+static inline int udpcp_xmit_is_sync(struct udpcp_dest *dest)
+{
+ struct sk_buff *skb = skb_peek(&dest->xmit);
+
+ return skb && !udpcp_hdr(skb)->msgid;
+}
+
+static inline struct udpcphdr *udpcp_ack_scan(struct sk_buff *skb)
+{
+ struct udpcphdr *uh;
+
+ for (;;) {
+ uh = udpcp_hdr(skb);
+
+ if (!(ntohs(uh->msginfo) & UDPCP_SINGLE_ACK_FLAG)
+ || udpcp_is_last_frag(uh))
+ return uh;
+
+ skb = skb->next;
+ }
+}
+
+/*
+ * Handle an incoming ack
+ */
+static void udpcp_handle_ack(struct sock *sk, struct sk_buff *skb,
+ struct udpcp_dest *dest)
+{
+ struct udpcphdr *r_uh;
+ struct udpcphdr *q_uh;
+
+ if (!dest->acks)
+ return;
+
+ r_uh = udpcp_hdr(skb);
+
+ /*
+ * acks doesn't have a payload
+ */
+ if (r_uh->length)
+ return;
+
+ q_uh = udpcp_ack_scan(dest->xmit_wait);
+
+ /*
+ * message id, fragnum and fragamount must match the awaited message
+ * fragment
+ */
+ if (r_uh->msgid != q_uh->msgid)
+ return;
+
+ if (r_uh->fragnum != q_uh->fragnum)
+ return;
+
+ if (r_uh->fragamount != q_uh->fragamount)
+ return;
+
+ dest->acks--;
+
+ /*
+ * if last fragment release message
+ */
+ if (udpcp_is_last_frag(q_uh)) {
+ udpcp_release_xmit(sk, dest);
+
+ /*
+ * special handling for sync messages
+ */
+ if (r_uh->msgid == 0)
+ udpcp_sync(sk, dest);
+ } else
+ dest->xmit_wait = dest->xmit_wait->next;
+
+ /*
+ * try to transmit next message/fragment
+ */
+ udpcp_xmit(sk, dest);
+}
+
+/*
+ * Queue incoming message as owned by udpcp socket
+ */
+static void udpcp_set_owner_r(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct sk_buff *skb;
+
+ skb = dest->recv_msg;
+ skb_set_owner_r(skb, sk);
+
+ skb = skb_shinfo(skb)->frag_list;
+ if (!skb)
+ return;
+
+ for (;;) {
+ skb_set_owner_r(skb, sk);
+ if (udpcp_is_last_frag(udpcp_hdr(skb)))
+ break;
+ skb = skb->next;
+ }
+}
+
+/*
+ * Handle an incoming data message fragment
+ */
+static int udpcp_handle_data(struct sock *sk, struct sk_buff *skb,
+ struct udpcp_dest *dest)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct udpcphdr *uh = udpcp_hdr(skb);
+ unsigned short msginfo = ntohs(uh->msginfo);
+ unsigned short length = ntohs(uh->length);
+
+ /*
+ * special handling for sync messages
+ */
+ if (uh->msgid == 0) {
+ /*
+ * sync messages doesn't have a payload
+ */
+ if (length)
+ return 1;
+
+ /*
+ * sync messages doesn't have a ack rules
+ */
+ if (msginfo & (UDPCP_NO_ACK_FLAG | UDPCP_SINGLE_ACK_FLAG))
+ return 1;
+
+ udpcp_send_ack(sk, skb, dest,
+ memcmp(uh, &dest->lastmsg,
+ sizeof(dest->lastmsg)) ? 0 : 1);
+
+ udpcp_purge_incoming(sk, dest);
+
+ /*
+ * skip the first message in the queue if it is a sync messages
+ */
+ if (udpcp_xmit_is_sync(dest)) {
+ dest->acks--;
+ udpcp_dec_pending(sk);
+ kfree_skb(skb_dequeue(&dest->xmit));
+ }
+
+ if (!dest->insync)
+ udpcp_sync(sk, dest);
+
+ udpcp_xmit(sk, dest);
+
+ return -1;
+ }
+
+ if (!dest->insync)
+ return 1;
+
+ if (length > UDPCP_MAX_MSGSIZE)
+ return 1;
+
+ length += sizeof(struct udpcphdr);
+
+ /*
+ * if the message was still handled, send a duplicate ack
+ */
+ if (!memcmp(uh, &dest->lastmsg, sizeof(dest->lastmsg))) {
+ udpcp_send_ack(sk, skb, dest, 1);
+ return 1;
+ }
+
+ if (dest->recv_msg) {
+ /*
+ * if a fragment is already received validate the fragment
+ */
+ if ((uh->msgid != udpcp_hdr(dest->recv_msg)->msgid) ||
+ (uh->msginfo != udpcp_hdr(dest->recv_msg)->msginfo) ||
+ (uh->length != udpcp_hdr(dest->recv_msg)->length) ||
+ (uh->fragamount != udpcp_hdr(dest->recv_msg)->fragamount)
+ ) {
+ udpcp_purge_incoming(sk, dest);
+ goto newmsg;
+ }
+
+ if (uh->fragnum != udpcp_hdr(dest->recv_last)->fragnum + 1)
+ return 1;
+
+ if (dest->recv_msg->len + skb->len - sizeof(struct udpcphdr) >
+ length)
+ return 1;
+ } else {
+newmsg:
+ /*
+ * first fragment must have the number 0
+ */
+ if (uh->fragnum != 0)
+ return 1;
+
+ /*
+ * UDPCP data length cannot be smaller then the UDP data length
+ */
+ if (skb->len > length)
+ return 1;
+
+ /*
+ * id of the last received is not valid
+ */
+ if (dest->lastmsg.msgid == uh->msgid)
+ return 1;
+
+ /*
+ * check against receive buffer limit
+ */
+ if (atomic_read(&sk->sk_rmem_alloc) + length > sk->sk_rcvbuf)
+ return 1;
+ }
+
+ memset(&dest->lastmsg, 0, sizeof(dest->lastmsg));
+
+ if (!dest->recv_msg) {
+ /*
+ * store the first message fragment
+ */
+ if (skb->cloned) {
+ struct sk_buff *skbc;
+
+ skbc = skb_copy(skb, sk->sk_allocation);
+ if (skbc == NULL)
+ return 1;
+ kfree_skb(skb);
+ skb = skbc;
+ }
+ dest->recv_msg = skb;
+ } else {
+ /*
+ * store the consecutively message fragment
+ */
+ struct skb_shared_info *shinfo;
+
+ shinfo = skb_shinfo(dest->recv_msg);
+
+ if (!shinfo->frag_list)
+ shinfo->frag_list = skb;
+ else
+ dest->recv_last->next = skb;
+
+ skb_pull(skb, sizeof(struct udpcphdr));
+ dest->recv_msg->len += skb->len;
+ dest->recv_msg->data_len += skb->len;
+ }
+ dest->recv_last = skb;
+
+ msginfo = ntohs(uh->msginfo);
+
+ if (udpcp_is_last_frag(uh) || uh->fragamount == 0) {
+ /*
+ * last fragment: queue it to the socket sk_receive_queue
+ * and ack it
+ */
+
+ if (dest->recv_msg->len != length) {
+ udpcp_purge_incoming(sk, dest);
+ return 0;
+ }
+
+ if (!(msginfo & UDPCP_NO_ACK_FLAG))
+ udpcp_send_ack(sk, skb, dest, 0);
+
+ memcpy(dest->recv_msg->data + UDPCP_HDRSIZE,
+ dest->recv_msg->data, sizeof(struct udphdr));
+ skb_pull(dest->recv_msg, UDPCP_HDRSIZE);
+
+ usk->stat.rxMsgs++;
+ atomic_inc(&udpcp_rxMsgs);
+
+ /*
+ * set a flag for UDPCP message
+ */
+ UDP_SKB_CB(skb)->udpcp_flag = 1;
+
+ udpcp_set_owner_r(sk, dest);
+ skb_queue_tail(&sk->sk_receive_queue, dest->recv_msg);
+
+ /*
+ * call the original data available handler
+ */
+ if (usk->udp_data_ready)
+ usk->udp_data_ready(sk, dest->recv_msg->len);
+
+ dest->recv_msg = 0;
+ dest->recv_last = 0;
+ } else {
+ /*
+ * ack fragment if requiered
+ */
+ if (!(msginfo & UDPCP_NO_ACK_FLAG)
+ && !(msginfo & UDPCP_SINGLE_ACK_FLAG))
+ udpcp_send_ack(sk, skb, dest, 0);
+
+ /*
+ * setup timeout handler
+ */
+ dest->rx_time = jiffies;
+
+ if (!timer_pending(&usk->timer))
+ udpcp_timer(sk, dest->rx_time + usk->rx_timeout);
+ }
+
+ return 0;
+}
+
+/*
+ * Deal with received UDPCP frames - sort out what type source it is
+ * and hand of it to the udpcp_handle_packet function.
+ */
+static void udpcp_data_ready(struct sock *sk, int slen)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct sk_buff *skb;
+ struct udpcp_dest *dest;
+ struct udpcphdr *uh;
+ unsigned short msginfo;
+ int ret;
+
+ skb = skb_peek_tail(&sk->sk_receive_queue);
+
+ /*
+ * don't handle NULL pointer buffer and UDPCP messages
+ */
+ if (skb == NULL || UDP_SKB_CB(skb)->udpcp_flag) {
+ if (usk->udp_data_ready)
+ usk->udp_data_ready(sk, slen);
+ return;
+ }
+
+ __skb_unlink(skb, &sk->sk_receive_queue);
+ if (udpcp_validate_skb(skb)) {
+ kfree_skb(skb);
+
+ return;
+ }
+
+ skb_orphan(skb);
+
+ /*
+ * do UDP checksum
+ */
+ if (udp_lib_checksum_complete(skb)) {
+ UDP_INC_STATS_BH(sock_net(sk), UDP_MIB_INERRORS,
+ IS_UDPLITE(sk));
+ return;
+ }
+
+ if (unlikely(debug))
+ dump_msg("receive", skb, ip_hdr(skb)->saddr,
+ ip_hdr(skb)->daddr);
+
+ uh = udpcp_hdr(skb);
+ msginfo = ntohs(uh->msginfo);
+
+ /*
+ * handle only UDPCP protocol version 2
+ */
+ if ((msginfo & UDPCP_PROTOCOL_MASK) != UDPCP_PROTOCOL_VERSION_2) {
+ kfree_skb(skb);
+ return;
+ }
+
+ /*
+ * handle UDPCP checksum
+ */
+ if (msginfo & UDPCP_CHECKSUM_FLAG) {
+ u8 *data;
+ u32 data_len;
+ u32 chksum;
+
+ chksum = ntohl(uh->chksum);
+ data = (u8 *) skb->data + sizeof(struct udphdr);
+ data_len = skb->len - sizeof(struct udphdr);
+
+ uh->chksum = 0;
+
+ if (chksum != zlib_adler32(1, data, data_len)) {
+ kfree_skb(skb);
+ usk->stat.crcErrors++;
+ atomic_inc(&udpcp_crcErrors);
+ return;
+ }
+ }
+
+ dest = __find_dest(sk, ip_hdr(skb)->saddr, udp_hdr(skb)->source);
+
+ if (!dest) {
+ /*
+ * new communication destination must start with an sync message
+ */
+ if (((msginfo & UDPCP_MSG_TYPE_MASK) != UDPCP_MSG_TYPE_DATA) ||
+ (uh->msgid != 0)) {
+ kfree_skb(skb);
+ return;
+ }
+
+ dest = new_dest(sk, ip_hdr(skb)->saddr, udp_hdr(skb)->source);
+
+ if (!dest) {
+ kfree_skb(skb);
+ return;
+ }
+ }
+
+ /*
+ * handle message type
+ */
+ switch (msginfo & UDPCP_MSG_TYPE_MASK) {
+ case UDPCP_MSG_TYPE_DATA:
+ if (!(dest->use_flag & RX_NODE)) {
+ dest->use_flag |= RX_NODE;
+ usk->stat.rxNodes++;
+ atomic_inc(&udpcp_rxNodes);
+ }
+
+ ret = udpcp_handle_data(sk, skb, dest);
+
+ if (ret > 0) {
+ dest->rxDiscardedFrags++;
+ usk->stat.rxDiscardedFrags++;
+ atomic_inc(&udpcp_rxDiscardedFrags);
+ }
+ break;
+ case UDPCP_MSG_TYPE_ACK:
+ udpcp_handle_ack(sk, skb, dest);
+ default:
+ ret = 1;
+ break;
+ }
+ if (ret)
+ kfree_skb(skb);
+}
+
+/*
+ * Set socket options
+ */
+static int udpcp_setsockopt(struct sock *sk, int level, int optname,
+ char __user *optval, unsigned int optlen)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int val, ret;
+
+ if (level != SOL_UDPCP) {
+ if (udp_prot.setsockopt) {
+ ret = udp_prot.setsockopt(sk, level, optname, optval,
+ optlen);
+ check_timeout(sk);
+ return ret;
+ }
+ return -ENOPROTOOPT;
+ }
+
+ if (optlen < sizeof(int))
+ return -EINVAL;
+
+ if (get_user(val, (int __user *)optval))
+ return -EFAULT;
+
+ switch (optname) {
+ case UDPCP_OPT_TRANSFER_MODE:
+ switch (val) {
+ case UDPCP_NOACK:
+ case UDPCP_ACK:
+ case UDPCP_SINGLE_ACK:
+ usk->ackmode = val;
+ break;
+ default:
+ return -EINVAL;
+ }
+ break;
+ case UDPCP_OPT_CHECKSUM_MODE:
+ switch (val) {
+ case UDPCP_NOCHECKSUM:
+ case UDPCP_CHECKSUM:
+ usk->chkmode = val;
+ break;
+ default:
+ return -EINVAL;
+ }
+ break;
+
+ case UDPCP_OPT_TX_TIMEOUT:
+ if ((val < 1) || (val > UDPCP_MAX_WAIT_SEC * 1000))
+ return -EINVAL;
+ usk->tx_timeout = msecs_to_jiffies(val);
+ break;
+
+ case UDPCP_OPT_RX_TIMEOUT:
+ if ((val < 1) || (val > UDPCP_MAX_WAIT_SEC * 1000))
+ return -EINVAL;
+ usk->rx_timeout = msecs_to_jiffies(val);
+ break;
+
+ case UDPCP_OPT_MAXTRY:
+ if ((val < 1) || (val > 10))
+ return -EINVAL;
+ usk->maxtry = val;
+ break;
+
+ case UDPCP_OPT_OUTSTANDING_ACKS:
+ if ((val < 1) || (val > 255))
+ return -EINVAL;
+ usk->acks = val;
+ break;
+
+ default:
+ return -ENOPROTOOPT;
+ }
+ return 0;
+}
+
+/*
+ * Get socket options
+ */
+static int udpcp_getsockopt(struct sock *sk, int level, int optname,
+ char __user *optval, int __user *optlen)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int val, len, ret;
+
+ if (level != SOL_UDPCP) {
+ if (udp_prot.getsockopt) {
+ ret = udp_prot.getsockopt(sk, level, optname, optval,
+ optlen);
+ check_timeout(sk);
+ return ret;
+ }
+ return -ENOPROTOOPT;
+ }
+
+ if (get_user(len, optlen))
+ return -EFAULT;
+
+ len = min_t(unsigned int, len, sizeof(int));
+
+ if (len < 0)
+ return -EINVAL;
+
+ switch (optname) {
+ case UDPCP_OPT_TRANSFER_MODE:
+ val = usk->ackmode;
+ break;
+
+ case UDPCP_OPT_CHECKSUM_MODE:
+ val = usk->chkmode;
+ break;
+
+ case UDPCP_OPT_TX_TIMEOUT:
+ val = jiffies_to_msecs(usk->tx_timeout);
+ break;
+
+ case UDPCP_OPT_MAXTRY:
+ val = usk->maxtry;
+ break;
+
+ case UDPCP_OPT_OUTSTANDING_ACKS:
+ val = usk->acks;
+ break;
+
+ default:
+ return -ENOPROTOOPT;
+ }
+
+ if (put_user(len, optlen))
+ return -EFAULT;
+ if (copy_to_user(optval, &val, len))
+ return -EFAULT;
+ return 0;
+}
+
+/*
+ * ioctl() requests applicable to the UDPCP protocol
+ */
+int udpcp_ioctl(struct sock *sk, int cmd, unsigned long arg)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int ret = 0;
+
+ switch (cmd) {
+ case UDPCP_IOCTL_GET_STATISTICS:
+ lock_sock(sk);
+ if (copy_to_user((void *)arg, &usk->stat, sizeof(usk->stat)))
+ ret = -EFAULT;
+ udpcp_release_sock(sk);
+ break;
+
+ case UDPCP_IOCTL_RESET_STATISTICS:
+ lock_sock(sk);
+ usk->stat.txMsgs = 0;
+ usk->stat.rxMsgs = 0;
+ usk->stat.txTimeout = 0;
+ usk->stat.rxTimeout = 0;
+ usk->stat.txRetries = 0;
+ usk->stat.rxDiscardedFrags = 0;
+ usk->stat.crcErrors = 0;
+ udpcp_release_sock(sk);
+ break;
+
+ case UDPCP_IOCTL_SYNC:
+ if (arg)
+ ret = wait_event_interruptible_timeout(usk->wq,
+ !usk->pending, msecs_to_jiffies(arg));
+ else
+ ret = wait_event_interruptible(usk->wq, !usk->pending);
+
+ break;
+
+ default:
+ if (udp_prot.ioctl) {
+ ret = udp_prot.ioctl(sk, cmd, arg);
+ check_timeout(sk);
+ } else
+ ret = -ENOIOCTLCMD;
+ break;
+ }
+ return ret;
+}
+
+/*
+ * This function will be called by recv(), recvfrom() and revmsg()
+ */
+int udpcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
+ size_t len, int noblock, int flags, int *addr_len)
+{
+ int ret;
+
+ ret = udp_prot.recvmsg(iocb, sk, msg, len, noblock, flags, addr_len);
+ check_timeout(sk);
+ return ret;
+}
+
+/*
+ * This function will be called by socket() and initialized the socket
+ */
+static int udpcp_sockinit(struct sock *sk)
+{
+ int ret;
+ struct udpcp_sock *usk;
+
+ sk->sk_protocol = SOL_UDP;
+ sk->sk_allocation = GFP_ATOMIC;
+ if (udp_prot.init) {
+ ret = udp_prot.init(sk);
+
+ if (ret)
+ return ret;
+ }
+
+ usk = udpcp_sk(sk);
+ usk->timer.expires = 0;
+ usk->timer.function = udpcp_timeout;
+ usk->timer.data = (long)sk;
+ init_timer(&usk->timer);
+ INIT_LIST_HEAD(&usk->destlist);
+ init_waitqueue_head(&usk->wq);
+ usk->pending = 0;
+ usk->ackmode = UDPCP_ACK;
+ usk->chkmode = UDPCP_CHECKSUM;
+ usk->maxtry = UDPCP_TX_MAXTRY;
+ usk->acks = UDPCP_OUTSTANDING_ACKS;
+ usk->tx_timeout = msecs_to_jiffies(UDPCP_TX_TIMEOUT);
+ usk->rx_timeout = msecs_to_jiffies(UDPCP_RX_TIMEOUT);
+ usk->udp_data_ready = sk->sk_data_ready;
+ sk->sk_data_ready = udpcp_data_ready;
+ usk->udpsock.pending = 0;
+ skb_queue_head_init(&usk->assembly);
+ usk->assembly_len = 0;
+ usk->assembly_dest = NULL;
+
+ spin_lock_irq(&udpcp_lock);
+ list_add_tail(&usk->udpcplist, &udpcp_list);
+ spin_unlock_irq(&udpcp_lock);
+
+#ifdef MODULE
+ try_module_get(THIS_MODULE);
+#endif
+ return 0;
+}
+
+/*
+ * This function will be called by close()
+ */
+static void udpcp_destroy(struct sock *sk)
+{
+ struct list_head *p;
+ struct list_head *n;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ spin_lock_irq(&udpcp_lock);
+ list_del(&usk->udpcplist);
+ spin_unlock_irq(&udpcp_lock);
+
+ if (udp_prot.destroy)
+ udp_prot.destroy(sk);
+
+ lock_sock(sk);
+
+ del_timer_sync(&usk->timer);
+ sk->sk_data_ready = usk->udp_data_ready;
+
+ skb_queue_purge(&usk->assembly);
+
+ list_for_each_safe(p, n, &usk->destlist) {
+ struct udpcp_dest *dest;
+
+ dest = list_to_udpcpdest(p);
+
+ skb_queue_purge(&dest->xmit);
+
+ kfree_skb(dest->recv_msg);
+
+ if (dest->rt)
+ dst_release(&dest->rt->dst);
+
+ kfree(dest);
+ }
+
+ atomic_sub(usk->stat.txNodes, &udpcp_txNodes);
+ atomic_sub(usk->stat.rxNodes, &udpcp_rxNodes);
+
+ usk->pending = 0;
+
+ if (waitqueue_active(&usk->wq))
+ wake_up_interruptible(&usk->wq);
+
+ release_sock(sk);
+
+#ifdef MODULE
+ module_put(THIS_MODULE);
+#endif
+}
+
+static struct proto udpcp_prot;
+
+/*
+ * inet protocol stack descriptor
+ */
+static struct inet_protosw udpcp_protosw = {
+ .type = SOCK_DGRAM,
+ .protocol = PF_UDPCP,
+ .prot = &udpcp_prot,
+ .ops = &inet_dgram_ops,
+ .no_check = UDP_CSUM_DEFAULT,
+ .flags = 0,
+};
+
+#ifdef CONFIG_PROC_FS
+/*
+ * The following functions handles the /proc/net/udpcp entry
+ */
+struct udpcp_seq_afinfo {
+ char *name;
+ const struct file_operations seq_fops;
+ const struct seq_operations seq_ops;
+};
+
+struct udpcp_iter_state {
+ struct seq_net_private p;
+ struct sock *sk;
+ struct list_head *list;
+ int bucket;
+};
+
+static int udpcp_get_destlist(struct udpcp_sock *usk,
+ struct udpcp_iter_state *state)
+{
+ struct sock *sk = (struct sock *)usk;
+
+ if (sock_flag(sk, SOCK_DEAD))
+ return 0;
+
+ sock_hold(sk);
+ if (!list_empty(&usk->destlist)) {
+ state->sk = sk;
+ state->list = &usk->destlist;
+ return 1;
+ }
+ sock_put(sk);
+
+ return 0;
+}
+
+static inline int udpcp_next_dest(struct udpcp_iter_state *state)
+{
+ struct sock *sk = state->sk;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int found = 0;
+
+ if (sock_flag(sk, SOCK_DEAD))
+ return 0;
+
+ lock_sock(sk);
+ if (!list_is_last(state->list, &usk->destlist)) {
+ state->list = state->list->next;
+ state->bucket++;
+ found = 1;
+ }
+ udpcp_release_sock(sk);
+ return found;
+}
+
+static void *udpcp_get_next(struct seq_file *seq)
+{
+ struct udpcp_iter_state *state = seq->private;
+ struct udpcp_sock *usk;
+ struct sock *sk;
+
+ while (state) {
+ if (udpcp_next_dest(state))
+ return state;
+
+ sk = state->sk;
+ usk = udpcp_sk(sk);
+
+ spin_lock_irq(&udpcp_lock);
+ while (!list_is_last(&usk->udpcplist, &udpcp_list)) {
+ usk = list_entry(usk->udpcplist.next, struct udpcp_sock,
+ udpcplist);
+
+ if (udpcp_get_destlist(usk, state))
+ goto found;
+ }
+ state->sk = NULL;
+ state = NULL;
+found:
+ spin_unlock_irq(&udpcp_lock);
+ sock_put(sk);
+ }
+ return state;
+}
+
+static void *udpcp_get_first(struct seq_file *seq)
+{
+ struct list_head *p;
+ struct udpcp_iter_state *state = seq->private;
+ int found = 0;
+
+ if (!state)
+ return NULL;
+
+ spin_lock_irq(&udpcp_lock);
+ list_for_each(p, &udpcp_list) {
+ found = udpcp_get_destlist(list_to_udpcpsock(p), state);
+ if (found)
+ goto found;
+ }
+found:
+ spin_unlock_irq(&udpcp_lock);
+
+ if (!found)
+ return NULL;
+ return udpcp_get_next(seq);
+}
+
+static void *udpcp_get_idx(struct seq_file *seq, loff_t pos)
+{
+ if (!udpcp_get_first(seq))
+ return NULL;
+
+ while (pos--) {
+ if (!udpcp_get_next(seq))
+ return NULL;
+ }
+ return seq->private;
+}
+
+static void *udpcp_seq_start(struct seq_file *seq, loff_t * pos)
+{
+ return *pos ? udpcp_get_idx(seq, *pos - 1) : SEQ_START_TOKEN;
+}
+
+static void *udpcp_seq_next(struct seq_file *seq, void *v, loff_t * pos)
+{
+ void *private;
+
+ if (v == SEQ_START_TOKEN)
+ private = udpcp_get_idx(seq, 0);
+ else
+ private = udpcp_get_next(seq);
+
+ ++*pos;
+ return private;
+}
+
+static void udpcp_seq_stop(struct seq_file *seq, void *v)
+{
+ struct udpcp_iter_state *state = seq->private;
+
+ if (state->sk)
+ sock_put(state->sk);
+}
+
+static int udpcp_seq_open(struct inode *inode, struct file *file)
+{
+ struct udpcp_seq_afinfo *afinfo = PDE(inode)->data;
+ int err;
+
+ err = seq_open_net(inode, file, &afinfo->seq_ops,
+ sizeof(struct udpcp_iter_state));
+ if (err < 0)
+ return err;
+
+ return err;
+}
+
+int udpcp_proc_register(struct net *net, struct udpcp_seq_afinfo *afinfo)
+{
+ struct proc_dir_entry *p;
+ int rc = 0;
+
+ p = proc_create_data(afinfo->name, S_IRUGO, net->proc_net,
+ &afinfo->seq_fops, afinfo);
+ if (!p)
+ rc = -ENOMEM;
+ return rc;
+}
+
+void udpcp_proc_unregister(struct net *net, struct udpcp_seq_afinfo *afinfo)
+{
+ proc_net_remove(net, afinfo->name);
+}
+
+static unsigned int udpcp_tx_queue_len(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct sk_buff *skb;
+ unsigned int n;
+
+ n = 0;
+ skb_queue_walk(&dest->xmit, skb)
+ n += skb->len;
+ return n;
+}
+
+static unsigned int udpcp_rx_queue_len(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct sk_buff *skb;
+ unsigned int n;
+
+ n = 0;
+ skb_queue_walk(&sk->sk_receive_queue, skb) {
+ if (udp_hdr(skb)->source == dest->port
+ && ip_hdr(skb)->saddr == dest->addr)
+ n += skb->len;
+ }
+ return n;
+}
+
+static void udpcp_format_sock(struct seq_file *seq, int *len)
+{
+ struct udpcp_iter_state *state = seq->private;
+ struct sock *sk = state->sk;
+ struct inet_sock *inet = inet_sk(sk);
+ struct udpcp_dest *p = list_to_udpcpdest(state->list);
+ __be32 src = inet->inet_rcv_saddr;
+ __u16 srcp = ntohs(inet->inet_sport);
+ __be32 dest = p->addr;
+ __u16 destp = ntohs(p->port);
+
+ lock_sock(sk);
+ seq_printf(seq, "%4d: %08X:%04X %08X:%04X"
+ " %02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p %u%n",
+ state->bucket, src, srcp, dest, destp, sk->sk_state,
+ udpcp_tx_queue_len(sk, p),
+ udpcp_rx_queue_len(sk, p),
+ 0, 0L, p->txRetries, sock_i_uid(sk),
+ p->txTimeout, sock_i_ino(sk),
+ atomic_read(&sk->sk_refcnt), sk, p->rxTimeout,
+ len);
+ udpcp_release_sock(sk);
+}
+
+int udpcp_seq_show(struct seq_file *seq, void *v)
+{
+ if (v == SEQ_START_TOKEN)
+ seq_printf(seq, "%-127s\n",
+ " sl local_address rem_address st tx_queue "
+ "rx_queue tr tm->when retrnsmt uid timeout "
+ "inode ref pointer drops");
+ else {
+ int len;
+
+ udpcp_format_sock(seq, &len);
+ seq_printf(seq, "%*s\n", 127 - len, "");
+ }
+ return 0;
+}
+
+static struct udpcp_seq_afinfo udpcp_seq_afinfo = {
+ .name = "udpcp",
+ .seq_fops = {
+ .owner = THIS_MODULE,
+ .open = udpcp_seq_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = seq_release_net,
+ },
+ .seq_ops = {
+ .show = udpcp_seq_show,
+ .start = udpcp_seq_start,
+ .next = udpcp_seq_next,
+ .stop = udpcp_seq_stop,
+ },
+};
+
+static int udpcp_proc_init_net(struct net *net)
+{
+ return udpcp_proc_register(net, &udpcp_seq_afinfo);
+}
+
+static void udpcp_proc_exit_net(struct net *net)
+{
+ udpcp_proc_unregister(net, &udpcp_seq_afinfo);
+}
+
+static struct pernet_operations udpcp_net_ops = {
+ .init = udpcp_proc_init_net,
+ .exit = udpcp_proc_exit_net,
+};
+
+int __init udpcp_proc_init(void)
+{
+ return register_pernet_subsys(&udpcp_net_ops);
+}
+
+void udpcp_proc_exit(void)
+{
+ unregister_pernet_subsys(&udpcp_net_ops);
+}
+#endif /* CONFIG_PROC_FS */
+
+/*
+ * Install and init module
+ */
+static int __init udpcp_init(void)
+{
+ int ret;
+ struct proc_dir_entry *proc_entry = NULL;
+
+ spin_lock_init(&udpcp_lock);
+
+ INIT_LIST_HEAD(&udpcp_list);
+
+ /*
+ * to prevent to rewrite the whole UDP protocol,
+ * assign struct proto udp to the struct proto udpcp
+ */
+ udpcp_prot = udp_prot;
+
+ /*
+ * change the protocol name
+ */
+ strcpy(udpcp_prot.name, "UDPCP");
+
+ /*
+ * overload the following function, all other
+ * functions will use the UDP protocol functions
+ */
+ udpcp_prot.sendmsg = udpcp_sendmsg;
+ udpcp_prot.sendpage = udpcp_sendpage;
+ udpcp_prot.init = udpcp_sockinit;
+ udpcp_prot.destroy = udpcp_destroy;
+ udpcp_prot.setsockopt = udpcp_setsockopt;
+ udpcp_prot.getsockopt = udpcp_getsockopt;
+ udpcp_prot.ioctl = udpcp_ioctl;
+ udpcp_prot.recvmsg = udpcp_recvmsg;
+
+ /*
+ * fix the object size for the embedded udpcp_sock structure
+ */
+ udpcp_prot.obj_size = sizeof(struct udpcp_sock);
+
+ /*
+ * register the UDPCP protocol
+ */
+ ret = proto_register(&udpcp_prot, 1);
+ if (ret)
+ return ret;
+
+ /*
+ * register the inet socket for UDPCP
+ */
+ inet_register_protosw(&udpcp_protosw);
+
+#ifdef CONFIG_PROC_FS
+ /*
+ * register /proc/driver/udpcp entry
+ */
+ proc_entry =
+ create_proc_read_entry(UDPCP_PROC, S_IRUSR | S_IRGRP | S_IROTH,
+ NULL, udpcp_proc, NULL);
+
+ if (!proc_entry) {
+ ret = -ENOMEM;
+ goto err;
+ }
+
+ /*
+ * register /proc/net/udpcp entry
+ */
+ ret = udpcp_proc_init();
+
+ if (ret)
+ goto err;
+#endif
+ pr_info("UDPCP protocol stack version " VERSION "\n");
+ return 0;
+#ifdef CONFIG_PROC_FS
+err:
+ if (proc_entry)
+ remove_proc_entry(UDPCP_PROC, NULL);
+ proto_unregister(&udpcp_prot);
+ return ret;
+#endif
+}
+
+/*
+ * Cleanup and exit module
+ */
+static void __exit udpcp_exit(void)
+{
+#ifdef CONFIG_PROC_FS
+ udpcp_proc_exit();
+ remove_proc_entry(UDPCP_PROC, NULL);
+#endif
+ inet_unregister_protosw(&udpcp_protosw);
+ proto_unregister(&udpcp_prot);
+}
+
+module_init(udpcp_init);
+module_exit(udpcp_exit);
+
+MODULE_AUTHOR("Stefani Seibold <stefani@seibold.net>");
+MODULE_DESCRIPTION("UDPCP protocol stack v" VERSION);
+MODULE_LICENSE("GPL");
+
--
1.7.3.4
^ permalink raw reply related
* Re: [PATCH net-next-2.6] tg3: fix warnings
From: David Miller @ 2011-01-01 21:51 UTC (permalink / raw)
To: eric.dumazet; +Cc: rjw, netdev, mcarlson, mchan, linux-pm, jesse
In-Reply-To: <1293895366.2535.43.camel@edumazet-laptop>
From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Sat, 01 Jan 2011 16:22:46 +0100
> Happy new year everybody ;)
Same to you Eric :)
> [PATCH net-next-2.6] tg3: fix warnings
>
> In case CONFIG_PM_SLEEP is disabled, we dont need tg3_suspend() and
> tg3_resume().
>
> drivers/net/tg3.c:15056: warning: ‘tg3_suspend’ defined but not used
> drivers/net/tg3.c:15110: warning: ‘tg3_resume’ defined but not used
>
> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Applied, thanks.
^ permalink raw reply
* Re: [PATCH net-next 00/20] Delete more semi-useless stuff from TIPC.
From: David Miller @ 2011-01-01 21:59 UTC (permalink / raw)
To: paul.gortmaker; +Cc: netdev, allan.stephens
In-Reply-To: <1293857975-30267-1-git-send-email-paul.gortmaker@windriver.com>
From: Paul Gortmaker <paul.gortmaker@windriver.com>
Date: Fri, 31 Dec 2010 23:59:15 -0500
> There really isn't a lot to see here technically. There is a bunch of
> stuff in TIPC that I guess falls into the category of "seemed like a
> good idea at the time", but is now just inactive or at best unstable
> prototype code for things like slave nodes and routing features and
> mulitcluster support. My favourite is this one:
>
> ----------
> config TIPC_CLUSTERS
> int "Maximum number of clusters in a zone"
> depends on TIPC_ADVANCED
> range 1 1
> default "1"
> help
> Specifies how many clusters can be supported in a TIPC zone.
>
> *** Currently TIPC only supports a single cluster per zone. ***
> ----------
>
> The remaining commits are just de-uglifying what is left, fixing whitespace
> crimes and other common errors typically picked up in janitor patches.
All applied, thanks Paul.
net/tipc/log.h still uses "_TIPC_DBG_H" for it's multiple-include protection
tests. Please send me a patch to rename that to _TIPC_LOG_H
Thanks.
^ permalink raw reply
* Re: [PATCH net-next] sky2: implement 64 bit stats
From: David Miller @ 2011-01-01 22:02 UTC (permalink / raw)
To: shemminger; +Cc: netdev
In-Reply-To: <20101231173427.17f07ad3@nehalam>
From: Stephen Hemminger <shemminger@vyatta.com>
Date: Fri, 31 Dec 2010 17:34:27 -0800
> This implements 64 bit statistics support and fixes races when reading
> counter values. The PHY counters can only be accessed 16 bits at a time,
> so they are subject to carry races.
>
> NB:
> * TX/RX counters are maintained in software because the the hardware packet count
> is only a 32 bit value.
>
> * Error counters are really only 32 bit.
>
> * Old 32 bit counter fields in dev->stats still used for some
> software counters
>
> Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Applied, thanks Stephen.
^ permalink raw reply
* Re: [PATCH] new UDPCP Communication Protocol
From: Eric Dumazet @ 2011-01-01 22:23 UTC (permalink / raw)
To: stefani; +Cc: linux-kernel, akpm, davem, netdev, shemminger
In-Reply-To: <1293918276-28773-1-git-send-email-stefani@seibold.net>
Le samedi 01 janvier 2011 à 22:44 +0100, stefani@seibold.net a écrit :
> From: Stefani Seibold <stefani@seibold.net>
>
> Changelog:
> 31.12.2010 first proposal
> 01.01.2011 code cleanup and fixes suggest by Eric Dumazet
>
> UDPCP is a communication protocol specified by the Open Base Station
> Architecture Initiative Special Interest Group (OBSAI SIG). The
> protocol is based on UDP and is designed to meet the needs of "Mobile
> Communcation Base Station" internal communications. It is widely used by
> the major networks infrastructure supplier.
>
> The UDPCP communication service supports the following features:
>
> -Connectionless communication for serial mode data transfer
> -Acknowledged and unacknowledged transfer modes
> -Retransmissions Algorithm
> -Checksum Algorithm using Adler32
> -Fragmentation of long messages (disassembly/reassembly) to match to the MTU
> during transport:
> -Broadcasting and multicasting messages to multiple peers in unacknowledged
> transfer mode
>
> UDPCP supports application level messages up to 64 KBytes (limited by 16-bit
> packet data length field). Messages that are longer than the MTU will be
> fragmented to the MTU.
>
> UDPCP provides a reliable transport service that will perform message
> retransmissions in case transport failures occur.
>
> The code is also a nice example how to implement a UDP based protocol as
> a kernel socket modules.
>
> Due the nature of UDPCP which has no sliding windows support, the latency has a
> huge impact. The perfomance increase by implementing as a kernel module is
> about the factor 10, because there are no context switches and data packets or
> ACKs will be handled in the interrupt service.
>
> There are no side effects to the network subsystems so i ask for merge it
> into linux-next. Hope you like it.
>
> The patch is against 2.6.37-rc8
>
Hi Stefani
1) Please base your next trys on net-next-2.6 : This is the reference
for stuff like that. It probably does not matter a lot, but still...
2) I still see some _irq() variants of spinlock(). Its not necessary in
network stack at the level you are working (process context, and
softirqs)
Please only use _bh() variants, it's enough.
3) I see UDPLITE references in your code. Are you sure UDPCP protocol
can really work on top of UDPLITE ? I think not, so please remove dead
code.
4) udpcp_release_sock() seems expensive to me. Why not testing
usk->timeout before releasing sock lock, and save a lock/unlock pair ?
static inline void udpcp_release_sock(struct sock *sk)
{
struct udpcp_sock *usk = udpcp_sk(sk);
if (usk->timeout)
udpcp_handle_timeout(sk);
release_sock(sk);
}
5) In udpcp_timeout(), if you find socket is locked by user, you set
timeout to one and rearm a timer to udpcp_timer(sk, jiffies + 1);
Why is it needed, since user process is going to handle the timeout
indication from udpcp_release_sock() ?
Thanks
^ permalink raw reply
* Re: [PATCH net-next 00/20] Delete more semi-useless stuff from TIPC.
From: Paul Gortmaker @ 2011-01-01 22:53 UTC (permalink / raw)
To: David Miller; +Cc: netdev, allan.stephens
In-Reply-To: <20110101.135952.226766850.davem@davemloft.net>
[Re: [PATCH net-next 00/20] Delete more semi-useless stuff from TIPC.] On 01/01/2011 (Sat 13:59) David Miller wrote:
> From: Paul Gortmaker <paul.gortmaker@windriver.com>
> Date: Fri, 31 Dec 2010 23:59:15 -0500
>
> > There really isn't a lot to see here technically. There is a bunch of
> > stuff in TIPC that I guess falls into the category of "seemed like a
> > good idea at the time", but is now just inactive or at best unstable
> > prototype code for things like slave nodes and routing features and
> > mulitcluster support. My favourite is this one:
> >
> > ----------
> > config TIPC_CLUSTERS
> > int "Maximum number of clusters in a zone"
> > depends on TIPC_ADVANCED
> > range 1 1
> > default "1"
> > help
> > Specifies how many clusters can be supported in a TIPC zone.
> >
> > *** Currently TIPC only supports a single cluster per zone. ***
> > ----------
> >
> > The remaining commits are just de-uglifying what is left, fixing whitespace
> > crimes and other common errors typically picked up in janitor patches.
>
> All applied, thanks Paul.
>
> net/tipc/log.h still uses "_TIPC_DBG_H" for it's multiple-include protection
> tests. Please send me a patch to rename that to _TIPC_LOG_H
Ah crap, I should have spotted that. Thanks for noticing it. Patch
follows.
Paul.
----
>From 443eb42b4e3e45b54698ee787ce7e5fb16acb57d Mon Sep 17 00:00:00 2001
From: Paul Gortmaker <paul.gortmaker@windriver.com>
Date: Sat, 1 Jan 2011 17:32:04 -0500
Subject: [PATCH] tipc: update log.h re-include protection to reflect new name
The tipc/dbg.h file was recently renamed to tipc/log.h,
but the re-include define was not updated accordingly.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
---
net/tipc/log.h | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/tipc/log.h b/net/tipc/log.h
index f4343bb..2248d96 100644
--- a/net/tipc/log.h
+++ b/net/tipc/log.h
@@ -34,8 +34,8 @@
* POSSIBILITY OF SUCH DAMAGE.
*/
-#ifndef _TIPC_DBG_H
-#define _TIPC_DBG_H
+#ifndef _TIPC_LOG_H
+#define _TIPC_LOG_H
/**
* struct print_buf - TIPC print buffer structure
--
1.7.3.3
^ permalink raw reply related
* Re: [PATCH net-next 00/20] Delete more semi-useless stuff from TIPC.
From: David Miller @ 2011-01-01 22:56 UTC (permalink / raw)
To: paul.gortmaker; +Cc: netdev, allan.stephens
In-Reply-To: <20110101225330.GF30150@windriver.com>
From: Paul Gortmaker <paul.gortmaker@windriver.com>
Date: Sat, 1 Jan 2011 17:53:31 -0500
> Subject: [PATCH] tipc: update log.h re-include protection to reflect new name
>
> The tipc/dbg.h file was recently renamed to tipc/log.h,
> but the re-include define was not updated accordingly.
>
> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Applied, thanks Paul.
^ permalink raw reply
* [PATCH net-next] net/Space: delete orphaned externs from deleted drivers
From: Paul Gortmaker @ 2011-01-01 23:15 UTC (permalink / raw)
To: davem; +Cc: netdev, Paul Gortmaker
The drivers associated with the prototypes in this commit have
been deleted some time ago, but the externs escaped detection.
Using a simple "git grep" shows that these references are
historical artefacts, only mentioned by the deleted lines.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
---
drivers/net/Space.c | 5 -----
1 files changed, 0 insertions(+), 5 deletions(-)
diff --git a/drivers/net/Space.c b/drivers/net/Space.c
index 9bb405b..068c356 100644
--- a/drivers/net/Space.c
+++ b/drivers/net/Space.c
@@ -55,8 +55,6 @@ extern struct net_device *eth16i_probe(int unit);
extern struct net_device *i82596_probe(int unit);
extern struct net_device *ewrk3_probe(int unit);
extern struct net_device *el1_probe(int unit);
-extern struct net_device *wavelan_probe(int unit);
-extern struct net_device *arlan_probe(int unit);
extern struct net_device *el16_probe(int unit);
extern struct net_device *elmc_probe(int unit);
extern struct net_device *elplus_probe(int unit);
@@ -68,7 +66,6 @@ extern struct net_device *ni5010_probe(int unit);
extern struct net_device *ni52_probe(int unit);
extern struct net_device *ni65_probe(int unit);
extern struct net_device *sonic_probe(int unit);
-extern struct net_device *SK_init(int unit);
extern struct net_device *seeq8005_probe(int unit);
extern struct net_device *smc_init(int unit);
extern struct net_device *atarilance_probe(int unit);
@@ -76,8 +73,6 @@ extern struct net_device *sun3lance_probe(int unit);
extern struct net_device *sun3_82586_probe(int unit);
extern struct net_device *apne_probe(int unit);
extern struct net_device *cs89x0_probe(int unit);
-extern struct net_device *hplance_probe(int unit);
-extern struct net_device *bagetlance_probe(int unit);
extern struct net_device *mvme147lance_probe(int unit);
extern struct net_device *tc515_probe(int unit);
extern struct net_device *lance_probe(int unit);
--
1.7.3.3
^ permalink raw reply related
* [PATCH net-next] netdev: Update status of 8390 based drivers in MAINTAINERS
From: Paul Gortmaker @ 2011-01-01 23:28 UTC (permalink / raw)
To: davem; +Cc: netdev, Paul Gortmaker
With the original 8 bit ISA ne1000 card being over 20 years old, it
only makes sense to consider ne.c and all the other toplevel 8390
based driver files as legacy for obsolete hardware. The most
recent thing made in large quantities that was 8390 based were
those crazy PCI ne2k clones - and even they are now 10+ years old.
Also remove myself as maintainer, since the only changes to these
drivers going forward will be the generic API type changes that
touch all drivers.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
---
MAINTAINERS | 3 +--
1 files changed, 1 insertions(+), 2 deletions(-)
diff --git a/MAINTAINERS b/MAINTAINERS
index 4641289..aa835f7 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -166,9 +166,8 @@ F: drivers/serial/8250*
F: include/linux/serial_8250.h
8390 NETWORK DRIVERS [WD80x3/SMC-ELITE, SMC-ULTRA, NE2000, 3C503, etc.]
-M: Paul Gortmaker <p_gortmaker@yahoo.com>
L: netdev@vger.kernel.org
-S: Maintained
+S: Orphan / Obsolete
F: drivers/net/*8390*
F: drivers/net/ax88796.c
--
1.7.3.3
^ permalink raw reply related
* Re: [PATCH 2.6.36] vlan: Avoid hwaccel vlan packets when vid not used
From: Jesse Gross @ 2011-01-02 0:27 UTC (permalink / raw)
To: Eric Dumazet
Cc: Matt Carlson, Michael Leun, Michael Chan, David Miller,
Ben Greear, linux-kernel@vger.kernel.org, netdev@vger.kernel.org
In-Reply-To: <1293901393.2535.49.camel@edumazet-laptop>
On Sat, Jan 1, 2011 at 12:03 PM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> Le mardi 14 décembre 2010 à 11:15 -0800, Matt Carlson a écrit :
>
>> Thanks for the comments Jesse. Below is an updated patch.
>>
>> Michael, I'm wondering if the difference in behavior can be explained by
>> the presence or absence of management firmware. Can you look at the
>> driver sign-on messages in your syslogs for ASF[]? I'm half expecting
>> the 5752 to show "ASF[0]" and the 5714 to show "ASF[1]". If you see
>> this, and the below patch doesn't fix the problem, let me know. I have
>> another test I'd like you to run.
>>
>> ----
>>
>> [PATCH] tg3: Use new VLAN code
>>
>> This patch pivots the tg3 driver to the new VLAN infrastructure.
>> All references to vlgrp have been removed and all VLAN code is
>> unconditionally active.
>>
>> Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
[...]
> Hi Matt.
>
> Any news on this patch ?
>
> Without it, net-next-2.6 doesnt work for me on a vlan setup on top of
> bonding.
>
> (bond0 : eth1 & eth2, eth1 being bnx2, eth2 beging tg3)
>
> ip link add link bond0 vlan.103 type vlan id 103
> ip addr add 192.168.20.110/24 dev vlan.103
> ip link set vlan.103 up
>
>
> If active slave is eth1 (bnx2), everything works, but if active slave is
> eth2 (tg3), incoming tagged frames (on vlan 103) are lost.
This patch isn't quite right - it always disables vlan stripping
unless management firmware is in use, so it's not really a correct
fix.
You said that this used to work correctly on this NIC? Does it work
without a bond, just a vlan on the tg3 device? It sounds like Michael
has a problem with vlan stripping on one of his NICs but if it works
with just a vlan or on older kernels, it's probably not the same
thing.
If it works on bnx2, it would seem to be a driver problem but it would
be good to confirm that the tag in skb->vlan_tci is not being
delievered to the networking core in this case.
^ permalink raw reply
* Re: [PATCH V7 7/8] ptp: Added a clock driver for the IXP46x.
From: Pavel Machek @ 2011-01-02 8:45 UTC (permalink / raw)
To: Richard Cochran
Cc: linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-api-u79uwXL29TY76Z2rM5mHXA, netdev-u79uwXL29TY76Z2rM5mHXA,
Alan Cox, Arnd Bergmann, Christoph Lameter, David Miller,
John Stultz, Krzysztof Halasa, Peter Zijlstra, Rodolfo Giometti,
Thomas Gleixner
In-Reply-To: <d2e0a5ac5eb51e9e4c7fcc94723b67c72c2f57c1.1292512461.git.richard.cochran-3mrvs1K0uXizZXS1Dc/lvw@public.gmane.org>
Hi!!
> +struct ixp46x_channel_ctl {
> + u32 Ch_Control; /* 0x40 Time Synchronization Channel Control */
> + u32 Ch_Event; /* 0x44 Time Synchronization Channel Event */
> + u32 TxSnapLo; /* 0x48 Transmit Snapshot Low Register */
> + u32 TxSnapHi; /* 0x4C Transmit Snapshot High Register */
CouldWeGetRidOfCamelCase?
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
^ permalink raw reply
* Re: [PATCH V7 7/8] ptp: Added a clock driver for the IXP46x.
From: Richard Cochran @ 2011-01-02 9:12 UTC (permalink / raw)
To: Pavel Machek
Cc: linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-api-u79uwXL29TY76Z2rM5mHXA, netdev-u79uwXL29TY76Z2rM5mHXA,
Alan Cox, Arnd Bergmann, Christoph Lameter, David Miller,
John Stultz, Krzysztof Halasa, Peter Zijlstra, Rodolfo Giometti,
Thomas Gleixner
In-Reply-To: <20110102084505.GA3980-+ZI9xUNit7I@public.gmane.org>
On Sun, Jan 02, 2011 at 09:45:05AM +0100, Pavel Machek wrote:
> Hi!!
>
> > +struct ixp46x_channel_ctl {
> > + u32 Ch_Control; /* 0x40 Time Synchronization Channel Control */
> > + u32 Ch_Event; /* 0x44 Time Synchronization Channel Event */
> > + u32 TxSnapLo; /* 0x48 Transmit Snapshot Low Register */
> > + u32 TxSnapHi; /* 0x4C Transmit Snapshot High Register */
>
> CouldWeGetRidOfCamelCase?
I agree that CamelCase is ugly and in bad taste. However, I make an
exception when the register level programmer's manual uses this style.
IMHO, it is better to use the exact same mnemonics as in the
manual. That way, when the next person comes along to make a change to
the code, with manual in hand, he will immediately know what is what.
Thanks,
Richard
^ permalink raw reply
* Re: [PATCH V7 7/8] ptp: Added a clock driver for the IXP46x.
From: Richard Cochran @ 2011-01-02 9:19 UTC (permalink / raw)
To: Pavel Machek
Cc: linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-api-u79uwXL29TY76Z2rM5mHXA, netdev-u79uwXL29TY76Z2rM5mHXA,
Alan Cox, Arnd Bergmann, Christoph Lameter, David Miller,
John Stultz, Krzysztof Halasa, Peter Zijlstra, Rodolfo Giometti,
Thomas Gleixner
In-Reply-To: <20110102084505.GA3980-+ZI9xUNit7I@public.gmane.org>
BTW, I posted an updated V8 of this patch series on December 31.
Richard
^ permalink raw reply
* Re: [PATCH V7 7/8] ptp: Added a clock driver for the IXP46x.
From: Pavel Machek @ 2011-01-02 9:20 UTC (permalink / raw)
To: Richard Cochran
Cc: linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-api-u79uwXL29TY76Z2rM5mHXA, netdev-u79uwXL29TY76Z2rM5mHXA,
Alan Cox, Arnd Bergmann, Christoph Lameter, David Miller,
John Stultz, Krzysztof Halasa, Peter Zijlstra, Rodolfo Giometti,
Thomas Gleixner
In-Reply-To: <20110102091233.GA2847-7KxsofuKt4IfAd9E5cN8NEzG7cXyKsk/@public.gmane.org>
On Sun 2011-01-02 10:12:33, Richard Cochran wrote:
> On Sun, Jan 02, 2011 at 09:45:05AM +0100, Pavel Machek wrote:
> > Hi!!
> >
> > > +struct ixp46x_channel_ctl {
> > > + u32 Ch_Control; /* 0x40 Time Synchronization Channel Control */
> > > + u32 Ch_Event; /* 0x44 Time Synchronization Channel Event */
> > > + u32 TxSnapLo; /* 0x48 Transmit Snapshot Low Register */
> > > + u32 TxSnapHi; /* 0x4C Transmit Snapshot High Register */
> >
> > CouldWeGetRidOfCamelCase?
>
> I agree that CamelCase is ugly and in bad taste. However, I make an
> exception when the register level programmer's manual uses this style.
>
> IMHO, it is better to use the exact same mnemonics as in the
> manual. That way, when the next person comes along to make a change to
Given the comments -- does manual really use camelCase crap?
And... the identifiers actually combine camelCase with _. Better fix
it.
Pavel
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
^ permalink raw reply
* Re: [*v3 PATCH 14/22] IPVS: netns awareness to ip_vs_sync
From: Hans Schillstrom @ 2011-01-02 9:47 UTC (permalink / raw)
To: Simon Horman
Cc: ja, daniel.lezcano, wensong, lvs-devel, netdev, netfilter-devel,
Hans Schillstrom
In-Reply-To: <20101231004433.GA14681@verge.net.au>
On Friday, December 31, 2010 01:44:33 Simon Horman wrote:
> On Thu, Dec 30, 2010 at 11:50:58AM +0100, hans@schillstrom.com wrote:
> > From: Hans Schillstrom <hans.schillstrom@ericsson.com>
> >
> > All global variables moved to struct ipvs,
> > most external changes fixed (i.e. init_net removed)
> > in sync_buf create + 4 replaced by sizeof(struct..)
>
> [ snip ]
>
> > diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c
> > index 0454a11..5d6e250 100644
> > --- a/net/netfilter/ipvs/ip_vs_core.c
> > +++ b/net/netfilter/ipvs/ip_vs_core.c
> > @@ -1477,6 +1477,8 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
> > struct ip_vs_proto_data *pd;
> > struct ip_vs_conn *cp;
> > int ret, restart, pkts;
> > + struct net *net;
>
> The line above adds a duplicate declaration of net as
> just before the diff-context starts there is
>
> struct net *net = NULL;
>
I'll remove that
^ permalink raw reply
* Re: [PATCH] new UDPCP Communication Protocol
From: Stefani Seibold @ 2011-01-02 11:17 UTC (permalink / raw)
To: Eric Dumazet; +Cc: linux-kernel, akpm, davem, netdev, shemminger
In-Reply-To: <1293920601.2535.64.camel@edumazet-laptop>
Am Samstag, den 01.01.2011, 23:23 +0100 schrieb Eric Dumazet:
> Le samedi 01 janvier 2011 à 22:44 +0100, stefani@seibold.net a écrit :
> > From: Stefani Seibold <stefani@seibold.net>
> >
> > Changelog:
> > 31.12.2010 first proposal
> > 01.01.2011 code cleanup and fixes suggest by Eric Dumazet
> >
> > UDPCP is a communication protocol specified by the Open Base Station
> > Architecture Initiative Special Interest Group (OBSAI SIG). The
> > protocol is based on UDP and is designed to meet the needs of "Mobile
> > Communcation Base Station" internal communications. It is widely used by
> > the major networks infrastructure supplier.
> >
> > The UDPCP communication service supports the following features:
> >
> > -Connectionless communication for serial mode data transfer
> > -Acknowledged and unacknowledged transfer modes
> > -Retransmissions Algorithm
> > -Checksum Algorithm using Adler32
> > -Fragmentation of long messages (disassembly/reassembly) to match to the MTU
> > during transport:
> > -Broadcasting and multicasting messages to multiple peers in unacknowledged
> > transfer mode
> >
> > UDPCP supports application level messages up to 64 KBytes (limited by 16-bit
> > packet data length field). Messages that are longer than the MTU will be
> > fragmented to the MTU.
> >
> > UDPCP provides a reliable transport service that will perform message
> > retransmissions in case transport failures occur.
> >
> > The code is also a nice example how to implement a UDP based protocol as
> > a kernel socket modules.
> >
> > Due the nature of UDPCP which has no sliding windows support, the latency has a
> > huge impact. The perfomance increase by implementing as a kernel module is
> > about the factor 10, because there are no context switches and data packets or
> > ACKs will be handled in the interrupt service.
> >
> > There are no side effects to the network subsystems so i ask for merge it
> > into linux-next. Hope you like it.
> >
> > The patch is against 2.6.37-rc8
> >
>
> Hi Stefani
>
> 1) Please base your next trys on net-next-2.6 : This is the reference
> for stuff like that. It probably does not matter a lot, but still...
>
Okay.
>
> 2) I still see some _irq() variants of spinlock(). Its not necessary in
> network stack at the level you are working (process context, and
> softirqs)
>
> Please only use _bh() variants, it's enough.
>
Will be fixed...
> 3) I see UDPLITE references in your code. Are you sure UDPCP protocol
> can really work on top of UDPLITE ? I think not, so please remove dead
> code.
>
I have not tested it yet. But i think i should work, the code is mostly
stolen from the udp.c file. Since the stack not depend on the len field
of the UDP header, it should not matter if UPD or UDP-Lite is used.
> 4) udpcp_release_sock() seems expensive to me. Why not testing
> usk->timeout before releasing sock lock, and save a lock/unlock pair ?
>
> static inline void udpcp_release_sock(struct sock *sk)
> {
> struct udpcp_sock *usk = udpcp_sk(sk);
>
> if (usk->timeout)
> udpcp_handle_timeout(sk);
> release_sock(sk);
> }
>
No... What is when a timer occurs between exiting udpcp_handle_timeout()
and release_sock(). This timer will not longer handled.
What about this:?
static inline void udpcp_release_sock(struct sock *sk)
{
while (usk->timeout) {
udpcp_handle_timeout(sk);
release_sock(sk);
check_timeout(sk);
}
> 5) In udpcp_timeout(), if you find socket is locked by user, you set
> timeout to one and rearm a timer to udpcp_timer(sk, jiffies + 1);
>
> Why is it needed, since user process is going to handle the timeout
> indication from udpcp_release_sock() ?
>
It is only for paranoid reasons, maybe there is a bug in my wonderful
code and than the stack will stop to work.... If it is okay, i didn't
want to remove it, my cuts feels a little bit nervous.
^ permalink raw reply
* Re: [PATCH] new UDPCP Communication Protocol
From: Eric Dumazet @ 2011-01-02 11:33 UTC (permalink / raw)
To: Stefani Seibold; +Cc: linux-kernel, akpm, davem, netdev, shemminger
In-Reply-To: <1293967023.29525.16.camel@wall-e>
Le dimanche 02 janvier 2011 à 12:17 +0100, Stefani Seibold a écrit :
> Am Samstag, den 01.01.2011, 23:23 +0100 schrieb Eric Dumazet:
> > 3) I see UDPLITE references in your code. Are you sure UDPCP protocol
> > can really work on top of UDPLITE ? I think not, so please remove dead
> > code.
> >
>
> I have not tested it yet. But i think i should work, the code is mostly
> stolen from the udp.c file. Since the stack not depend on the len field
> of the UDP header, it should not matter if UPD or UDP-Lite is used.
>
Hmm, but isnt your :
udpcp_prot = udp_prot; /* in udpcp_init() */
implies that all UDPCP sockets you are going to provide to user are on
top of UDP, not UDPLITE ?
Also, could you provide a link to UDPCP protocol specs ?
Thanks
^ permalink raw reply
* Re: [PATCH] new UDPCP Communication Protocol
From: Stefani Seibold @ 2011-01-02 11:57 UTC (permalink / raw)
To: Eric Dumazet; +Cc: linux-kernel, akpm, davem, netdev, shemminger
In-Reply-To: <1293967996.2535.76.camel@edumazet-laptop>
Am Sonntag, den 02.01.2011, 12:33 +0100 schrieb Eric Dumazet:
> Le dimanche 02 janvier 2011 à 12:17 +0100, Stefani Seibold a écrit :
> > Am Samstag, den 01.01.2011, 23:23 +0100 schrieb Eric Dumazet:
>
> > > 3) I see UDPLITE references in your code. Are you sure UDPCP protocol
> > > can really work on top of UDPLITE ? I think not, so please remove dead
> > > code.
> > >
> >
> > I have not tested it yet. But i think i should work, the code is mostly
> > stolen from the udp.c file. Since the stack not depend on the len field
> > of the UDP header, it should not matter if UPD or UDP-Lite is used.
> >
>
> Hmm, but isnt your :
>
> udpcp_prot = udp_prot; /* in udpcp_init() */
>
> implies that all UDPCP sockets you are going to provide to user are on
> top of UDP, not UDPLITE ?
>
Okay, i kick to udplite support away!
> Also, could you provide a link to UDPCP protocol specs ?
>
http://read.pudn.com/downloads76/doc/project/283718/OBSAI/OBSAI/RP1_V2.0.PDF
> Thanks
>
>
^ permalink raw reply
* Final CFP and deadline extension: IFIP/IEEE DANMS 2011 (co-located with IM 2011)
From: danms2011 @ 2011-01-02 13:10 UTC (permalink / raw)
To: danms2011
(We apologize if you receive multiple copies of this message)
(Due to multiple requests paper submission deadline has been extended to 7 Jan 2011)
CALL FOR PAPERS
IFIP/IEEE DANMS 2011
Forth International Workshop on Distributed Autonomous Network Management Systems
Co-located with IM 2011, Dublin, Ireland
www.danms.org
Important Dates
Paper submission deadline: 7 Jan 2011 (extended)
Notification of acceptance: 30 Jan 2011
Camera-ready paper: 15 Feb 2011
Paper Submission URL
https://jems.sbc.org.br/danms2011
The DANMS 2011 workshop is part of a series of workshops dedicated to advances in network management and the application of new management principles in network design.
This year’s workshop emphasizes on “Efficient Management of Loosely Collaborative Service Networks” where network connectivity providers, content and service providers come together to provide high-level services such as Over-The-Top (OTT) services to end-users in a loosely collaborative way.
In the recent past, end-user services such as high quality video calls, web-based video content servers, HD video streaming and VoD services etc. have driven the telecom market. The appearance of these new content/service providers, which include so called Over-The-Top (OTT) service providers, has driven network usage, but also created huge network management problems for the operators. In this context, content/service providers together with network operators provide value to the subscribers in a “loosely collaborative” fashion. Service assurance in this loose collaborative environment is challenging, particularly in the presence of a network with limited or no guarantees. This nascent eco-system is driving fundamental changes in how networks are deployed and managed as well as how they interact with content and service providers. In this context, together with wider network management contributions, we expect to have technical contributions in the areas of automated service assurance and in loosely collaborative service networks.
Topics of interest include (but are not limited to) the following:
* Federated network management
* Impact of Over the Top (OTT) services in resource management
* Case studies of service assurance
* Efficient use of terminal reports in service assurance
* SLA for OTT service assurance
* Service and resource modelling approaches for management
* Business rules and organizational modelling
* Use of semantics in service deployment and quality assurance
* Extensions and refinements of NM standards
* Aspects of service management and assurance
* Automated service provisioning across multiple service providers
* Fault and performance management, diagnosis, and troubleshooting
* End-to-end QoS management for enterprise networks
* Measurements and insights from network operations
* Metrics, techniques, and experiments for evaluating network
* Convergence of fixed and mobile networks
Steering Committee
Nazim Agoulmine, University of Evry Val d'Essonne France
Raouf Boutaba, University of Waterloo Canada
Gabriel Hogan, Network Management Lab, Ericsson, Ireland
Workshop Co-Chairs
Sidath Handurukande, Network Management Lab, Ericsson Ireland
Jose Neuman de Souza, Federal University of Ceara, Brazil
Publicity Chair
Yangcheng Huang, Ericsson Ireland
Technical Program Committee
Arosha Bandara, Open University, UK
Javier Baliosian, Uni. of the Republic, Uruguay
Saleem Bhatti, Uni. of St Andrews, UK
Sami Bhiri, DERI, Ireland
Anne-Marie Bosneag, Ericsson, Ireland
Anca Chandra, IBM Research, USA
Dave Cleary, Ericsson, Ireland
Zoran Despotovic, NTT DoCoMo Euro-Labs, Germany
Dominique Dudkowski, NEC, Germany
Liam Fallon, Ericsson, Ireland
James Hong, POSTECH, Korea
Jesse Kielthy, TSSG, Ireland
Francine Krief, Bordeaux 1 University, France
Petr Kuznetsov, Deutsche Telekom Laboratories, Germany
Brian Lee, AIT, Ireland
Declan O'Sullivan, Trinity College Dublin, Ireland
Gerard Parr, University of Ulster, UK
Luis Rodrigues, INESC-ID/IST, Portugal
Florian Schreiner, Fraunhofer FOKUS, Germany
Rolf Stadler, KTH, Sweden
Martin van Steen, Vrje University, Netherlands
Filip De Turck, Ghent University-IBBT, Belgium
Stefan Wallin, Luleå University of Technology, Sweden
Martin Zach, Siemens AG Austria
Paper Submission Instructions
Authors are invited to submit papers of maximum six pages or extended papers of up to eight pages (including figures, tables, and references), in the standard two-column, 11 pt font, IEEE conference paper format. Standard IEEE Transactions templates for Microsoft Word or LaTeX formats can be found at
http://www.ieee.org/portal/pages/pubs/transactions/stylesheets.html.
All work submitted must be original, not previously published or under submission at other venues.
At least one author of accepted papers must be present at the workshop, to present the paper. Accepted papers will be published by IEEE Xplore and indexed accordingly.
Only PDF files will be accepted for the review process and all submissions must be done through JEMS at the URL below:
https://jems.sbc.org.br/danms2011
^ permalink raw reply
* [PATCH net-2.6 V2] bridge: fix br_multicast_ipv6_rcv for paged skbs
From: Tomas Winkler @ 2011-01-02 13:24 UTC (permalink / raw)
To: davem; +Cc: netdev, Tomas Winkler, Johannes Berg, Stephen Hemminger
use pskb_may_pull to access ipv6 header correctly for paged skbs
It was omitted in the bridge code leading to crash in blind
__skb_pull
this fixes bug https://bugzilla.kernel.org/show_bug.cgi?id=25202
Dec 15 14:36:40 User-PC hostapd: wlan0: STA 00:15:00:60:5d:34 IEEE 802.11: authenticated
Dec 15 14:36:40 User-PC hostapd: wlan0: STA 00:15:00:60:5d:34 IEEE 802.11: associated (aid 2)
Dec 15 14:36:40 User-PC hostapd: wlan0: STA 00:15:00:60:5d:34 RADIUS: starting accounting session 4D0608A3-00000005
Dec 15 14:36:41 User-PC kernel: [175576.120287] ------------[ cut here ]------------
Dec 15 14:36:41 User-PC kernel: [175576.120452] kernel BUG at include/linux/skbuff.h:1178!
Dec 15 14:36:41 User-PC kernel: [175576.120609] invalid opcode: 0000 [#1] SMP
Dec 15 14:36:41 User-PC kernel: [175576.120749] last sysfs file: /sys/devices/pci0000:00/0000:00:1f.2/host0/target0:0:0/0:0:0:0/block/sda/uevent
Dec 15 14:36:41 User-PC kernel: [175576.121035] Modules linked in: oprofile binfmt_misc bridge stp llc parport_pc ppdev arc4 iwlagn snd_hda_codec_realtek iwlcore i915 snd_hda_intel mac80211 joydev snd_hda_codec snd_hwdep snd_pcm snd_seq_midi drm_kms_helper snd_rawmidi drm snd_seq_midi_event snd_seq snd_timer snd_seq_device cfg80211 eeepc_wmi usbhid psmouse intel_agp i2c_algo_bit intel_gtt uvcvideo agpgart videodev sparse_keymap snd shpchp v4l1_compat lp hid video serio_raw soundcore output snd_page_alloc ahci libahci atl1c
Dec 15 14:36:41 User-PC kernel: [175576.122712]
Dec 15 14:36:41 User-PC kernel: [175576.122769] Pid: 0, comm: kworker/0:0 Tainted: G W 2.6.37-rc5-wl+ #3 1015PE/1016P
Dec 15 14:36:41 User-PC kernel: [175576.123012] EIP: 0060:[<f83edd65>] EFLAGS: 00010283 CPU: 1
Dec 15 14:36:41 User-PC kernel: [175576.123193] EIP is at br_multicast_rcv+0xc95/0xe1c [bridge]
Dec 15 14:36:41 User-PC kernel: [175576.123362] EAX: 0000001c EBX: f5626318 ECX: 00000000 EDX: 00000000
Dec 15 14:36:41 User-PC kernel: [175576.123550] ESI: ec512262 EDI: f5626180 EBP: f60b5ca0 ESP: f60b5bd8
Dec 15 14:36:41 User-PC kernel: [175576.123737] DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
Dec 15 14:36:41 User-PC kernel: [175576.123902] Process kworker/0:0 (pid: 0, ti=f60b4000 task=f60a8000 task.ti=f60b0000)
Dec 15 14:36:41 User-PC kernel: [175576.124137] Stack:
Dec 15 14:36:41 User-PC kernel: [175576.124181] ec556500 f6d06800 f60b5be8 c01087d8 ec512262 00000030 00000024 f5626180
Dec 15 14:36:41 User-PC kernel: [175576.124181] f572c200 ef463440 f5626300 3affffff f6d06dd0 e60766a4 000000c4 f6d06860
Dec 15 14:36:41 User-PC kernel: [175576.124181] ffffffff ec55652c 00000001 f6d06844 f60b5c64 c0138264 c016e451 c013e47d
Dec 15 14:36:41 User-PC kernel: [175576.124181] Call Trace:
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c01087d8>] ? sched_clock+0x8/0x10
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c0138264>] ? enqueue_entity+0x174/0x440
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c016e451>] ? sched_clock_cpu+0x131/0x190
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c013e47d>] ? select_task_rq_fair+0x2ad/0x730
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c0524fc1>] ? nf_iterate+0x71/0x90
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f83e4914>] ? br_handle_frame_finish+0x184/0x220 [bridge]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f83e4790>] ? br_handle_frame_finish+0x0/0x220 [bridge]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f83e46e9>] ? br_handle_frame+0x189/0x230 [bridge]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f83e4790>] ? br_handle_frame_finish+0x0/0x220 [bridge]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f83e4560>] ? br_handle_frame+0x0/0x230 [bridge]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c04ff026>] ? __netif_receive_skb+0x1b6/0x5b0
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c04f7a30>] ? skb_copy_bits+0x110/0x210
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c0503a7f>] ? netif_receive_skb+0x6f/0x80
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f82cb74c>] ? ieee80211_deliver_skb+0x8c/0x1a0 [mac80211]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f82cc836>] ? ieee80211_rx_handlers+0xeb6/0x1aa0 [mac80211]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c04ff1f0>] ? __netif_receive_skb+0x380/0x5b0
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c016e242>] ? sched_clock_local+0xb2/0x190
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c012b688>] ? default_spin_lock_flags+0x8/0x10
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c05d83df>] ? _raw_spin_lock_irqsave+0x2f/0x50
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f82cd621>] ? ieee80211_prepare_and_rx_handle+0x201/0xa90 [mac80211]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f82ce154>] ? ieee80211_rx+0x2a4/0x830 [mac80211]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f815a8d6>] ? iwl_update_stats+0xa6/0x2a0 [iwlcore]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f8499212>] ? iwlagn_rx_reply_rx+0x292/0x3b0 [iwlagn]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c05d83df>] ? _raw_spin_lock_irqsave+0x2f/0x50
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f8483697>] ? iwl_rx_handle+0xe7/0x350 [iwlagn]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<f8486ab7>] ? iwl_irq_tasklet+0xf7/0x5c0 [iwlagn]
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c01aece1>] ? __rcu_process_callbacks+0x201/0x2d0
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c0150d05>] ? tasklet_action+0xc5/0x100
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c0150a07>] ? __do_softirq+0x97/0x1d0
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c05d910c>] ? nmi_stack_correct+0x2f/0x34
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c0150970>] ? __do_softirq+0x0/0x1d0
Dec 15 14:36:41 User-PC kernel: [175576.124181] <IRQ>
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c01508f5>] ? irq_exit+0x65/0x70
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c05df062>] ? do_IRQ+0x52/0xc0
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c01036b0>] ? common_interrupt+0x30/0x38
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c03a1fc2>] ? intel_idle+0xc2/0x160
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c04daebb>] ? cpuidle_idle_call+0x6b/0x100
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c0101dea>] ? cpu_idle+0x8a/0xf0
Dec 15 14:36:41 User-PC kernel: [175576.124181] [<c05d2702>] ? start_secondary+0x1e8/0x1ee
Cc: David Miller <davem@davemloft.net>
Cc: Johannes Berg <johannes@sipsolutions.net>
Cc: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
---
V2: Implement David Miller's suggestion
net/bridge/br_multicast.c | 26 ++++++++++++++++++++------
1 files changed, 20 insertions(+), 6 deletions(-)
diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
index f19e347..5d1449f 100644
--- a/net/bridge/br_multicast.c
+++ b/net/bridge/br_multicast.c
@@ -1469,15 +1469,16 @@ static int br_multicast_ipv6_rcv(struct net_bridge *br,
if (!skb2)
return -ENOMEM;
+
+ err = -EINVAL;
+ if (!pskb_may_pull(skb2, offset + sizeof(struct icmp6hdr)))
+ goto out_nopush;
+
len -= offset - skb_network_offset(skb2);
__skb_pull(skb2, offset);
skb_reset_transport_header(skb2);
- err = -EINVAL;
- if (!pskb_may_pull(skb2, sizeof(*icmp6h)))
- goto out;
-
icmp6h = icmp6_hdr(skb2);
switch (icmp6h->icmp6_type) {
@@ -1516,7 +1517,13 @@ static int br_multicast_ipv6_rcv(struct net_bridge *br,
switch (icmp6h->icmp6_type) {
case ICMPV6_MGM_REPORT:
{
- struct mld_msg *mld = (struct mld_msg *)icmp6h;
+ struct mld_msg *mld;
+ if (skb->len < sizeof(*mld) ||
+ !pskb_may_pull(skb2, sizeof(*mld))) {
+ err = -EINVAL;
+ goto out;
+ }
+ mld = (struct mld_msg *)icmp6h;
BR_INPUT_SKB_CB(skb2)->mrouters_only = 1;
err = br_ip6_multicast_add_group(br, port, &mld->mld_mca);
break;
@@ -1529,13 +1536,20 @@ static int br_multicast_ipv6_rcv(struct net_bridge *br,
break;
case ICMPV6_MGM_REDUCTION:
{
- struct mld_msg *mld = (struct mld_msg *)icmp6h;
+ struct mld_msg *mld;
+ if (skb->len < sizeof(*mld) ||
+ !pskb_may_pull(skb2, sizeof(*mld))) {
+ err = -EINVAL;
+ goto out;
+ }
+ mld = (struct mld_msg *)icmp6h;
br_ip6_multicast_leave_group(br, port, &mld->mld_mca);
}
}
out:
__skb_push(skb2, offset);
+out_nopush:
if (skb2 != skb)
kfree_skb(skb2);
return err;
--
1.7.3.4
---------------------------------------------------------------------
Intel Israel (74) Limited
This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
^ permalink raw reply related
* [PATCH net-next-2.6] bonding: remove meaningless /sys/module/bonding/parameters entries.
From: Nicolas de Pesloüan @ 2011-01-02 14:35 UTC (permalink / raw)
To: bonding-devel, netdev, fubar, davem; +Cc: nicolas.2p.debian
Only two bonding parameters are exposed in /sys/module/bonding:
num_grat_arp
num_unsol_na
Those values are not module global, but per device.
The per device values are available in /sys/class/net/<device>/bonding.
The values exposed in /sys/module/bonding are those given at module load time
and only used as default values when creating a device. They are read-only and
cannot change in any way.
As such, they are mostly meaningless.
---
drivers/net/bonding/bond_main.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index b1025b8..b7fd966 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -116,9 +116,9 @@ module_param(max_bonds, int, 0);
MODULE_PARM_DESC(max_bonds, "Max number of bonded devices");
module_param(tx_queues, int, 0);
MODULE_PARM_DESC(tx_queues, "Max number of transmit queues (default = 16)");
-module_param(num_grat_arp, int, 0644);
+module_param(num_grat_arp, int, 0);
MODULE_PARM_DESC(num_grat_arp, "Number of gratuitous ARP packets to send on failover event");
-module_param(num_unsol_na, int, 0644);
+module_param(num_unsol_na, int, 0);
MODULE_PARM_DESC(num_unsol_na, "Number of unsolicited IPv6 Neighbor Advertisements packets to send on failover event");
module_param(miimon, int, 0);
MODULE_PARM_DESC(miimon, "Link check interval in milliseconds");
--
1.7.2.3
^ permalink raw reply related
* Re: [PATCH] net: eepro testing positive EBUSY return by request_irq()?
From: roel kluin @ 2011-01-02 14:52 UTC (permalink / raw)
To: Ben Hutchings; +Cc: davem, netdev, Andrew Morton, LKML
In-Reply-To: <1293809262.2870.45.camel@localhost>
>> - if (request_irq (*irqp, NULL, IRQF_SHARED, "bogus", dev) != EBUSY) {
>> + if (request_irq (*irqp, NULL, IRQF_SHARED, "bogus", dev) != -EBUSY) {
> This condition is completely bogus - request_irq() with a NULL handler
> now returns -EINVAL before even checking whether the IRQ is in use. The
> code should be fixed along the lines of what I did for 3c503 in commit
> b0cf4dfb7cd21556efd9a6a67edcba0840b4d98d.
>
> The e2100 and hp net drivers have the same bug.
>
> Ben.
I've made an attempt to fix the mentioned issues, but I'm not sure it is complete.
The commit made some other changes as well which didn't seem appropriate here to me.
And I didn't compile or run test this.
As a side note, in the function changed by the commit you mentioned, el2_open(), it
appears that in current git, boolean `seen' cannot ever become true, can it? So it
loops forever until the request_irq() returns an error other than -EBUSY. Not sure
how it should be fixed though.
------------->8-------------------8<----------------
Handle errors returned by request_irq() appropriately
Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
---
drivers/net/e2100.c | 14 +++++++++-----
drivers/net/eepro.c | 26 +++++++++++++++-----------
drivers/net/hp.c | 25 ++++++++++++++-----------
3 files changed, 38 insertions(+), 27 deletions(-)
diff --git a/drivers/net/e2100.c b/drivers/net/e2100.c
index 06e72fb..115fa16 100644
--- a/drivers/net/e2100.c
+++ b/drivers/net/e2100.c
@@ -217,11 +217,15 @@ static int __init e21_probe1(struct net_device *dev, int ioaddr)
if (dev->irq < 2) {
int irqlist[] = {15, 11, 10, 12, 5, 9, 3, 4};
- for (i = 0; i < ARRAY_SIZE(irqlist); i++)
- if (request_irq (irqlist[i], NULL, 0, "bogus", NULL) != -EBUSY) {
- dev->irq = irqlist[i];
- break;
- }
+ for (i = 0; i < ARRAY_SIZE(irqlist); i++) {
+ retval = request_irq (irqlist[i], NULL, 0, "bogus", NULL);
+ if (retval != -EBUSY)
+ continue;
+ if (retval < 0)
+ goto out;
+ dev->irq = irqlist[i];
+ break;
+ }
if (i >= ARRAY_SIZE(irqlist)) {
printk(" unable to get IRQ %d.\n", dev->irq);
retval = -EAGAIN;
diff --git a/drivers/net/eepro.c b/drivers/net/eepro.c
index 7c82631..24ff522 100644
--- a/drivers/net/eepro.c
+++ b/drivers/net/eepro.c
@@ -897,6 +897,7 @@ static int eepro_grab_irq(struct net_device *dev)
{
int irqlist[] = { 3, 4, 5, 7, 9, 10, 11, 12, 0 };
int *irqp = irqlist, temp_reg, ioaddr = dev->base_addr;
+ int retval;
eepro_sw2bank1(ioaddr); /* be CAREFUL, BANK 1 now */
@@ -919,21 +920,24 @@ static int eepro_grab_irq(struct net_device *dev)
outb((temp_reg & 0xf8) | irqrmap[*irqp], ioaddr + INT_NO_REG);
eepro_sw2bank0(ioaddr); /* Switch back to Bank 0 */
+ retval = request_irq (*irqp, NULL, IRQF_SHARED, "bogus", dev);
+ if (retval == -EBUSY)
+ continue;
+ if (retval < 0)
+ return retval;
- if (request_irq (*irqp, NULL, IRQF_SHARED, "bogus", dev) != EBUSY) {
- unsigned long irq_mask;
- /* Twinkle the interrupt, and check if it's seen */
- irq_mask = probe_irq_on();
+ unsigned long irq_mask;
+ /* Twinkle the interrupt, and check if it's seen */
+ irq_mask = probe_irq_on();
- eepro_diag(ioaddr); /* RESET the 82595 */
- mdelay(20);
+ eepro_diag(ioaddr); /* RESET the 82595 */
+ mdelay(20);
- if (*irqp == probe_irq_off(irq_mask)) /* It's a good IRQ line */
- break;
+ if (*irqp == probe_irq_off(irq_mask)) /* It's a good IRQ line */
+ break;
- /* clear all interrupts */
- eepro_clear_int(ioaddr);
- }
+ /* clear all interrupts */
+ eepro_clear_int(ioaddr);
} while (*++irqp);
eepro_sw2bank1(ioaddr); /* Switch back to Bank 1 */
diff --git a/drivers/net/hp.c b/drivers/net/hp.c
index d15d2f2..aeb68a0 100644
--- a/drivers/net/hp.c
+++ b/drivers/net/hp.c
@@ -167,17 +167,20 @@ static int __init hp_probe1(struct net_device *dev, int ioaddr)
int *irqp = wordmode ? irq_16list : irq_8list;
do {
int irq = *irqp;
- if (request_irq (irq, NULL, 0, "bogus", NULL) != -EBUSY) {
- unsigned long cookie = probe_irq_on();
- /* Twinkle the interrupt, and check if it's seen. */
- outb_p(irqmap[irq] | HP_RUN, ioaddr + HP_CONFIGURE);
- outb_p( 0x00 | HP_RUN, ioaddr + HP_CONFIGURE);
- if (irq == probe_irq_off(cookie) /* It's a good IRQ line! */
- && request_irq (irq, eip_interrupt, 0, DRV_NAME, dev) == 0) {
- printk(" selecting IRQ %d.\n", irq);
- dev->irq = *irqp;
- break;
- }
+ retval = request_irq (irq, NULL, 0, "bogus", NULL);
+ if (retval == -EBUSY)
+ continue;
+ if (retval < 0)
+ goto out;
+ unsigned long cookie = probe_irq_on();
+ /* Twinkle the interrupt, and check if it's seen. */
+ outb_p(irqmap[irq] | HP_RUN, ioaddr + HP_CONFIGURE);
+ outb_p( 0x00 | HP_RUN, ioaddr + HP_CONFIGURE);
+ if (irq == probe_irq_off(cookie) /* It's a good IRQ line! */
+ && request_irq (irq, eip_interrupt, 0, DRV_NAME, dev) == 0) {
+ printk(" selecting IRQ %d.\n", irq);
+ dev->irq = *irqp;
+ break;
}
} while (*++irqp);
if (*irqp == 0) {
^ permalink raw reply related
* [PATCH] new UDPCP Communication Protocol
From: stefani @ 2011-01-02 15:31 UTC (permalink / raw)
To: linux-kernel, akpm, davem, netdev, eric.dumazet, shemminger; +Cc: stefani
From: Stefani Seibold <stefani@seibold.net>
Changelog:
31.12.2010 first proposal
01.01.2011 code cleanup and fixes suggest by Eric Dumazet
02.01.2011 kick away UDP-Lite support
change spin_lock_irq into spin_lock_bh
faster udpcp_release_sock
base is now linux-next
UDPCP is a communication protocol specified by the Open Base Station
Architecture Initiative Special Interest Group (OBSAI SIG). The
protocol is based on UDP and is designed to meet the needs of "Mobile
Communcation Base Station" internal communications. It is widely used by
the major networks infrastructure supplier.
The UDPCP communication service supports the following features:
-Connectionless communication for serial mode data transfer
-Acknowledged and unacknowledged transfer modes
-Retransmissions Algorithm
-Checksum Algorithm using Adler32
-Fragmentation of long messages (disassembly/reassembly) to match to the MTU
during transport:
-Broadcasting and multicasting messages to multiple peers in unacknowledged
transfer mode
UDPCP supports application level messages up to 64 KBytes (limited by 16-bit
packet data length field). Messages that are longer than the MTU will be
fragmented to the MTU.
UDPCP provides a reliable transport service that will perform message
retransmissions in case transport failures occur.
The code is also a nice example how to implement a UDP based protocol as
a kernel socket modules.
Due the nature of UDPCP which has no sliding windows support, the latency has
a huge impact. The perfomance increase by implementing as a kernel module is
about the factor 10, because there are no context switches and data packets or
ACKs will be handled in the interrupt service.
There are no side effects to the network subsystems so i ask for merge it
into linux-next. Hope you like it.
The patch is against linux next-20101231
- Stefani
Signed-off-by: Stefani Seibold <stefani@seibold.net>
---
include/linux/socket.h | 9 +-
include/net/udp.h | 1 +
include/net/udpcp.h | 47 +
net/Kconfig | 1 +
net/Makefile | 1 +
net/ipv4/ip_output.c | 2 +
net/ipv4/ip_sockglue.c | 2 +
net/udpcp/Kconfig | 34 +
net/udpcp/Makefile | 5 +
net/udpcp/udpcp.c | 2841 ++++++++++++++++++++++++++++++++++++++++++++++++
10 files changed, 2940 insertions(+), 3 deletions(-)
create mode 100644 include/net/udpcp.h
create mode 100644 net/udpcp/Kconfig
create mode 100644 net/udpcp/Makefile
create mode 100644 net/udpcp/udpcp.c
diff --git a/include/linux/socket.h b/include/linux/socket.h
index 2dccbeb..2e9157c 100644
--- a/include/linux/socket.h
+++ b/include/linux/socket.h
@@ -171,7 +171,7 @@ struct ucred {
#define AF_DECnet 12 /* Reserved for DECnet project */
#define AF_NETBEUI 13 /* Reserved for 802.2LLC project*/
#define AF_SECURITY 14 /* Security callback pseudo AF */
-#define AF_KEY 15 /* PF_KEY key management API */
+#define AF_KEY 15 /* PF_KEY key management API */
#define AF_NETLINK 16
#define AF_ROUTE AF_NETLINK /* Alias to emulate 4.4BSD */
#define AF_PACKET 17 /* Packet family */
@@ -194,7 +194,8 @@ struct ucred {
#define AF_IEEE802154 36 /* IEEE802154 sockets */
#define AF_CAIF 37 /* CAIF sockets */
#define AF_ALG 38 /* Algorithm sockets */
-#define AF_MAX 39 /* For now.. */
+#define AF_UDPCP 39 /* UDPCP sockets */
+#define AF_MAX 40 /* For now.. */
/* Protocol families, same as address families. */
#define PF_UNSPEC AF_UNSPEC
@@ -204,7 +205,7 @@ struct ucred {
#define PF_AX25 AF_AX25
#define PF_IPX AF_IPX
#define PF_APPLETALK AF_APPLETALK
-#define PF_NETROM AF_NETROM
+#define PF_NETROM AF_NETROM
#define PF_BRIDGE AF_BRIDGE
#define PF_ATMPVC AF_ATMPVC
#define PF_X25 AF_X25
@@ -236,6 +237,7 @@ struct ucred {
#define PF_IEEE802154 AF_IEEE802154
#define PF_CAIF AF_CAIF
#define PF_ALG AF_ALG
+#define PF_UDPCP AF_UDPCP
#define PF_MAX AF_MAX
/* Maximum queue length specifiable by listen. */
@@ -310,6 +312,7 @@ struct ucred {
#define SOL_IUCV 277
#define SOL_CAIF 278
#define SOL_ALG 279
+#define SOL_UDPCP 280
/* IPX options */
#define IPX_TYPE 1
diff --git a/include/net/udp.h b/include/net/udp.h
index bb967dd..82c95a7 100644
--- a/include/net/udp.h
+++ b/include/net/udp.h
@@ -47,6 +47,7 @@ struct udp_skb_cb {
} header;
__u16 cscov;
__u8 partial_cov;
+ __u8 udpcp_flag;
};
#define UDP_SKB_CB(__skb) ((struct udp_skb_cb *)((__skb)->cb))
diff --git a/include/net/udpcp.h b/include/net/udpcp.h
new file mode 100644
index 0000000..45180a5
--- /dev/null
+++ b/include/net/udpcp.h
@@ -0,0 +1,47 @@
+/* Definitions for UDPCP sockets. */
+
+#ifndef __LINUX_IF_UDPCP
+#define __LINUX_IF_UDPCP
+
+#include "linux/ioctl.h"
+
+#define UDPCP_MAX_MSGSIZE 65487
+
+#define UDPCP_MAX_WAIT_SEC 60
+
+#define UDPCP_OPT_TRANSFER_MODE 0
+#define UDPCP_OPT_CHECKSUM_MODE 1
+#define UDPCP_OPT_TX_TIMEOUT 2
+#define UDPCP_OPT_RX_TIMEOUT 3
+#define UDPCP_OPT_MAXTRY 4
+#define UDPCP_OPT_OUTSTANDING_ACKS 5
+
+#define UDPCP_NOACK 0
+#define UDPCP_ACK 1
+#define UDPCP_SINGLE_ACK 2
+#define UDPCP_NOCHECKSUM 3
+#define UDPCP_CHECKSUM 4
+
+#define UDPCP_IOC_MAGIC 251
+
+#define UDPCP_IOCTL_GET_STATISTICS \
+ _IOR(UDPCP_IOC_MAGIC, 0x01, struct udpcp_statistics *)
+#define UDPCP_IOCTL_RESET_STATISTICS \
+ _IO(UDPCP_IOC_MAGIC, 0x02)
+#define UDPCP_IOCTL_SYNC \
+ _IOR(UDPCP_IOC_MAGIC, 0x03, unsigned long)
+
+struct udpcp_statistics {
+ unsigned int txMsgs; /* Num of transmitted messages */
+ unsigned int rxMsgs; /* Num of received messages */
+ unsigned int txNodes; /* Num of receiver nodes */
+ unsigned int rxNodes; /* Num of transmitter nodes */
+ unsigned int txTimeout; /* Num of unsuccessful transmissions */
+ unsigned int rxTimeout; /* Num of partial message receptions */
+ unsigned int txRetries; /* Num of resends */
+ unsigned int rxDiscardedFrags; /* Num of discarded fragments */
+ unsigned int crcErrors; /* Num of crc errors detected */
+};
+
+#endif
+
diff --git a/net/Kconfig b/net/Kconfig
index 7284062..4b3b619 100644
--- a/net/Kconfig
+++ b/net/Kconfig
@@ -302,6 +302,7 @@ source "net/rfkill/Kconfig"
source "net/9p/Kconfig"
source "net/caif/Kconfig"
source "net/ceph/Kconfig"
+source "net/udpcp/Kconfig"
endif # if NET
diff --git a/net/Makefile b/net/Makefile
index a3330eb..388a582 100644
--- a/net/Makefile
+++ b/net/Makefile
@@ -70,3 +70,4 @@ obj-$(CONFIG_WIMAX) += wimax/
obj-$(CONFIG_DNS_RESOLVER) += dns_resolver/
obj-$(CONFIG_CEPH_LIB) += ceph/
obj-$(CONFIG_BATMAN_ADV) += batman-adv/
+obj-$(CONFIG_UDPCP) += udpcp/
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 04c7b3b..41f9276 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -1084,6 +1084,7 @@ error:
IP_INC_STATS(sock_net(sk), IPSTATS_MIB_OUTDISCARDS);
return err;
}
+EXPORT_SYMBOL(ip_append_data);
ssize_t ip_append_page(struct sock *sk, struct page *page,
int offset, size_t size, int flags)
@@ -1340,6 +1341,7 @@ error:
IP_INC_STATS(net, IPSTATS_MIB_OUTDISCARDS);
goto out;
}
+EXPORT_SYMBOL(ip_push_pending_frames);
/*
* Throw away all pending data on the socket.
diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c
index 3948c86..310369c 100644
--- a/net/ipv4/ip_sockglue.c
+++ b/net/ipv4/ip_sockglue.c
@@ -226,6 +226,7 @@ int ip_cmsg_send(struct net *net, struct msghdr *msg, struct ipcm_cookie *ipc)
}
return 0;
}
+EXPORT_SYMBOL(ip_cmsg_send);
/* Special input handler for packets caught by router alert option.
@@ -369,6 +370,7 @@ void ip_local_error(struct sock *sk, int err, __be32 daddr, __be16 port, u32 inf
if (sock_queue_err_skb(sk, skb))
kfree_skb(skb);
}
+EXPORT_SYMBOL(ip_local_error);
/*
* Handle MSG_ERRQUEUE
diff --git a/net/udpcp/Kconfig b/net/udpcp/Kconfig
new file mode 100644
index 0000000..a58c1b0
--- /dev/null
+++ b/net/udpcp/Kconfig
@@ -0,0 +1,34 @@
+#
+# UDPCP protocol
+#
+
+config UDPCP
+ tristate "UDPCP Communication Protocol"
+ depends on INET
+ ---help---
+ UDPCP is a communication protocol specified by the Open Base Station
+ Architecture Initiative Special Interest Group (OBSAI SIG). The
+ protocol is based on UDP and is designed to meet the needs of "Mobile
+ Communcation Base Station" internal communications.
+
+ The UDPCP communication service supports the following features:
+
+ -Connectionless communication for serial mode data transfer
+ -Acknowledged and unacknowledged transfer modes
+ -Retransmissions Algorithm
+ -Checksum Algorithm using Adler32
+ -Fragmentation of long messages (disassembly/reassembly) to
+ match to the MTU during transport:
+ -Broadcasting and multicasting messages to multiple peers in
+ unacknowledged transfer mode
+
+ UDPCP supports application level messages up to 64 KBytes (limited
+ by 16-bit packet data length field). Messages that are longer than the
+ MTU will be fragmented to the MTU.
+
+ UDPCP provides a reliable transport service that will perform message
+ retransmissions in case transport failures occur.
+
+ To compile this driver as a module, choose M here: the module
+ will be called udpcp.
+
diff --git a/net/udpcp/Makefile b/net/udpcp/Makefile
new file mode 100644
index 0000000..37f87c5
--- /dev/null
+++ b/net/udpcp/Makefile
@@ -0,0 +1,5 @@
+#
+# Makefile for UDPCP support code.
+#
+
+obj-$(CONFIG_UDPCP) += udpcp.o
diff --git a/net/udpcp/udpcp.c b/net/udpcp/udpcp.c
new file mode 100644
index 0000000..62eab61
--- /dev/null
+++ b/net/udpcp/udpcp.c
@@ -0,0 +1,2841 @@
+/*
+ * UDPCP communication protocol
+ *
+ * Copyright (C) 2010 Stefani Seibold <stefani@seibold.net>
+ * in order of NSN Ulm/Germany
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ */
+
+#include <net/xfrm.h>
+#include <net/protocol.h>
+#include <net/ip.h>
+#include <net/udp.h>
+#include <net/inet_common.h>
+#include <linux/zutil.h>
+#include <linux/module.h>
+#include <linux/proc_fs.h>
+#include <linux/spinlock.h>
+#include <linux/errqueue.h>
+#include <linux/atomic.h>
+
+#include <net/udpcp.h>
+
+#define VERSION "0.71"
+
+/*
+ * UDPCP Protocol default parameters
+ */
+#define UDPCP_TX_TIMEOUT 100 /* milliseconds */
+#define UDPCP_RX_TIMEOUT 1000 /* milliseconds */
+#define UDPCP_TX_MAXTRY 5
+#define UDPCP_OUTSTANDING_ACKS 1
+
+/*
+ * UDPCP Protocol definitions
+ */
+#define UDPCP_MSG_TYPE_BIT 14
+#define UDPCP_PROTOCOL_VERSION_BIT 11
+#define UDPCP_NO_ACK_BIT 10
+#define UDPCP_CHECKSUM_BIT 9
+#define UDPCP_SINGLE_ACK_BIT 8
+#define UDPCP_DUPLICATE_BIT 7
+
+#define UDPCP_MSG_TYPE_MASK (3 << UDPCP_MSG_TYPE_BIT)
+#define UDPCP_PROTOCOL_MASK (7 << UDPCP_PROTOCOL_VERSION_BIT)
+
+#define UDPCP_MSG_TYPE_DATA (1 << UDPCP_MSG_TYPE_BIT)
+#define UDPCP_MSG_TYPE_ACK (2 << UDPCP_MSG_TYPE_BIT)
+#define UDPCP_PROTOCOL_VERSION_2 (2 << UDPCP_PROTOCOL_VERSION_BIT)
+
+#define UDPCP_NO_ACK_FLAG (1 << UDPCP_NO_ACK_BIT)
+#define UDPCP_CHECKSUM_FLAG (1 << UDPCP_CHECKSUM_BIT)
+#define UDPCP_SINGLE_ACK_FLAG (1 << UDPCP_SINGLE_ACK_BIT)
+#define UDPCP_DUPLICATE_FLAG (1 << UDPCP_DUPLICATE_BIT)
+
+/*
+ * helper macros
+ */
+#define list_to_udpcpdest(d) container_of(d, struct udpcp_dest, list)
+#define list_to_udpcpsock(d) container_of(d, struct udpcp_sock, udpcplist)
+
+#define UDPCP_HDRSIZE (sizeof(struct udpcphdr)-sizeof(struct udphdr))
+
+#define RX_NODE 1
+#define TX_NODE 2
+
+/*
+ * name of the /proc entry
+ */
+#define UDPCP_PROC "driver/udpcp"
+
+/*
+ * UDPCP message header
+ */
+struct udpcphdr {
+ struct udphdr udphdr;
+ __be32 chksum;
+ __be16 msginfo;
+ u8 fragamount;
+ u8 fragnum;
+ __be16 msgid;
+ __be16 length;
+};
+
+/*
+ * UDPCP destination descriptor
+ *
+ * For each communication address an individual destination descriptor will
+ * be create.
+ *
+ * The fields has the following meanings:
+ *
+ * list: link list: part of udpcp_sock.destlist
+ * xmit: messages fragments to be transmit
+ * tx_time: timestamp of the last transmitted message fragment
+ * rx_time: timestamp ot the last received message fragment
+ * txTimeout: statistic use only: number of transmit timeout
+ * rxTimeout: statistic use only: number of receive timeout
+ * txRetries: statistic use only: number of transmit retries
+ * rxDiscardedFrags: statistic use only: number of discarded messages
+ * xmit_wait: message fragment which is waiting for an ACK
+ * xmit_last: last fragment transmitted
+ * recv_msg: first fragment of the received message
+ * recv_last: last fragment of the received message
+ * lastmsg: last messages fragment header received
+ * ipc: linux internal ipc cookie
+ * fl: flow/routing information
+ * rt: routing entry currently used for this destination
+ * addr: ipv4 destination address
+ * port: destination port number
+ * msgid: current message id for outgoing data messages
+ * use_flag: statistic use only: flag for dest using TX and/or RX
+ * insync: flag for protocol synchronization
+ * ackmode; ack mode for the current assembled message
+ * chkmode; checksum mode for the current assembled message
+ * try: current number of retries xmit_wait message
+ * acks: number of outstandig ack's
+ */
+struct udpcp_dest {
+ struct list_head list;
+ struct sk_buff_head xmit;
+ unsigned long tx_time;
+ unsigned long rx_time;
+ u32 txTimeout;
+ u32 rxTimeout;
+ u32 txRetries;
+ u32 rxDiscardedFrags;
+ struct sk_buff *xmit_wait;
+ struct sk_buff *xmit_last;
+ struct sk_buff *recv_msg;
+ struct sk_buff *recv_last;
+ struct udpcphdr lastmsg;
+ struct ipcm_cookie ipc;
+ struct flowi fl;
+ struct rtable *rt;
+ __be32 addr;
+ __be16 port;
+ u16 msgid;
+ u8 use_flag;
+ u8 insync;
+ u8 ackmode;
+ u8 chkmode;
+ u8 try;
+ u8 acks;
+};
+
+/*
+ * UDPCP socket descriptor
+ *
+ * For each opened socket individual socket descriptor will
+ * be created
+ *
+ * The fields has the following meanings:
+ *
+ * udpsock: UDP socket has to be the first member of udpcp_sock
+ * assembly: messages fragments currently assembled
+ * assembly_len: current length of the assembled message
+ * assembly_dest: current destination assembled
+ * wq: wait queue for UDPCP_IOCTL_SYNC
+ * destlist: head of destination descriptors link list
+ * udpcplist: link list: part of udpcp_list
+ * timer: timeout handler
+ * stat: statistics for this socket
+ * pending: number of pending messages fragment in the queues
+ * tx_timeout: transmit timeout in jiffies
+ * rx_timeout: receive timeout in jiffies
+ * udp_data_ready: original data_ready handler for this socket
+ * ackmode: default ack mode
+ * chkmode: default checksum mode
+ * maxtry: max. number of resends
+ * acks: max. number of outstandig ack's
+ * timeout: flag for unhandled timeout
+ */
+struct udpcp_sock {
+ struct udp_sock udpsock;
+ struct sk_buff_head assembly;
+ u32 assembly_len;
+ struct udpcp_dest *assembly_dest;
+ wait_queue_head_t wq;
+ struct list_head destlist;
+ struct list_head udpcplist;
+ struct timer_list timer;
+ struct udpcp_statistics stat;
+ u32 pending;
+ unsigned long tx_timeout;
+ unsigned long rx_timeout;
+ void (*udp_data_ready) (struct sock *sk, int bytes);
+ u8 ackmode;
+ u8 chkmode;
+ u8 maxtry;
+ u8 acks;
+ u8 timeout;
+};
+
+/* head of struct udpcp_sock.udpcplist link list */
+static struct list_head udpcp_list;
+
+/* spinlock for race free access to the static variables */
+static spinlock_t udpcp_lock;
+
+/* debug flag, set != 0 to enable debug */
+static int debug;
+
+/* overall UDPCP statistics */
+static atomic_t udpcp_txMsgs;
+static atomic_t udpcp_rxMsgs;
+static atomic_t udpcp_txNodes;
+static atomic_t udpcp_rxNodes;
+static atomic_t udpcp_txTimeout;
+static atomic_t udpcp_rxTimeout;
+static atomic_t udpcp_txRetries;
+static atomic_t udpcp_rxDiscardedFrags;
+static atomic_t udpcp_crcErrors;
+
+module_param(debug, int, 0);
+MODULE_PARM_DESC(debug, "Debug enabled or not");
+
+#ifdef CONFIG_PROC_FS
+/*
+ * Handle /proc/driver/udpcp
+ *
+ * Show the statistics information
+ */
+static int udpcp_proc(char *page, char **start, off_t off, int count, int *eof,
+ void *data)
+{
+ int len;
+
+ len = snprintf(page, count,
+ "txMsgs: %u\n"
+ "rxMsgs: %u\n"
+ "txNodes: %u\n"
+ "rxNodes: %u\n"
+ "txTimeout: %u\n"
+ "rxTimeout: %u\n"
+ "txRetries: %u\n"
+ "rxDiscaredFrags: %u\n"
+ "crcErrors: %u\n",
+ atomic_read(&udpcp_txMsgs),
+ atomic_read(&udpcp_rxMsgs),
+ atomic_read(&udpcp_txNodes),
+ atomic_read(&udpcp_rxNodes),
+ atomic_read(&udpcp_txTimeout),
+ atomic_read(&udpcp_rxTimeout),
+ atomic_read(&udpcp_txRetries),
+ atomic_read(&udpcp_rxDiscardedFrags),
+ atomic_read(&udpcp_crcErrors)
+ );
+
+ if (len <= off)
+ return 0;
+
+ len -= off;
+
+ if (len > count)
+ return count;
+
+ return len;
+}
+#endif
+
+/*
+ * Helper for the UDPCP header from a socket buffer
+ */
+static inline struct udpcphdr *udpcp_hdr(const struct sk_buff *skb)
+{
+ return (struct udpcphdr *)skb_transport_header(skb);
+}
+
+/*
+ * Helper for conversion a basic socket into a UDPCP socket
+ */
+static inline struct udpcp_sock *udpcp_sk(const struct sock *sk)
+{
+ return (struct udpcp_sock *)sk;
+}
+
+/*
+ * Dump the transport data of a socket buffer
+ */
+static inline void dump_data(struct sk_buff *skb, unsigned int max)
+{
+ unsigned int i;
+ unsigned char *data;
+ int data_len;
+
+ data = skb_transport_header(skb) + sizeof(struct udpcphdr);
+ data_len = skb_tail_pointer(skb) - data;
+
+ pr_debug(" data: ");
+
+ if (!data_len) {
+ pr_cont("<none>\n");
+ return;
+ }
+
+ if (max > data_len)
+ max = data_len;
+
+ for (i = 0; i < max; i++)
+ pr_cont("%02x ", data[i]);
+
+ if (data_len > max)
+ pr_cont("...");
+ pr_cont("\n");
+}
+
+/*
+ * Dump and decode a msginfo value
+ */
+static inline void dump_msginfo(u16 msginfo)
+{
+ pr_debug(" msginfo:0x%04x (", msginfo);
+
+ pr_cont("PCKT:");
+ switch (msginfo & UDPCP_MSG_TYPE_MASK) {
+ case UDPCP_MSG_TYPE_DATA:
+ pr_cont("DATA");
+ break;
+ case UDPCP_MSG_TYPE_ACK:
+ pr_cont("ACK");
+ break;
+ default:
+ pr_cont("UNKNOWN");
+ break;
+ }
+ pr_cont(" VER:%d",
+ (msginfo & UDPCP_PROTOCOL_MASK) >> UDPCP_PROTOCOL_VERSION_BIT);
+
+ if (msginfo & UDPCP_NO_ACK_FLAG)
+ pr_cont(" NO_ACK");
+ if (msginfo & UDPCP_CHECKSUM_FLAG)
+ pr_cont(" CHECKSUM");
+ if (msginfo & UDPCP_SINGLE_ACK_FLAG)
+ pr_cont(" SINGLE_ACK");
+ if (msginfo & UDPCP_DUPLICATE_FLAG)
+ pr_cont(" DUPLICATE");
+ pr_cont(")\n");
+}
+
+/*
+ * Dump and decode a UDPCP message fragment
+ */
+static void dump_msg(const char *action, struct sk_buff *skb, __be32 saddr,
+ __be32 daddr)
+{
+ struct udpcphdr *uh = udpcp_hdr(skb);
+
+ pr_debug("udpcp: %s (%lu)\n", action, jiffies);
+
+ pr_debug(" src:0x%08x:%d dst:0x%08x:%d fraglen:%d\n",
+ saddr, uh->udphdr.source, daddr, uh->udphdr.dest, skb->len);
+
+ pr_debug(" fragamount:%u fragnum:%u msgid:%u%s"
+ " length:%u checksum:0x%08x\n",
+ uh->fragamount, uh->fragnum, ntohs(uh->msgid),
+ (!uh->msgid) ? "(Sync)" : "", ntohs(uh->length),
+ ntohl(uh->chksum)
+ );
+
+ dump_msginfo(ntohs(uh->msginfo));
+ dump_data(skb, 16);
+}
+
+/*
+ * Create a new destination descriptor for the given IPV4 address and port
+ */
+static struct udpcp_dest *new_dest(struct sock *sk, __be32 addr, __be16 port)
+{
+ struct udpcp_dest *dest;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ dest = kzalloc(sizeof(*dest), sk->sk_allocation);
+
+ if (dest) {
+ skb_queue_head_init(&dest->xmit);
+ dest->addr = addr;
+ dest->port = port;
+ dest->ackmode = UDPCP_ACK;
+ list_add_tail(&dest->list, &usk->destlist);
+ }
+
+ return dest;
+}
+
+/*
+ * Lookup for a destination descriptor for the given IPV4 address and port
+ */
+static struct udpcp_dest *__find_dest(struct sock *sk, __be32 addr, __be16 port)
+{
+ struct udpcp_dest *dest;
+ struct list_head *p;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ list_for_each(p, &usk->destlist) {
+ dest = list_to_udpcpdest(p);
+
+ if ((dest->addr == addr) && (dest->port == port))
+ return dest;
+ }
+ return NULL;
+}
+
+/*
+ * Lookup for a destination descriptor and create a new one if no
+ * descriptor was found.
+ */
+static struct udpcp_dest *find_dest(struct sock *sk, __be32 addr, __be16 port)
+{
+ struct udpcp_dest *dest;
+
+ dest = __find_dest(sk, addr, port);
+
+ if (!dest)
+ dest = new_dest(sk, addr, port);
+
+ return dest;
+}
+
+/*
+ * Calculate udp checksum, mostly stolen from udp stack
+ */
+static void udpcp_do_csum(struct sock *sk, struct sk_buff *skb,
+ struct udpcp_dest *dest)
+{
+ struct flowi *fl = &dest->fl;
+ struct udphdr *uh = udp_hdr(skb);
+ __wsum csum = 0;
+ unsigned short len = ntohs(uh->len);
+
+ if (sk->sk_no_check == UDP_CSUM_NOXMIT) {
+ skb->ip_summed = CHECKSUM_NONE;
+ return;
+ }
+ if (skb->ip_summed == CHECKSUM_PARTIAL) {
+ /* UDP hardware csum */
+ skb->csum_start = skb_transport_header(skb) - skb->head;
+ skb->csum_offset = offsetof(struct udphdr, check);
+ uh->check =
+ ~csum_tcpudp_magic(fl->fl4_src, fl->fl4_dst, len,
+ sk->sk_protocol, 0);
+ return;
+ }
+ csum = csum_partial(uh, sizeof(struct udpcphdr), 0);
+ csum = csum_add(csum, skb->csum);
+
+ /* add protocol-dependent pseudo-header */
+ uh->check =
+ csum_tcpudp_magic(fl->fl4_src, fl->fl4_dst, len, sk->sk_protocol,
+ csum);
+ if (uh->check == 0)
+ uh->check = CSUM_MANGLED_0;
+}
+
+/*
+ * Fetch data from kernel space and fill in checksum if needed.
+ */
+static int ip_reply_glue_bits(void *dptr, char *to, int offset,
+ int len, int odd, struct sk_buff *skb)
+{
+ __wsum csum;
+
+ csum = csum_partial_copy_nocheck(dptr+offset, to, len, 0);
+ skb->csum = csum_block_add(skb->csum, csum, odd);
+ return 0;
+}
+
+/*
+ * Send an ack for a received data message fragment
+ *
+ * If the argument duplicate is true a ACK with UDPCP_DUPLICATE_FLAG set will
+ * be send
+ */
+static void udpcp_send_ack(struct sock *sk, struct sk_buff *skb,
+ struct udpcp_dest *dest, int duplicate)
+{
+ struct inet_sock *inet = inet_sk(sk);
+ struct udpcphdr *uh = udpcp_hdr(skb);
+ struct rtable *rt = NULL;
+ __wsum csum;
+ struct ipcm_cookie ipc;
+ struct udpcphdr rep;
+
+ memset(&rep, 0, sizeof(rep));
+
+ /* Swap the send and the receive ports. */
+ rep.udphdr.source = uh->udphdr.dest;
+ rep.udphdr.dest = uh->udphdr.source;
+ rep.udphdr.len = htons(sizeof(struct udpcphdr));
+
+ rep.msginfo = htons(UDPCP_MSG_TYPE_ACK |
+ UDPCP_NO_ACK_FLAG |
+ UDPCP_SINGLE_ACK_FLAG | UDPCP_PROTOCOL_VERSION_2);
+ if (duplicate)
+ rep.msginfo |= htons(UDPCP_DUPLICATE_FLAG);
+ else
+ memcpy(&dest->lastmsg, uh, sizeof(dest->lastmsg));
+ rep.msgid = uh->msgid;
+ rep.fragamount = uh->fragamount;
+ rep.fragnum = uh->fragnum;
+ rep.length = 0;
+ rep.chksum = 0;
+ if (ntohs(uh->msginfo) & UDPCP_CHECKSUM_FLAG) {
+ u8 *data;
+ u32 data_len;
+
+ data = (u8 *) &rep + sizeof(struct udphdr);
+ data_len = sizeof(struct udpcphdr)-sizeof(struct udphdr);
+
+ rep.msginfo |= htons(UDPCP_CHECKSUM_FLAG);
+ rep.chksum = htonl(zlib_adler32(1, data, data_len));
+ }
+
+ if (unlikely(debug)) {
+ struct sk_buff tmp;
+
+ tmp.len = ntohs(rep.udphdr.len);
+ tmp.head = tmp.transport_header = tmp.data = (void *)&rep;
+ tmp.tail = tmp.head + tmp.len;
+
+ dump_msg("ack msg", &tmp, ip_hdr(skb)->daddr,
+ ip_hdr(skb)->saddr);
+ }
+
+ csum = csum_tcpudp_nofold(ip_hdr(skb)->daddr,
+ ip_hdr(skb)->saddr,
+ sizeof(rep), sk->sk_protocol, 0);
+
+ ipc.addr = dest->addr;
+ ipc.opt = NULL;
+ ipc.tx_flags = 0;
+
+ {
+ struct flowi fl = {
+ .nl_u = { .ip4_u = {
+ .daddr = ipc.addr,
+ .saddr = ip_hdr(skb)->daddr,
+ .tos = RT_TOS(ip_hdr(skb)->tos)
+ }
+ },
+ .uli_u = { .ports = {
+ .sport = udp_hdr(skb)->dest,
+ .dport = udp_hdr(skb)->source
+ }
+ },
+ .proto = sk->sk_protocol,
+ };
+ security_skb_classify_flow(skb, &fl);
+ if (ip_route_output_key(sock_net(sk), &rt, &fl))
+ return;
+ }
+
+ inet->tos = ip_hdr(skb)->tos;
+ sk->sk_priority = skb->priority;
+ sk->sk_protocol = ip_hdr(skb)->protocol;
+ sk->sk_bound_dev_if = 0;
+ ip_append_data(sk, ip_reply_glue_bits, &rep, sizeof(rep),
+ 0, &ipc, &rt, MSG_DONTWAIT);
+ skb = skb_peek(&sk->sk_write_queue);
+ if (skb) {
+ *((__sum16 *)skb_transport_header(skb) +
+ offsetof(struct udphdr, check) / 2) =
+ csum_fold(csum_add(skb->csum, csum));
+ skb->ip_summed = CHECKSUM_NONE;
+ ip_push_pending_frames(sk);
+ }
+
+ ip_rt_put(rt);
+
+ UDP_INC_STATS_BH(sock_net(sk), UDP_MIB_OUTDATAGRAMS, 0);
+}
+
+/*
+ * Pass a UDPCP skb buffer to the ip stack and send it
+ */
+static int udpcp_send_skb(struct sock *sk, struct sk_buff *skb,
+ struct udpcp_dest *dest, struct ip_options *opt)
+{
+ int err;
+
+ skb_dst_set(skb, dst_clone(&dest->rt->dst));
+
+ err = ip_build_and_send_pkt(skb, sk, dest->fl.fl4_src,
+ dest->fl.fl4_dst, opt);
+
+ if (!err)
+ UDP_INC_STATS_USER(sock_net(sk), UDP_MIB_OUTDATAGRAMS, 0);
+ return err;
+}
+
+/*
+ * Release a routing table entry if no packed will be assembled
+ */
+static void udpcp_dst_release(struct udpcp_sock *usk, struct udpcp_dest *dest)
+{
+ if (usk->assembly_dest != dest) {
+ dst_release(&dest->rt->dst);
+ dest->rt = NULL;
+ }
+}
+
+/*
+ * Return true it the passed skb socket buffer is the last in the list
+ */
+static inline bool skb_is_eoq(const struct sk_buff_head *list,
+ const struct sk_buff *skb)
+{
+ return (skb->next == (struct sk_buff *)list);
+}
+
+/*
+ * Arm the timeout handler for the socket
+ */
+static void udpcp_timer(struct sock *sk, unsigned long timeout)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ mod_timer(&usk->timer, timeout);
+}
+
+/*
+ * Decrement the socket pending counter and wakeup a waiting UDPCP_IOCTL_SYNC
+ */
+static inline void udpcp_dec_pending(struct sock *sk)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ if (!--usk->pending) {
+ if (waitqueue_active(&usk->wq))
+ wake_up_interruptible(&usk->wq);
+ }
+}
+
+/*
+ * Returns true is the passed message fragment is the last fragment
+ */
+static inline int udpcp_is_last_frag(struct udpcphdr *uh)
+{
+ return uh->fragamount == uh->fragnum + 1;
+}
+
+/*
+ * Transmit data message fragments
+ */
+static int _udpcp_xmit(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct sk_buff *skb = NULL;
+ struct sk_buff *skbc;
+ struct udpcphdr *uh;
+ int err = 0;
+
+ if (dest->acks >= usk->acks)
+ goto out;
+
+ if (!dest->xmit_last) {
+ /*
+ * handle data message fragments without an ack
+ */
+ while ((skb = skb_peek(&dest->xmit))) {
+ uh = udpcp_hdr(skb);
+
+ if (!(ntohs(uh->msginfo) & UDPCP_NO_ACK_FLAG))
+ break;
+ if (udpcp_is_last_frag(uh)) {
+ usk->stat.txMsgs++;
+ atomic_inc(&udpcp_txMsgs);
+ }
+ skb_unlink(skb, &dest->xmit);
+ udpcp_dec_pending(sk);
+ if (unlikely(debug))
+ dump_msg("send msg", skb, dest->fl.fl4_src,
+ dest->fl.fl4_dst);
+ err = udpcp_send_skb(sk, skb, dest,
+ (struct ip_options *)skb->cb);
+ if (err) {
+ kfree_skb(skb);
+ skb = NULL;
+ break;
+ }
+ }
+ dest->xmit_wait = skb;
+ } else {
+ /*
+ * handle next data message fragment waiting for an ack
+ */
+ uh = udpcp_hdr(dest->xmit_last);
+
+ if (udpcp_is_last_frag(uh))
+ goto out;
+
+ /*
+ * get next data message fragment
+ */
+ skb = dest->xmit_last->next;
+ }
+
+ /*
+ * send all data message fragment till the first which must be acked
+ */
+ while (skb) {
+ skbc = skb_clone(skb, sk->sk_allocation);
+
+ if (!skbc)
+ break;
+
+ if (unlikely(debug))
+ dump_msg("send msg", skbc, dest->fl.fl4_src,
+ dest->fl.fl4_dst);
+ err = udpcp_send_skb(sk, skbc, dest,
+ (struct ip_options *)skb->cb);
+ if (err) {
+ kfree_skb(skbc);
+ break;
+ }
+
+ uh = udpcp_hdr(skb);
+
+ if (!(ntohs(uh->msginfo) & UDPCP_SINGLE_ACK_FLAG)
+ || udpcp_is_last_frag(uh)) {
+ dest->xmit_last = skb;
+
+ if (++dest->acks >= usk->acks || udpcp_is_last_frag(uh))
+ break;
+ }
+
+ skb = skb_is_eoq(&dest->xmit, skb) ? NULL : skb->next;
+ }
+
+out:
+ if (skb_queue_empty(&dest->xmit))
+ udpcp_dst_release(usk, dest);
+
+ return err;
+}
+
+/*
+ * Transmit data message fragments and rearm the timeout handler if necessary
+ */
+static int udpcp_xmit(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int ret;
+
+ ret = _udpcp_xmit(sk, dest);
+
+ if (dest->xmit_wait) {
+ dest->tx_time = jiffies;
+
+ if (!timer_pending(&usk->timer))
+ udpcp_timer(sk, dest->tx_time + usk->tx_timeout);
+ }
+ return ret;
+}
+
+/*
+ * Queue the assembled message fragment into the transmit queue
+ */
+static void udpcp_queue_xmit(struct sock *sk, struct udpcp_dest *dest,
+ u8 ackmode, u8 chkmode)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct udpcphdr *uh;
+ struct sk_buff *skb;
+ u8 fragamount;
+ u8 fragnum;
+ unsigned short msginfo;
+ struct flowi *fl = &dest->fl;
+
+ msginfo = UDPCP_MSG_TYPE_DATA | UDPCP_PROTOCOL_VERSION_2;
+ switch (ackmode) {
+ case UDPCP_NOACK:
+ msginfo |= UDPCP_NO_ACK_FLAG;
+ break;
+ case UDPCP_SINGLE_ACK:
+ msginfo |= UDPCP_SINGLE_ACK_FLAG;
+ break;
+ case UDPCP_ACK:
+ default:
+ break;
+ }
+ switch (chkmode) {
+ case UDPCP_NOCHECKSUM:
+ break;
+ case UDPCP_CHECKSUM:
+ default:
+ msginfo |= UDPCP_CHECKSUM_FLAG;
+ break;
+ }
+
+ fragamount = skb_queue_len(&usk->assembly);
+
+ udpcp_sk(sk)->pending += fragamount;
+
+ for (fragnum = 0; fragnum != fragamount; fragnum++) {
+ unsigned char *data;
+ int data_len;
+
+ skb = skb_dequeue(&usk->assembly);
+ uh = udpcp_hdr(skb);
+
+ /*
+ * setup a UDPCP header
+ */
+ uh->chksum = 0;
+ uh->msginfo = htons(msginfo);
+ uh->fragnum = fragnum;
+ uh->fragamount = fragamount;
+ uh->msgid = htons(dest->msgid);
+ uh->length = htons(usk->assembly_len);
+
+ data = skb_transport_header(skb) + sizeof(struct udphdr);
+ data_len = skb_tail_pointer(skb) - data;
+
+ if (chkmode == UDPCP_CHECKSUM)
+ uh->chksum = htonl(zlib_adler32(1, data, data_len));
+ /*
+ * create a UDP header
+ */
+ uh->udphdr.source = fl->fl_ip_sport;
+ uh->udphdr.dest = fl->fl_ip_dport;
+ uh->udphdr.len = htons(sizeof(struct udphdr) + data_len);
+ uh->udphdr.check = 0;
+
+ /*
+ * create UDP checksum
+ */
+ udpcp_do_csum(sk, skb, dest);
+
+ /*
+ * add to xmit queue
+ */
+ skb_queue_tail(&dest->xmit, skb);
+ }
+
+ dest->msgid++;
+ usk->assembly_len = 0;
+ usk->assembly_dest = NULL;
+}
+
+/*
+ * Remove all data message fragments of the first message from the transmit
+ * queue all fragments will be merged together
+ */
+static struct sk_buff *udpcp_dequeue_msg(struct sock *sk,
+ struct udpcp_dest *dest)
+{
+ struct sk_buff *msg;
+ struct sk_buff *skb;
+ struct sk_buff **next;
+ struct udpcphdr *uh;
+
+ msg = skb_dequeue(&dest->xmit);
+ if (!msg)
+ return NULL;
+ skb_orphan(msg);
+
+ uh = udpcp_hdr(msg);
+ if (!uh->msgid) {
+ /*
+ * sync message
+ */
+ kfree_skb(msg);
+ return NULL;
+ }
+
+ skb_pull(msg, sizeof(struct udpcphdr));
+ if (udpcp_is_last_frag(uh))
+ return msg;
+
+ next = &skb_shinfo(msg)->frag_list;
+ for (;;) {
+ skb = skb_dequeue(&dest->xmit);
+ if (!skb)
+ break;
+ skb_orphan(skb);
+ uh = udpcp_hdr(skb);
+ skb_pull(msg, sizeof(struct udpcphdr));
+ msg->len += skb->len;
+ msg->data_len += skb->len;
+ *next = skb;
+ if (udpcp_is_last_frag(uh))
+ break;
+ next = &skb->next;
+ }
+ return msg;
+}
+
+static void udpcp_flush_err(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct inet_sock *inet = inet_sk(sk);
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ if (!inet->recverr)
+ skb_queue_purge(&dest->xmit);
+ else {
+ struct sock_exterr_skb *serr;
+ struct iphdr *iph;
+ struct sk_buff *skb;
+
+ while (!skb_queue_empty(&dest->xmit)) {
+ skb = udpcp_dequeue_msg(sk, dest);
+ if (!skb)
+ continue;
+
+ if (unlikely(debug))
+ dump_msg("flush outgoing message", skb,
+ dest->fl.fl4_src, dest->fl.fl4_dst);
+
+ skb_push(skb, sizeof(struct iphdr));
+ skb_reset_network_header(skb);
+ iph = ip_hdr(skb);
+ iph->daddr = dest->rt->rt_dst;
+
+ serr = SKB_EXT_ERR(skb);
+ serr->ee.ee_errno = EPROTO;
+ serr->ee.ee_origin = SO_EE_ORIGIN_LOCAL;
+ serr->ee.ee_type = 0;
+ serr->ee.ee_code = 0;
+ serr->ee.ee_pad = 0;
+ serr->ee.ee_info = 0;
+ serr->ee.ee_data = 0;
+ serr->addr_offset = (u8 *) &iph->daddr -
+ skb_network_header(skb);
+ serr->port = dest->fl.fl_ip_dport;
+
+ skb_reset_transport_header(skb);
+ skb_pull(skb, sizeof(struct iphdr));
+
+ /*
+ * set a flag for UDPCP message
+ */
+ UDP_SKB_CB(skb)->udpcp_flag = 1;
+
+ /*
+ * pass the dequeued message to the error queue of the
+ * socket
+ */
+ skb_set_owner_r(skb, sk);
+ skb_queue_tail(&sk->sk_error_queue, skb);
+ if (!sock_flag(sk, SOCK_DEAD)) {
+ if (usk->udp_data_ready)
+ usk->udp_data_ready(sk, skb->len);
+ }
+ }
+ }
+
+ dest->xmit_wait = 0;
+ dest->xmit_last = 0;
+ dest->try = 0;
+ dest->acks = 0;
+
+ usk->pending = 0;
+ if (waitqueue_active(&usk->wq))
+ wake_up_interruptible(&usk->wq);
+}
+
+/*
+ * Purge the current incoming data message
+ */
+static void udpcp_purge_incoming(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ if (dest->recv_last) {
+ u32 fragnum = udpcp_hdr(dest->recv_last)->fragnum + 1;
+
+ dest->rxDiscardedFrags += fragnum;
+ usk->stat.rxDiscardedFrags += fragnum;
+ atomic_add(fragnum, &udpcp_rxDiscardedFrags);
+
+ dest->lastmsg.msgid = 0;
+
+ if (unlikely(debug))
+ dump_msg("purge incoming message", dest->recv_msg,
+ dest->fl.fl4_src, dest->fl.fl4_dst);
+ }
+
+ kfree_skb(dest->recv_msg);
+ dest->recv_msg = 0;
+ dest->recv_last = 0;
+}
+
+/*
+ * Resend all data message fragments to the one which is currently waiting for
+ * an ack
+ */
+static int udpcp_resend(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct sk_buff *skb;
+ struct sk_buff *skbc;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int err;
+
+ if (++dest->try >= usk->maxtry) {
+ dest->insync = 0;
+ udpcp_flush_err(sk, dest);
+ udpcp_purge_incoming(sk, dest);
+ udpcp_dst_release(usk, dest);
+ return 0;
+ }
+
+ dest->txRetries++;
+ usk->stat.txRetries++;
+ atomic_inc(&udpcp_txRetries);
+
+ if (!dest->xmit_last)
+ _udpcp_xmit(sk, dest);
+ else {
+ skb = dest->xmit_wait;
+
+ for (;;) {
+ skbc = skb_clone(skb, sk->sk_allocation);
+
+ if (skbc == NULL)
+ break;
+
+ if (unlikely(debug))
+ dump_msg("resend msg", skbc, dest->fl.fl4_src,
+ dest->fl.fl4_dst);
+ err = udpcp_send_skb(sk, skbc, dest,
+ (struct ip_options *)skb->cb);
+ if (err) {
+ kfree_skb(skbc);
+ break;
+ }
+
+ if (skb == dest->xmit_last) {
+ _udpcp_xmit(sk, dest);
+ break;
+ }
+
+ skb = skb->next;
+ }
+ }
+ dest->tx_time = jiffies;
+
+ return 1;
+}
+
+/*
+ * Handle udpcp timeout
+ */
+static void udpcp_handle_timeout(struct sock *sk)
+{
+ struct udpcp_dest *dest;
+ struct list_head *p;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int wflag = 0;
+ unsigned long t = jiffies + UDPCP_MAX_WAIT_SEC * HZ + 1;
+
+ usk->timeout = 0;
+
+ /*
+ * walk through all destinations
+ */
+ list_for_each(p, &usk->destlist) {
+ dest = list_to_udpcpdest(p);
+
+ if (dest->xmit_wait) {
+ if (time_is_before_eq_jiffies
+ (dest->tx_time + usk->tx_timeout)) {
+ /*
+ * transmit timeout expired
+ */
+ if (unlikely(debug))
+ dump_msg("send timeout",
+ dest->xmit_wait,
+ dest->fl.fl4_src,
+ dest->fl.fl4_dst);
+ if (udpcp_resend(sk, dest) == 0) {
+ dest->txTimeout++;
+ usk->stat.txTimeout++;
+ atomic_inc(&udpcp_txTimeout);
+ goto check_incoming;
+ }
+ wflag = 1;
+ }
+ if (time_before(dest->tx_time + usk->tx_timeout, t)) {
+ /*
+ * calculate new timeout timer value
+ */
+ t = dest->tx_time + usk->tx_timeout;
+ wflag = 1;
+ }
+ }
+check_incoming:
+ if (dest->recv_msg) {
+ if (time_is_before_eq_jiffies
+ (dest->rx_time + usk->rx_timeout)) {
+ /*
+ * receive timeout occurred
+ */
+ if (unlikely(debug))
+ dump_msg("receive timeout",
+ dest->recv_last,
+ dest->fl.fl4_src,
+ dest->fl.fl4_dst);
+ udpcp_purge_incoming(sk, dest);
+ dest->rxTimeout++;
+ usk->stat.rxTimeout++;
+ atomic_inc(&udpcp_rxTimeout);
+ } else
+ if (time_before(dest->rx_time + usk->rx_timeout, t)) {
+ /*
+ * calculate new timeout timer value
+ */
+ t = dest->rx_time + usk->rx_timeout;
+ wflag = 1;
+ }
+ }
+ }
+ /*
+ * restart timer if necessary
+ */
+ if (wflag)
+ udpcp_timer(sk, t);
+}
+
+/*
+ * Timeout function
+ */
+static void udpcp_timeout(unsigned long data)
+{
+ struct sock *sk = (struct sock *)data;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ bh_lock_sock(sk);
+ if (!sock_owned_by_user(sk))
+ udpcp_handle_timeout(sk);
+ else {
+ /*
+ * bad, cannot handle the timeout because the socket is in use
+ * set flag for unhandled timeout and rearm the timer
+ */
+ usk->timeout = 1;
+ udpcp_timer(sk, jiffies + 1);
+ }
+ bh_unlock_sock(sk);
+}
+
+/*
+ * Handle timeout if an the unhandled timeout flag is set
+ */
+static inline void check_timeout(struct sock *sk)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ while (usk->timeout) {
+ lock_sock(sk);
+ while (usk->timeout)
+ udpcp_handle_timeout(sk);
+ release_sock(sk);
+ }
+}
+
+/*
+ * Release the socket lock and test for unhandled timeouts
+ */
+static inline void udpcp_release_sock(struct sock *sk)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ while (usk->timeout)
+ udpcp_handle_timeout(sk);
+ release_sock(sk);
+ check_timeout(sk);
+}
+
+/*
+ * Parse sendmsg() control message
+ */
+static int udpcp_cmsg_send(struct msghdr *msg, u8 * ackmode, u8 * chkmode)
+{
+ struct cmsghdr *cmsg;
+
+ for (cmsg = CMSG_FIRSTHDR(msg); cmsg; cmsg = CMSG_NXTHDR(msg, cmsg)) {
+ if (!CMSG_OK(msg, cmsg))
+ return -EINVAL;
+ if (cmsg->cmsg_level != SOL_UDPCP)
+ continue;
+ switch (cmsg->cmsg_type) {
+ case UDPCP_NOACK:
+ case UDPCP_ACK:
+ case UDPCP_SINGLE_ACK:
+ *ackmode = cmsg->cmsg_type;
+ break;
+ case UDPCP_CHECKSUM:
+ case UDPCP_NOCHECKSUM:
+ *chkmode = cmsg->cmsg_type;
+ break;
+ default:
+ return -EINVAL;
+ }
+ }
+ return 0;
+}
+
+/*
+ * Validate a skb buffer
+ */
+static int udpcp_validate_skb(struct sk_buff *skb)
+{
+ if (skb->next) {
+ pr_err("udpcp: unexpected skb_buff->next != NULL\n");
+ BUG();
+ return 1;
+ }
+ if (skb_shinfo(skb)->frag_list) {
+ pr_err("udpcp: unexpected skb_shinfo(skb)->frag_list != NULL\n");
+ BUG();
+ return 1;
+ }
+ return 0;
+}
+
+/*
+ * Split a message into fragments and store it into the assemble queue
+ * mostly stolen from UDP stack
+ */
+static int udpcp_data(struct sock *sk, struct udpcp_dest *dest,
+ struct iovec *from, int length, unsigned int flags)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct inet_sock *inet = inet_sk(sk);
+ struct sk_buff *skb;
+ struct ipcm_cookie *ipc = &dest->ipc;
+ struct ip_options *opt = ipc->opt;
+ int hh_len;
+ int exthdrlen;
+ int mtu;
+ int copy;
+ int err;
+ int offset = 0;
+ unsigned int maxfraglen, fragheaderlen;
+ int csummode = CHECKSUM_NONE;
+ int transhdrlen = sizeof(struct udpcphdr);
+ struct rtable *rt = dest->rt;
+
+ if (opt && sizeof(skb->cb) < optlength(opt)) {
+ err = -EFAULT;
+ goto error;
+ }
+
+ usk->assembly_len += length;
+ usk->assembly_dest = dest;
+
+ if (usk->assembly_len > UDPCP_MAX_MSGSIZE) {
+ ip_local_error(sk, EMSGSIZE, rt->rt_dst, dest->fl.fl_ip_dport,
+ usk->assembly_len);
+ err = -EMSGSIZE;
+ goto error;
+ }
+
+ mtu = (inet->pmtudisc == IP_PMTUDISC_PROBE) ?
+ rt->dst.dev->mtu : dst_mtu(rt->dst.path);
+ sk->sk_sndmsg_page = NULL;
+ sk->sk_sndmsg_off = 0;
+ exthdrlen = rt->dst.header_len;
+ length += exthdrlen;
+ transhdrlen += exthdrlen;
+
+ hh_len = LL_RESERVED_SPACE(rt->dst.dev);
+
+ fragheaderlen = sizeof(struct iphdr) + (opt ? opt->optlen : 0);
+ maxfraglen = ((mtu - fragheaderlen) & ~7) + fragheaderlen;
+
+ if (rt->dst.dev->features & NETIF_F_V4_CSUM && !exthdrlen)
+ csummode = CHECKSUM_PARTIAL;
+
+ skb = skb_peek_tail(&usk->assembly);
+ if (skb) {
+ unsigned int off;
+
+ off = skb->len;
+
+ copy = mtu - skb->len;
+ if (copy > length)
+ copy = length;
+
+ if (copy > 0 &&
+ ip_generic_getfrag(
+ from, skb_put(skb, copy), 0, copy, off, skb) < 0) {
+ __skb_trim(skb, off);
+ err = -EFAULT;
+ goto error;
+ }
+ length -= copy;
+ offset += copy;
+
+ if (!length)
+ return 0;
+ }
+
+ do {
+ char *data;
+ unsigned int datalen;
+ unsigned int fraglen;
+ unsigned int alloclen;
+
+ length += transhdrlen;
+ /*
+ * If remaining data exceeds the mtu,
+ * we know we need more fragment(s).
+ */
+ datalen = length;
+ if (datalen > mtu - fragheaderlen)
+ datalen = maxfraglen - fragheaderlen;
+ fraglen = datalen + fragheaderlen;
+
+ if ((flags & MSG_MORE)
+ && !(rt->dst.dev->features & NETIF_F_SG))
+ alloclen = mtu;
+ else
+ alloclen = fraglen;
+
+ alloclen += rt->dst.trailer_len + hh_len + 15;
+
+ udpcp_release_sock(sk);
+ skb = sock_alloc_send_skb(sk, alloclen,
+ (flags & MSG_DONTWAIT), &err);
+ lock_sock(sk);
+ if (skb == NULL)
+ goto error;
+
+ if (udpcp_validate_skb(skb)) {
+ kfree_skb(skb);
+
+ goto error;
+ }
+
+ /*
+ * Fill in the control structures
+ */
+ skb->ip_summed = csummode;
+ skb->csum = 0;
+ skb_reserve(skb, hh_len);
+
+ /*
+ * Find where to start putting bytes.
+ */
+ data = skb_put(skb, fraglen);
+ skb_set_network_header(skb, exthdrlen);
+ skb->transport_header = (skb->network_header + fragheaderlen);
+ data += fragheaderlen;
+
+ copy = datalen - transhdrlen;
+
+ if (copy > 0 &&
+ ip_generic_getfrag(
+ from, data + transhdrlen, offset, copy, 0, skb) < 0) {
+ err = -EFAULT;
+ kfree_skb(skb);
+ goto error;
+ }
+
+ offset += copy;
+ length -= datalen;
+
+ if (ipc->opt)
+ memcpy(skb->cb, &ipc->opt, optlength(opt));
+
+ skb_pull(skb, fragheaderlen);
+ skb_queue_tail(&usk->assembly, skb);
+ } while (length > 0);
+
+ return 0;
+error:
+ skb_queue_purge(&usk->assembly);
+ usk->assembly_len = 0;
+
+ IP_INC_STATS(sock_net(sk), IPSTATS_MIB_OUTDISCARDS);
+ return err;
+}
+
+/*
+ * This function will be called by send(), sento() and sendmsg()
+ */
+static int udpcp_sendmsg(struct kiocb *iocb, struct sock *sk,
+ struct msghdr *msg, size_t len)
+{
+ struct inet_sock *inet = inet_sk(sk);
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct ipcm_cookie *ipc;
+ struct rtable *rt = NULL;
+ int free = 0;
+ int connected = 0;
+ __be32 daddr, faddr, saddr;
+ __be16 dport;
+ u8 tos;
+ int err = 0;
+ int corkreq = usk->udpsock.corkflag || msg->msg_flags & MSG_MORE;
+ struct udpcp_dest *dest;
+
+ if (len > UDPCP_MAX_MSGSIZE)
+ return -EMSGSIZE;
+
+ /*
+ * Check the flags.
+ */
+ if (msg->msg_flags & MSG_OOB)
+ return -EOPNOTSUPP;
+
+ /*
+ * check if socket is binded to a port
+ */
+ if (!(sk->sk_userlocks & SOCK_BINDPORT_LOCK) || !inet->inet_num)
+ return -ENOTCONN;
+
+ /*
+ * Get and verify the address.
+ */
+ if (msg->msg_name) {
+ struct sockaddr_in *usin = (struct sockaddr_in *)msg->msg_name;
+ if (msg->msg_namelen < sizeof(*usin))
+ return -EINVAL;
+ if (usin->sin_family != AF_INET) {
+ if (usin->sin_family != AF_UNSPEC)
+ return -EAFNOSUPPORT;
+ }
+
+ daddr = usin->sin_addr.s_addr;
+ dport = usin->sin_port;
+ } else {
+ if (sk->sk_state != TCP_ESTABLISHED)
+ return -EDESTADDRREQ;
+ daddr = inet->inet_daddr;
+ dport = inet->inet_dport;
+ /* Open fast path for connected socket.
+ Route will not be used, if at least one option is set.
+ */
+ connected = 1;
+ }
+
+ if (dport == 0)
+ return -EINVAL;
+
+ dest = find_dest(sk, daddr, dport);
+
+ if (!(dest->use_flag & TX_NODE)) {
+ dest->use_flag |= TX_NODE;
+ usk->stat.txNodes++;
+ atomic_inc(&udpcp_txNodes);
+ }
+
+ ipc = &dest->ipc;
+
+ if (!skb_queue_empty(&usk->assembly)) {
+ /*
+ * assembly is ongoing
+ */
+ lock_sock(sk);
+ if (likely(!skb_queue_empty(&usk->assembly))) {
+ if (usk->assembly_dest != dest) {
+ udpcp_release_sock(sk);
+ return -EUSERS;
+ }
+ ipc->opt =
+ (struct ip_options *)skb_peek(&usk->assembly)->cb;
+ goto queue_data;
+ }
+ udpcp_release_sock(sk);
+ }
+
+ ipc->addr = inet->inet_saddr;
+ ipc->oif = sk->sk_bound_dev_if;
+
+ dest->ackmode = usk->ackmode;
+ dest->chkmode = usk->chkmode;
+
+ if (msg->msg_controllen) {
+ /*
+ * handle control message
+ */
+ err = udpcp_cmsg_send(msg, &dest->ackmode, &dest->chkmode);
+ if (err)
+ return err;
+ err = ip_cmsg_send(sock_net(sk), msg, ipc);
+ if (err)
+ return err;
+ if (ipc->opt)
+ free = 1;
+ connected = 0;
+ }
+
+ if (!ipc->opt)
+ ipc->opt = inet->opt;
+
+ saddr = ipc->addr;
+ ipc->addr = faddr = daddr;
+
+ if (ipc->opt && ipc->opt->srr) {
+ if (!daddr)
+ return -EINVAL;
+ faddr = ipc->opt->faddr;
+ connected = 0;
+ }
+ tos = RT_TOS(inet->tos);
+ if (sock_flag(sk, SOCK_LOCALROUTE) ||
+ (msg->msg_flags & MSG_DONTROUTE) ||
+ (ipc->opt && ipc->opt->is_strictroute)) {
+ tos |= RTO_ONLINK;
+ connected = 0;
+ }
+
+ if (ipv4_is_multicast(daddr)) {
+ if (dest->ackmode != UDPCP_NOACK) {
+ err = EOPNOTSUPP;
+ goto out;
+ }
+ if (!ipc->oif)
+ ipc->oif = inet->mc_index;
+ if (!saddr)
+ saddr = inet->mc_addr;
+ connected = 0;
+ }
+
+ lock_sock(sk);
+ rt = dest->rt;
+ if (rt)
+ goto queue_data;
+ udpcp_release_sock(sk);
+
+ /*
+ * calculate routing
+ */
+ if (connected)
+ rt = (struct rtable *)sk_dst_check(sk, 0);
+
+ if (rt == NULL) {
+ struct flowi fl = {.oif = ipc->oif,
+ .nl_u = {.ip4_u = {.daddr = faddr,
+ .saddr = saddr,
+ .tos = tos} },
+ .proto = sk->sk_protocol,
+ .uli_u = {.ports = {.sport = inet->inet_sport,
+ .dport = dport} }
+ };
+ struct net *net = sock_net(sk);
+
+ security_sk_classify_flow(sk, &fl);
+ err = ip_route_output_flow(net, &rt, &fl, sk, 1);
+ if (err) {
+ if (err == -ENETUNREACH)
+ IP_INC_STATS_BH(net, IPSTATS_MIB_OUTNOROUTES);
+ goto out;
+ }
+
+ err = -EACCES;
+ if ((rt->rt_flags & RTCF_BROADCAST) &&
+ !sock_flag(sk, SOCK_BROADCAST))
+ goto out;
+ if (connected)
+ sk_dst_set(sk, dst_clone(&rt->dst));
+ }
+
+ if (msg->msg_flags & MSG_CONFIRM)
+ goto do_confirm;
+back_from_confirm:
+
+ saddr = rt->rt_src;
+ if (!ipc->addr)
+ daddr = ipc->addr = rt->rt_dst;
+
+ lock_sock(sk);
+
+ dest->fl.fl4_dst = daddr;
+ dest->fl.fl_ip_dport = dport;
+ dest->fl.fl4_src = saddr;
+ dest->fl.fl_ip_sport = inet->inet_sport;
+ dest->rt = rt;
+
+queue_data:
+ if (msg->msg_flags & MSG_PROBE)
+ goto release;
+
+ if (!dest->insync && skb_queue_empty(&dest->xmit)) {
+ /*
+ * if not synced, queue a SYNC message
+ */
+ err = udpcp_data(sk, dest, NULL, 0, 0);
+ if (err)
+ goto release;
+ dest->msgid = 0;
+ udpcp_queue_xmit(sk, dest, UDPCP_ACK, UDPCP_CHECKSUM);
+ }
+
+ /*
+ * split message and store it to the assembly queue
+ */
+ err = udpcp_data(sk, dest, msg->msg_iov, len,
+ corkreq ? msg->msg_flags | MSG_MORE : msg->msg_flags);
+ if (err)
+ goto release;
+
+ if (!dest->msgid)
+ dest->msgid = 1;
+
+ if (!corkreq) {
+ /*
+ * message is complete, transfer it from the assembly queue
+ * into the transmit queue
+ */
+ udpcp_queue_xmit(sk, dest, dest->ackmode, dest->chkmode);
+ /*
+ * start transmit if possible
+ */
+ err = udpcp_xmit(sk, dest);
+ }
+release:
+ udpcp_release_sock(sk);
+out:
+ if (free)
+ kfree(ipc->opt);
+
+ if (!err)
+ return len;
+ /*
+ * ENOBUFS = no kernel mem, SOCK_NOSPACE = no sndbuf space. Reporting
+ * ENOBUFS might not be good (it's not tunable per se), but otherwise
+ * we don't have a good statistic (IpOutDiscards but it can be too many
+ * things). We could add another new stat but at least for now that
+ * seems like overkill.
+ */
+ if (err == -ENOBUFS || test_bit(SOCK_NOSPACE, &sk->sk_socket->flags))
+ UDP_INC_STATS_USER(sock_net(sk), UDP_MIB_SNDBUFERRORS, 0);
+ return err;
+
+do_confirm:
+ dst_confirm(&rt->dst);
+ if (!(msg->msg_flags & MSG_PROBE) || len)
+ goto back_from_confirm;
+
+ err = 0;
+ goto out;
+}
+
+/*
+ * Sendpage() is not really implemented
+ */
+static int udpcp_sendpage(struct sock *sk, struct page *page, int offset,
+ size_t size, int flags)
+{
+ return sock_no_sendpage(sk->sk_socket, page, offset, size, flags);
+}
+
+/*
+ * Release all message fragments of the first in the transmit queue
+ */
+static void udpcp_release_xmit(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct sk_buff *skb;
+ struct udpcphdr *uh;
+
+ for (;;) {
+ skb = skb_dequeue(&dest->xmit);
+
+ uh = udpcp_hdr(skb);
+
+ if (udpcp_is_last_frag(uh) && uh->msgid) {
+ usk->stat.txMsgs++;
+ atomic_inc(&udpcp_txMsgs);
+ }
+
+ udpcp_dec_pending(sk);
+
+ kfree_skb(skb);
+ if (skb == dest->xmit_last)
+ break;
+ }
+
+ dest->xmit_wait = 0;
+ dest->xmit_last = 0;
+ dest->try = 0;
+}
+
+/*
+ * Set the sync state
+ */
+static void udpcp_sync(struct sock *sk, struct udpcp_dest *dest)
+{
+ dest->xmit_wait = 0;
+ dest->xmit_last = 0;
+ dest->try = 0;
+ dest->acks = 0;
+ dest->insync = 1;
+}
+
+/*
+ * Returns true if the first message in the transmit queue is a sync message
+ */
+static inline int udpcp_xmit_is_sync(struct udpcp_dest *dest)
+{
+ struct sk_buff *skb = skb_peek(&dest->xmit);
+
+ return skb && !udpcp_hdr(skb)->msgid;
+}
+
+static inline struct udpcphdr *udpcp_ack_scan(struct sk_buff *skb)
+{
+ struct udpcphdr *uh;
+
+ for (;;) {
+ uh = udpcp_hdr(skb);
+
+ if (!(ntohs(uh->msginfo) & UDPCP_SINGLE_ACK_FLAG)
+ || udpcp_is_last_frag(uh))
+ return uh;
+
+ skb = skb->next;
+ }
+}
+
+/*
+ * Handle an incoming ack
+ */
+static void udpcp_handle_ack(struct sock *sk, struct sk_buff *skb,
+ struct udpcp_dest *dest)
+{
+ struct udpcphdr *r_uh;
+ struct udpcphdr *q_uh;
+
+ if (!dest->acks)
+ return;
+
+ r_uh = udpcp_hdr(skb);
+
+ /*
+ * acks doesn't have a payload
+ */
+ if (r_uh->length)
+ return;
+
+ q_uh = udpcp_ack_scan(dest->xmit_wait);
+
+ /*
+ * message id, fragnum and fragamount must match the awaited message
+ * fragment
+ */
+ if (r_uh->msgid != q_uh->msgid)
+ return;
+
+ if (r_uh->fragnum != q_uh->fragnum)
+ return;
+
+ if (r_uh->fragamount != q_uh->fragamount)
+ return;
+
+ dest->acks--;
+
+ /*
+ * if last fragment release message
+ */
+ if (udpcp_is_last_frag(q_uh)) {
+ udpcp_release_xmit(sk, dest);
+
+ /*
+ * special handling for sync messages
+ */
+ if (r_uh->msgid == 0)
+ udpcp_sync(sk, dest);
+ } else
+ dest->xmit_wait = dest->xmit_wait->next;
+
+ /*
+ * try to transmit next message/fragment
+ */
+ udpcp_xmit(sk, dest);
+}
+
+/*
+ * Queue incoming message as owned by udpcp socket
+ */
+static void udpcp_set_owner_r(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct sk_buff *skb;
+
+ skb = dest->recv_msg;
+ skb_set_owner_r(skb, sk);
+
+ skb = skb_shinfo(skb)->frag_list;
+ if (!skb)
+ return;
+
+ for (;;) {
+ skb_set_owner_r(skb, sk);
+ if (udpcp_is_last_frag(udpcp_hdr(skb)))
+ break;
+ skb = skb->next;
+ }
+}
+
+/*
+ * Handle an incoming data message fragment
+ */
+static int udpcp_handle_data(struct sock *sk, struct sk_buff *skb,
+ struct udpcp_dest *dest)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct udpcphdr *uh = udpcp_hdr(skb);
+ unsigned short msginfo = ntohs(uh->msginfo);
+ unsigned short length = ntohs(uh->length);
+
+ /*
+ * special handling for sync messages
+ */
+ if (uh->msgid == 0) {
+ /*
+ * sync messages doesn't have a payload
+ */
+ if (length)
+ return 1;
+
+ /*
+ * sync messages doesn't have a ack rules
+ */
+ if (msginfo & (UDPCP_NO_ACK_FLAG | UDPCP_SINGLE_ACK_FLAG))
+ return 1;
+
+ udpcp_send_ack(sk, skb, dest,
+ memcmp(uh, &dest->lastmsg,
+ sizeof(dest->lastmsg)) ? 0 : 1);
+
+ udpcp_purge_incoming(sk, dest);
+
+ /*
+ * skip the first message in the queue if it is a sync messages
+ */
+ if (udpcp_xmit_is_sync(dest)) {
+ dest->acks--;
+ udpcp_dec_pending(sk);
+ kfree_skb(skb_dequeue(&dest->xmit));
+ }
+
+ if (!dest->insync)
+ udpcp_sync(sk, dest);
+
+ udpcp_xmit(sk, dest);
+
+ return -1;
+ }
+
+ if (!dest->insync)
+ return 1;
+
+ if (length > UDPCP_MAX_MSGSIZE)
+ return 1;
+
+ length += sizeof(struct udpcphdr);
+
+ /*
+ * if the message was still handled, send a duplicate ack
+ */
+ if (!memcmp(uh, &dest->lastmsg, sizeof(dest->lastmsg))) {
+ udpcp_send_ack(sk, skb, dest, 1);
+ return 1;
+ }
+
+ if (dest->recv_msg) {
+ /*
+ * if a fragment is already received validate the fragment
+ */
+ if ((uh->msgid != udpcp_hdr(dest->recv_msg)->msgid) ||
+ (uh->msginfo != udpcp_hdr(dest->recv_msg)->msginfo) ||
+ (uh->length != udpcp_hdr(dest->recv_msg)->length) ||
+ (uh->fragamount != udpcp_hdr(dest->recv_msg)->fragamount)
+ ) {
+ udpcp_purge_incoming(sk, dest);
+ goto newmsg;
+ }
+
+ if (uh->fragnum != udpcp_hdr(dest->recv_last)->fragnum + 1)
+ return 1;
+
+ if (dest->recv_msg->len + skb->len - sizeof(struct udpcphdr) >
+ length)
+ return 1;
+ } else {
+newmsg:
+ /*
+ * first fragment must have the number 0
+ */
+ if (uh->fragnum != 0)
+ return 1;
+
+ /*
+ * UDPCP data length cannot be smaller then the UDP data length
+ */
+ if (skb->len > length)
+ return 1;
+
+ /*
+ * id of the last received is not valid
+ */
+ if (dest->lastmsg.msgid == uh->msgid)
+ return 1;
+
+ /*
+ * check against receive buffer limit
+ */
+ if (atomic_read(&sk->sk_rmem_alloc) + length > sk->sk_rcvbuf)
+ return 1;
+ }
+
+ memset(&dest->lastmsg, 0, sizeof(dest->lastmsg));
+
+ if (!dest->recv_msg) {
+ /*
+ * store the first message fragment
+ */
+ if (skb->cloned) {
+ struct sk_buff *skbc;
+
+ skbc = skb_copy(skb, sk->sk_allocation);
+ if (skbc == NULL)
+ return 1;
+ kfree_skb(skb);
+ skb = skbc;
+ }
+ dest->recv_msg = skb;
+ } else {
+ /*
+ * store the consecutively message fragment
+ */
+ struct skb_shared_info *shinfo;
+
+ shinfo = skb_shinfo(dest->recv_msg);
+
+ if (!shinfo->frag_list)
+ shinfo->frag_list = skb;
+ else
+ dest->recv_last->next = skb;
+
+ skb_pull(skb, sizeof(struct udpcphdr));
+ dest->recv_msg->len += skb->len;
+ dest->recv_msg->data_len += skb->len;
+ }
+ dest->recv_last = skb;
+
+ msginfo = ntohs(uh->msginfo);
+
+ if (udpcp_is_last_frag(uh) || uh->fragamount == 0) {
+ /*
+ * last fragment: queue it to the socket sk_receive_queue
+ * and ack it
+ */
+
+ if (dest->recv_msg->len != length) {
+ udpcp_purge_incoming(sk, dest);
+ return 0;
+ }
+
+ if (!(msginfo & UDPCP_NO_ACK_FLAG))
+ udpcp_send_ack(sk, skb, dest, 0);
+
+ memcpy(dest->recv_msg->data + UDPCP_HDRSIZE,
+ dest->recv_msg->data, sizeof(struct udphdr));
+ skb_pull(dest->recv_msg, UDPCP_HDRSIZE);
+
+ usk->stat.rxMsgs++;
+ atomic_inc(&udpcp_rxMsgs);
+
+ /*
+ * set a flag for UDPCP message
+ */
+ UDP_SKB_CB(skb)->udpcp_flag = 1;
+
+ udpcp_set_owner_r(sk, dest);
+ skb_queue_tail(&sk->sk_receive_queue, dest->recv_msg);
+
+ /*
+ * call the original data available handler
+ */
+ if (usk->udp_data_ready)
+ usk->udp_data_ready(sk, dest->recv_msg->len);
+
+ dest->recv_msg = 0;
+ dest->recv_last = 0;
+ } else {
+ /*
+ * ack fragment if requiered
+ */
+ if (!(msginfo & UDPCP_NO_ACK_FLAG)
+ && !(msginfo & UDPCP_SINGLE_ACK_FLAG))
+ udpcp_send_ack(sk, skb, dest, 0);
+
+ /*
+ * setup timeout handler
+ */
+ dest->rx_time = jiffies;
+
+ if (!timer_pending(&usk->timer))
+ udpcp_timer(sk, dest->rx_time + usk->rx_timeout);
+ }
+
+ return 0;
+}
+
+/*
+ * Deal with received UDPCP frames - sort out what type source it is
+ * and hand of it to the udpcp_handle_packet function.
+ */
+static void udpcp_data_ready(struct sock *sk, int slen)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ struct sk_buff *skb;
+ struct udpcp_dest *dest;
+ struct udpcphdr *uh;
+ unsigned short msginfo;
+ int ret;
+
+ skb = skb_peek_tail(&sk->sk_receive_queue);
+
+ /*
+ * don't handle NULL pointer buffer and UDPCP messages
+ */
+ if (skb == NULL || UDP_SKB_CB(skb)->udpcp_flag) {
+ if (usk->udp_data_ready)
+ usk->udp_data_ready(sk, slen);
+ return;
+ }
+
+ __skb_unlink(skb, &sk->sk_receive_queue);
+ if (udpcp_validate_skb(skb)) {
+ kfree_skb(skb);
+
+ return;
+ }
+
+ skb_orphan(skb);
+
+ /*
+ * do UDP checksum
+ */
+ if (udp_lib_checksum_complete(skb)) {
+ UDP_INC_STATS_BH(sock_net(sk), UDP_MIB_INERRORS, 0);
+ return;
+ }
+
+ if (unlikely(debug))
+ dump_msg("receive", skb, ip_hdr(skb)->saddr,
+ ip_hdr(skb)->daddr);
+
+ uh = udpcp_hdr(skb);
+ msginfo = ntohs(uh->msginfo);
+
+ /*
+ * handle only UDPCP protocol version 2
+ */
+ if ((msginfo & UDPCP_PROTOCOL_MASK) != UDPCP_PROTOCOL_VERSION_2) {
+ kfree_skb(skb);
+ return;
+ }
+
+ /*
+ * handle UDPCP checksum
+ */
+ if (msginfo & UDPCP_CHECKSUM_FLAG) {
+ u8 *data;
+ u32 data_len;
+ u32 chksum;
+
+ chksum = ntohl(uh->chksum);
+ data = (u8 *) skb->data + sizeof(struct udphdr);
+ data_len = skb->len - sizeof(struct udphdr);
+
+ uh->chksum = 0;
+
+ if (chksum != zlib_adler32(1, data, data_len)) {
+ kfree_skb(skb);
+ usk->stat.crcErrors++;
+ atomic_inc(&udpcp_crcErrors);
+ return;
+ }
+ }
+
+ dest = __find_dest(sk, ip_hdr(skb)->saddr, udp_hdr(skb)->source);
+
+ if (!dest) {
+ /*
+ * new communication destination must start with an sync message
+ */
+ if (((msginfo & UDPCP_MSG_TYPE_MASK) != UDPCP_MSG_TYPE_DATA) ||
+ (uh->msgid != 0)) {
+ kfree_skb(skb);
+ return;
+ }
+
+ dest = new_dest(sk, ip_hdr(skb)->saddr, udp_hdr(skb)->source);
+
+ if (!dest) {
+ kfree_skb(skb);
+ return;
+ }
+ }
+
+ /*
+ * handle message type
+ */
+ switch (msginfo & UDPCP_MSG_TYPE_MASK) {
+ case UDPCP_MSG_TYPE_DATA:
+ if (!(dest->use_flag & RX_NODE)) {
+ dest->use_flag |= RX_NODE;
+ usk->stat.rxNodes++;
+ atomic_inc(&udpcp_rxNodes);
+ }
+
+ ret = udpcp_handle_data(sk, skb, dest);
+
+ if (ret > 0) {
+ dest->rxDiscardedFrags++;
+ usk->stat.rxDiscardedFrags++;
+ atomic_inc(&udpcp_rxDiscardedFrags);
+ }
+ break;
+ case UDPCP_MSG_TYPE_ACK:
+ udpcp_handle_ack(sk, skb, dest);
+ default:
+ ret = 1;
+ break;
+ }
+ if (ret)
+ kfree_skb(skb);
+}
+
+/*
+ * Set socket options
+ */
+static int udpcp_setsockopt(struct sock *sk, int level, int optname,
+ char __user *optval, unsigned int optlen)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int val, ret;
+
+ if (level != SOL_UDPCP) {
+ if (udp_prot.setsockopt) {
+ ret = udp_prot.setsockopt(sk, level, optname, optval,
+ optlen);
+ check_timeout(sk);
+ return ret;
+ }
+ return -ENOPROTOOPT;
+ }
+
+ if (optlen < sizeof(int))
+ return -EINVAL;
+
+ if (get_user(val, (int __user *)optval))
+ return -EFAULT;
+
+ switch (optname) {
+ case UDPCP_OPT_TRANSFER_MODE:
+ switch (val) {
+ case UDPCP_NOACK:
+ case UDPCP_ACK:
+ case UDPCP_SINGLE_ACK:
+ usk->ackmode = val;
+ break;
+ default:
+ return -EINVAL;
+ }
+ break;
+ case UDPCP_OPT_CHECKSUM_MODE:
+ switch (val) {
+ case UDPCP_NOCHECKSUM:
+ case UDPCP_CHECKSUM:
+ usk->chkmode = val;
+ break;
+ default:
+ return -EINVAL;
+ }
+ break;
+
+ case UDPCP_OPT_TX_TIMEOUT:
+ if ((val < 1) || (val > UDPCP_MAX_WAIT_SEC * 1000))
+ return -EINVAL;
+ usk->tx_timeout = msecs_to_jiffies(val);
+ break;
+
+ case UDPCP_OPT_RX_TIMEOUT:
+ if ((val < 1) || (val > UDPCP_MAX_WAIT_SEC * 1000))
+ return -EINVAL;
+ usk->rx_timeout = msecs_to_jiffies(val);
+ break;
+
+ case UDPCP_OPT_MAXTRY:
+ if ((val < 1) || (val > 10))
+ return -EINVAL;
+ usk->maxtry = val;
+ break;
+
+ case UDPCP_OPT_OUTSTANDING_ACKS:
+ if ((val < 1) || (val > 255))
+ return -EINVAL;
+ usk->acks = val;
+ break;
+
+ default:
+ return -ENOPROTOOPT;
+ }
+ return 0;
+}
+
+/*
+ * Get socket options
+ */
+static int udpcp_getsockopt(struct sock *sk, int level, int optname,
+ char __user *optval, int __user *optlen)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int val, len, ret;
+
+ if (level != SOL_UDPCP) {
+ if (udp_prot.getsockopt) {
+ ret = udp_prot.getsockopt(sk, level, optname, optval,
+ optlen);
+ check_timeout(sk);
+ return ret;
+ }
+ return -ENOPROTOOPT;
+ }
+
+ if (get_user(len, optlen))
+ return -EFAULT;
+
+ len = min_t(unsigned int, len, sizeof(int));
+
+ if (len < 0)
+ return -EINVAL;
+
+ switch (optname) {
+ case UDPCP_OPT_TRANSFER_MODE:
+ val = usk->ackmode;
+ break;
+
+ case UDPCP_OPT_CHECKSUM_MODE:
+ val = usk->chkmode;
+ break;
+
+ case UDPCP_OPT_TX_TIMEOUT:
+ val = jiffies_to_msecs(usk->tx_timeout);
+ break;
+
+ case UDPCP_OPT_MAXTRY:
+ val = usk->maxtry;
+ break;
+
+ case UDPCP_OPT_OUTSTANDING_ACKS:
+ val = usk->acks;
+ break;
+
+ default:
+ return -ENOPROTOOPT;
+ }
+
+ if (put_user(len, optlen))
+ return -EFAULT;
+ if (copy_to_user(optval, &val, len))
+ return -EFAULT;
+ return 0;
+}
+
+/*
+ * ioctl() requests applicable to the UDPCP protocol
+ */
+int udpcp_ioctl(struct sock *sk, int cmd, unsigned long arg)
+{
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int ret = 0;
+
+ switch (cmd) {
+ case UDPCP_IOCTL_GET_STATISTICS:
+ lock_sock(sk);
+ if (copy_to_user((void *)arg, &usk->stat, sizeof(usk->stat)))
+ ret = -EFAULT;
+ udpcp_release_sock(sk);
+ break;
+
+ case UDPCP_IOCTL_RESET_STATISTICS:
+ lock_sock(sk);
+ usk->stat.txMsgs = 0;
+ usk->stat.rxMsgs = 0;
+ usk->stat.txTimeout = 0;
+ usk->stat.rxTimeout = 0;
+ usk->stat.txRetries = 0;
+ usk->stat.rxDiscardedFrags = 0;
+ usk->stat.crcErrors = 0;
+ udpcp_release_sock(sk);
+ break;
+
+ case UDPCP_IOCTL_SYNC:
+ if (arg)
+ ret = wait_event_interruptible_timeout(usk->wq,
+ !usk->pending, msecs_to_jiffies(arg));
+ else
+ ret = wait_event_interruptible(usk->wq, !usk->pending);
+
+ break;
+
+ default:
+ if (udp_prot.ioctl) {
+ ret = udp_prot.ioctl(sk, cmd, arg);
+ check_timeout(sk);
+ } else
+ ret = -ENOIOCTLCMD;
+ break;
+ }
+ return ret;
+}
+
+/*
+ * This function will be called by recv(), recvfrom() and revmsg()
+ */
+int udpcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
+ size_t len, int noblock, int flags, int *addr_len)
+{
+ int ret;
+
+ ret = udp_prot.recvmsg(iocb, sk, msg, len, noblock, flags, addr_len);
+ check_timeout(sk);
+ return ret;
+}
+
+/*
+ * This function will be called by socket() and initialized the socket
+ */
+static int udpcp_sockinit(struct sock *sk)
+{
+ int ret;
+ struct udpcp_sock *usk;
+
+ sk->sk_protocol = SOL_UDP;
+ sk->sk_allocation = GFP_ATOMIC;
+ if (udp_prot.init) {
+ ret = udp_prot.init(sk);
+
+ if (ret)
+ return ret;
+ }
+
+ usk = udpcp_sk(sk);
+ usk->timer.expires = 0;
+ usk->timer.function = udpcp_timeout;
+ usk->timer.data = (long)sk;
+ init_timer(&usk->timer);
+ INIT_LIST_HEAD(&usk->destlist);
+ init_waitqueue_head(&usk->wq);
+ usk->pending = 0;
+ usk->ackmode = UDPCP_ACK;
+ usk->chkmode = UDPCP_CHECKSUM;
+ usk->maxtry = UDPCP_TX_MAXTRY;
+ usk->acks = UDPCP_OUTSTANDING_ACKS;
+ usk->tx_timeout = msecs_to_jiffies(UDPCP_TX_TIMEOUT);
+ usk->rx_timeout = msecs_to_jiffies(UDPCP_RX_TIMEOUT);
+ usk->udp_data_ready = sk->sk_data_ready;
+ sk->sk_data_ready = udpcp_data_ready;
+ usk->udpsock.pending = 0;
+ skb_queue_head_init(&usk->assembly);
+ usk->assembly_len = 0;
+ usk->assembly_dest = NULL;
+
+ spin_lock_bh(&udpcp_lock);
+ list_add_tail(&usk->udpcplist, &udpcp_list);
+ spin_unlock_bh(&udpcp_lock);
+
+#ifdef MODULE
+ try_module_get(THIS_MODULE);
+#endif
+ return 0;
+}
+
+/*
+ * This function will be called by close()
+ */
+static void udpcp_destroy(struct sock *sk)
+{
+ struct list_head *p;
+ struct list_head *n;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+
+ spin_lock_bh(&udpcp_lock);
+ list_del(&usk->udpcplist);
+ spin_unlock_bh(&udpcp_lock);
+
+ if (udp_prot.destroy)
+ udp_prot.destroy(sk);
+
+ lock_sock(sk);
+
+ del_timer_sync(&usk->timer);
+ sk->sk_data_ready = usk->udp_data_ready;
+
+ skb_queue_purge(&usk->assembly);
+
+ list_for_each_safe(p, n, &usk->destlist) {
+ struct udpcp_dest *dest;
+
+ dest = list_to_udpcpdest(p);
+
+ skb_queue_purge(&dest->xmit);
+
+ kfree_skb(dest->recv_msg);
+
+ if (dest->rt)
+ dst_release(&dest->rt->dst);
+
+ kfree(dest);
+ }
+
+ atomic_sub(usk->stat.txNodes, &udpcp_txNodes);
+ atomic_sub(usk->stat.rxNodes, &udpcp_rxNodes);
+
+ usk->pending = 0;
+
+ if (waitqueue_active(&usk->wq))
+ wake_up_interruptible(&usk->wq);
+
+ release_sock(sk);
+
+#ifdef MODULE
+ module_put(THIS_MODULE);
+#endif
+}
+
+static struct proto udpcp_prot;
+
+/*
+ * inet protocol stack descriptor
+ */
+static struct inet_protosw udpcp_protosw = {
+ .type = SOCK_DGRAM,
+ .protocol = PF_UDPCP,
+ .prot = &udpcp_prot,
+ .ops = &inet_dgram_ops,
+ .no_check = UDP_CSUM_DEFAULT,
+ .flags = 0,
+};
+
+#ifdef CONFIG_PROC_FS
+/*
+ * The following functions handles the /proc/net/udpcp entry
+ */
+struct udpcp_seq_afinfo {
+ char *name;
+ const struct file_operations seq_fops;
+ const struct seq_operations seq_ops;
+};
+
+struct udpcp_iter_state {
+ struct seq_net_private p;
+ struct sock *sk;
+ struct list_head *list;
+ int bucket;
+};
+
+static int udpcp_get_destlist(struct udpcp_sock *usk,
+ struct udpcp_iter_state *state)
+{
+ struct sock *sk = (struct sock *)usk;
+
+ if (sock_flag(sk, SOCK_DEAD))
+ return 0;
+
+ sock_hold(sk);
+ if (!list_empty(&usk->destlist)) {
+ state->sk = sk;
+ state->list = &usk->destlist;
+ return 1;
+ }
+ sock_put(sk);
+
+ return 0;
+}
+
+static inline int udpcp_next_dest(struct udpcp_iter_state *state)
+{
+ struct sock *sk = state->sk;
+ struct udpcp_sock *usk = udpcp_sk(sk);
+ int found = 0;
+
+ if (sock_flag(sk, SOCK_DEAD))
+ return 0;
+
+ lock_sock(sk);
+ if (!list_is_last(state->list, &usk->destlist)) {
+ state->list = state->list->next;
+ state->bucket++;
+ found = 1;
+ }
+ udpcp_release_sock(sk);
+ return found;
+}
+
+static void *udpcp_get_next(struct seq_file *seq)
+{
+ struct udpcp_iter_state *state = seq->private;
+ struct udpcp_sock *usk;
+ struct sock *sk;
+
+ while (state) {
+ if (udpcp_next_dest(state))
+ return state;
+
+ sk = state->sk;
+ usk = udpcp_sk(sk);
+
+ spin_lock_bh(&udpcp_lock);
+ while (!list_is_last(&usk->udpcplist, &udpcp_list)) {
+ usk = list_entry(usk->udpcplist.next, struct udpcp_sock,
+ udpcplist);
+
+ if (udpcp_get_destlist(usk, state))
+ goto found;
+ }
+ state->sk = NULL;
+ state = NULL;
+found:
+ spin_unlock_bh(&udpcp_lock);
+ sock_put(sk);
+ }
+ return state;
+}
+
+static void *udpcp_get_first(struct seq_file *seq)
+{
+ struct list_head *p;
+ struct udpcp_iter_state *state = seq->private;
+ int found = 0;
+
+ if (!state)
+ return NULL;
+
+ spin_lock_bh(&udpcp_lock);
+ list_for_each(p, &udpcp_list) {
+ found = udpcp_get_destlist(list_to_udpcpsock(p), state);
+ if (found)
+ goto found;
+ }
+found:
+ spin_unlock_bh(&udpcp_lock);
+
+ if (!found)
+ return NULL;
+ return udpcp_get_next(seq);
+}
+
+static void *udpcp_get_idx(struct seq_file *seq, loff_t pos)
+{
+ if (!udpcp_get_first(seq))
+ return NULL;
+
+ while (pos--) {
+ if (!udpcp_get_next(seq))
+ return NULL;
+ }
+ return seq->private;
+}
+
+static void *udpcp_seq_start(struct seq_file *seq, loff_t * pos)
+{
+ return *pos ? udpcp_get_idx(seq, *pos - 1) : SEQ_START_TOKEN;
+}
+
+static void *udpcp_seq_next(struct seq_file *seq, void *v, loff_t * pos)
+{
+ void *private;
+
+ if (v == SEQ_START_TOKEN)
+ private = udpcp_get_idx(seq, 0);
+ else
+ private = udpcp_get_next(seq);
+
+ ++*pos;
+ return private;
+}
+
+static void udpcp_seq_stop(struct seq_file *seq, void *v)
+{
+ struct udpcp_iter_state *state = seq->private;
+
+ if (state->sk)
+ sock_put(state->sk);
+}
+
+static int udpcp_seq_open(struct inode *inode, struct file *file)
+{
+ struct udpcp_seq_afinfo *afinfo = PDE(inode)->data;
+ int err;
+
+ err = seq_open_net(inode, file, &afinfo->seq_ops,
+ sizeof(struct udpcp_iter_state));
+ if (err < 0)
+ return err;
+
+ return err;
+}
+
+int udpcp_proc_register(struct net *net, struct udpcp_seq_afinfo *afinfo)
+{
+ struct proc_dir_entry *p;
+ int rc = 0;
+
+ p = proc_create_data(afinfo->name, S_IRUGO, net->proc_net,
+ &afinfo->seq_fops, afinfo);
+ if (!p)
+ rc = -ENOMEM;
+ return rc;
+}
+
+void udpcp_proc_unregister(struct net *net, struct udpcp_seq_afinfo *afinfo)
+{
+ proc_net_remove(net, afinfo->name);
+}
+
+static unsigned int udpcp_tx_queue_len(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct sk_buff *skb;
+ unsigned int n;
+
+ n = 0;
+ skb_queue_walk(&dest->xmit, skb)
+ n += skb->len;
+ return n;
+}
+
+static unsigned int udpcp_rx_queue_len(struct sock *sk, struct udpcp_dest *dest)
+{
+ struct sk_buff *skb;
+ unsigned int n;
+
+ n = 0;
+ skb_queue_walk(&sk->sk_receive_queue, skb) {
+ if (udp_hdr(skb)->source == dest->port
+ && ip_hdr(skb)->saddr == dest->addr)
+ n += skb->len;
+ }
+ return n;
+}
+
+static void udpcp_format_sock(struct seq_file *seq, int *len)
+{
+ struct udpcp_iter_state *state = seq->private;
+ struct sock *sk = state->sk;
+ struct inet_sock *inet = inet_sk(sk);
+ struct udpcp_dest *p = list_to_udpcpdest(state->list);
+ __be32 src = inet->inet_rcv_saddr;
+ __u16 srcp = ntohs(inet->inet_sport);
+ __be32 dest = p->addr;
+ __u16 destp = ntohs(p->port);
+
+ lock_sock(sk);
+ seq_printf(seq, "%4d: %08X:%04X %08X:%04X"
+ " %02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p %u%n",
+ state->bucket, src, srcp, dest, destp, sk->sk_state,
+ udpcp_tx_queue_len(sk, p),
+ udpcp_rx_queue_len(sk, p),
+ 0, 0L, p->txRetries, sock_i_uid(sk),
+ p->txTimeout, sock_i_ino(sk),
+ atomic_read(&sk->sk_refcnt), sk, p->rxTimeout,
+ len);
+ udpcp_release_sock(sk);
+}
+
+int udpcp_seq_show(struct seq_file *seq, void *v)
+{
+ if (v == SEQ_START_TOKEN)
+ seq_printf(seq, "%-127s\n",
+ " sl local_address rem_address st tx_queue "
+ "rx_queue tr tm->when retrnsmt uid timeout "
+ "inode ref pointer drops");
+ else {
+ int len;
+
+ udpcp_format_sock(seq, &len);
+ seq_printf(seq, "%*s\n", 127 - len, "");
+ }
+ return 0;
+}
+
+static struct udpcp_seq_afinfo udpcp_seq_afinfo = {
+ .name = "udpcp",
+ .seq_fops = {
+ .owner = THIS_MODULE,
+ .open = udpcp_seq_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = seq_release_net,
+ },
+ .seq_ops = {
+ .show = udpcp_seq_show,
+ .start = udpcp_seq_start,
+ .next = udpcp_seq_next,
+ .stop = udpcp_seq_stop,
+ },
+};
+
+static int udpcp_proc_init_net(struct net *net)
+{
+ return udpcp_proc_register(net, &udpcp_seq_afinfo);
+}
+
+static void udpcp_proc_exit_net(struct net *net)
+{
+ udpcp_proc_unregister(net, &udpcp_seq_afinfo);
+}
+
+static struct pernet_operations udpcp_net_ops = {
+ .init = udpcp_proc_init_net,
+ .exit = udpcp_proc_exit_net,
+};
+
+int __init udpcp_proc_init(void)
+{
+ return register_pernet_subsys(&udpcp_net_ops);
+}
+
+void udpcp_proc_exit(void)
+{
+ unregister_pernet_subsys(&udpcp_net_ops);
+}
+#endif /* CONFIG_PROC_FS */
+
+/*
+ * Install and init module
+ */
+static int __init udpcp_init(void)
+{
+ int ret;
+ struct proc_dir_entry *proc_entry = NULL;
+
+ spin_lock_init(&udpcp_lock);
+
+ INIT_LIST_HEAD(&udpcp_list);
+
+ /*
+ * to prevent to rewrite the whole UDP protocol,
+ * assign struct proto udp to the struct proto udpcp
+ */
+ udpcp_prot = udp_prot;
+
+ /*
+ * change the protocol name
+ */
+ strcpy(udpcp_prot.name, "UDPCP");
+
+ /*
+ * overload the following function, all other
+ * functions will use the UDP protocol functions
+ */
+ udpcp_prot.sendmsg = udpcp_sendmsg;
+ udpcp_prot.sendpage = udpcp_sendpage;
+ udpcp_prot.init = udpcp_sockinit;
+ udpcp_prot.destroy = udpcp_destroy;
+ udpcp_prot.setsockopt = udpcp_setsockopt;
+ udpcp_prot.getsockopt = udpcp_getsockopt;
+ udpcp_prot.ioctl = udpcp_ioctl;
+ udpcp_prot.recvmsg = udpcp_recvmsg;
+
+ /*
+ * fix the object size for the embedded udpcp_sock structure
+ */
+ udpcp_prot.obj_size = sizeof(struct udpcp_sock);
+
+ /*
+ * register the UDPCP protocol
+ */
+ ret = proto_register(&udpcp_prot, 1);
+ if (ret)
+ return ret;
+
+ /*
+ * register the inet socket for UDPCP
+ */
+ inet_register_protosw(&udpcp_protosw);
+
+#ifdef CONFIG_PROC_FS
+ /*
+ * register /proc/driver/udpcp entry
+ */
+ proc_entry =
+ create_proc_read_entry(UDPCP_PROC, S_IRUSR | S_IRGRP | S_IROTH,
+ NULL, udpcp_proc, NULL);
+
+ if (!proc_entry) {
+ ret = -ENOMEM;
+ goto err;
+ }
+
+ /*
+ * register /proc/net/udpcp entry
+ */
+ ret = udpcp_proc_init();
+
+ if (ret)
+ goto err;
+#endif
+ pr_info("UDPCP protocol stack version " VERSION "\n");
+ return 0;
+#ifdef CONFIG_PROC_FS
+err:
+ if (proc_entry)
+ remove_proc_entry(UDPCP_PROC, NULL);
+ proto_unregister(&udpcp_prot);
+ return ret;
+#endif
+}
+
+/*
+ * Cleanup and exit module
+ */
+static void __exit udpcp_exit(void)
+{
+#ifdef CONFIG_PROC_FS
+ udpcp_proc_exit();
+ remove_proc_entry(UDPCP_PROC, NULL);
+#endif
+ inet_unregister_protosw(&udpcp_protosw);
+ proto_unregister(&udpcp_prot);
+}
+
+module_init(udpcp_init);
+module_exit(udpcp_exit);
+
+MODULE_AUTHOR("Stefani Seibold <stefani@seibold.net>");
+MODULE_DESCRIPTION("UDPCP protocol stack v" VERSION);
+MODULE_LICENSE("GPL");
+
--
1.7.3.4
^ permalink raw reply related
* Re: [net-next-2.6 PATCH v2 2/4] dcbnl: adding DCBX feature flags get-set
From: Shmulik Ravid @ 2011-01-02 16:01 UTC (permalink / raw)
To: John Fastabend
Cc: davem@davemloft.net, Eilon Greenstein, Liu, Lucy,
netdev@vger.kernel.org
In-Reply-To: <4D1CD814.2010903@intel.com>
> One more nit ;)
>
> > +
> > + ret = nla_parse_nested(data, DCB_FEATCFG_ATTR_MAX, tb[DCB_ATTR_FEATCFG],
> > + dcbnl_featcfg_nest);
> > + if (ret) {
> > + ret = -EINVAL;
> > + goto err_out;
> > + }
>
> Why do you set EINVAL here if you use the returned error code from nla_parse_nested you get a more descriptive error. See ./lib/nlattr.c:nla_parse()/validate_nla().
>
> [...]
>
> > +static int dcbnl_setfeatcfg(struct net_device *netdev, struct nlattr **tb,
> > + u32 pid, u32 seq, u16 flags)
> > +{
> > + struct nlattr *data[DCB_FEATCFG_ATTR_MAX + 1];
> > + int ret = -EINVAL;
> > + u8 value;
> > + int i;
> > +
> > + if (!tb[DCB_ATTR_FEATCFG] || !netdev->dcbnl_ops->setfeatcfg)
> > + return ret;
> > +
> > + ret = nla_parse_nested(data, DCB_FEATCFG_ATTR_MAX, tb[DCB_ATTR_FEATCFG],
> > + dcbnl_featcfg_nest);
> > +
> > + if (ret) {
> > + ret = -EINVAL;
> > + goto err;
> > + }
>
> Same here.
>
I'll send a patch with the improved return values for the the new dcbnl
routines. While I'm at it, is it safe to fix on the same lines the
older already established dcbnl routines?
Thanks,
Shmulik
^ permalink raw reply
* Re: [PATCH 2.6.36] vlan: Avoid hwaccel vlan packets when vid not used
From: Eric Dumazet @ 2011-01-02 16:05 UTC (permalink / raw)
To: Jesse Gross
Cc: Matt Carlson, Michael Leun, Michael Chan, David Miller,
Ben Greear, linux-kernel@vger.kernel.org, netdev@vger.kernel.org
In-Reply-To: <AANLkTinq9ABi0prxe0m1Ma6rno4UnAOzuW-b5L362ij9@mail.gmail.com>
Le samedi 01 janvier 2011 à 19:27 -0500, Jesse Gross a écrit :
> On Sat, Jan 1, 2011 at 12:03 PM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> > Le mardi 14 décembre 2010 à 11:15 -0800, Matt Carlson a écrit :
> >
> >> Thanks for the comments Jesse. Below is an updated patch.
> >>
> >> Michael, I'm wondering if the difference in behavior can be explained by
> >> the presence or absence of management firmware. Can you look at the
> >> driver sign-on messages in your syslogs for ASF[]? I'm half expecting
> >> the 5752 to show "ASF[0]" and the 5714 to show "ASF[1]". If you see
> >> this, and the below patch doesn't fix the problem, let me know. I have
> >> another test I'd like you to run.
> >>
> >> ----
> >>
> >> [PATCH] tg3: Use new VLAN code
> >>
> >> This patch pivots the tg3 driver to the new VLAN infrastructure.
> >> All references to vlgrp have been removed and all VLAN code is
> >> unconditionally active.
> >>
> >> Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
>
> [...]
>
> > Hi Matt.
> >
> > Any news on this patch ?
> >
> > Without it, net-next-2.6 doesnt work for me on a vlan setup on top of
> > bonding.
> >
> > (bond0 : eth1 & eth2, eth1 being bnx2, eth2 beging tg3)
> >
> > ip link add link bond0 vlan.103 type vlan id 103
> > ip addr add 192.168.20.110/24 dev vlan.103
> > ip link set vlan.103 up
> >
> >
> > If active slave is eth1 (bnx2), everything works, but if active slave is
> > eth2 (tg3), incoming tagged frames (on vlan 103) are lost.
>
> This patch isn't quite right - it always disables vlan stripping
> unless management firmware is in use, so it's not really a correct
> fix.
>
> You said that this used to work correctly on this NIC? Does it work
> without a bond, just a vlan on the tg3 device? It sounds like Michael
> has a problem with vlan stripping on one of his NICs but if it works
> with just a vlan or on older kernels, it's probably not the same
> thing.
>
1) current linux-2.6 works OK for me (and previous versions as well, I
am using this vlan/bonding setup since 3 years or so on one of my dev
machine)
Only net-next-2.6 has the problem.
If I remove bonding of the equation, I still have the problem, and can
see the 'dropped' counter increasing while I send packets to eth2 (tg3)
$ ifconfig eth2
eth2 Link encap:Ethernet HWaddr 00:1E:0B:92:78:50
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:94 errors:0 dropped:38686 overruns:0 frame:0
TX packets:18 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:8332 (8.1 Kb) TX bytes:1392 (1.3 Kb)
Interrupt:19
$ ifconfig vlan.103
vlan.103 Link encap:Ethernet HWaddr 00:1E:0B:92:78:50
inet addr:192.168.20.110 Bcast:0.0.0.0 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:15 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:0 (0.0 b) TX bytes:846 (846.0 b)
> If it works on bnx2, it would seem to be a driver problem but it would
> be good to confirm that the tag in skb->vlan_tci is not being
> delievered to the networking core in this case.
Hmm, where do you want me to check this ?
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox