Generic support (header files, configuration, and documentation)
for the UDP-Lite protocol (RFC 3828).
This has been tested successfully on AMD, i386/i686, and SMP
architectures (compiles cleanly and works under different configurations).
With regard to the previously submitted patch set, no conceptual
changes (it works), but a lot of tidying up and re-organisation:
* the number of changes to other files has been reduced to the
bare minimum (would greatly value ideas for further reduction),
* ifdefs have been replaced by generic functions (whose implementation
depends on CONFIG_IP_UDPLITE),
* much of the declarations and concepts shared between v4 and v6 UDP-
Lite is now concentrated in the header file
I hope that this makes the code a lot more accessible. Both ipv4/udplite.c
and ipv6/udplite.c derive from the respective udp.c variants.
There is documentation enclosed, as well as a few (v6-enabled) applications
to test the kernel implementation. These should be easy to built - please
give it a test run if you have a little time.
Previous comments have been a great help in improving the code: I would greatly
value any further ideas, suggestions, and comments -- in particular ACKS & NAKS
or application.
Signed-off-by: Gerrit Renker <[email protected]>
Signed-off-by: William Stanislaus <[email protected]>
---
Documentation/networking/udplite.txt | 334 +++++++++++++++++++++++++++++++++++
include/linux/in.h | 1
include/linux/socket.h | 1
include/linux/udplite.h | 96 ++++++++++
include/net/udplite.h | 196 ++++++++++++++++++++
net/core/sock.c | 7
net/ipv4/Kconfig | 21 ++
7 files changed, 655 insertions(+), 1 deletion(-)
diff -Nurp a/net/ipv4/Kconfig b/net/ipv4/Kconfig
--- a/net/ipv4/Kconfig 2006-07-14 10:15:27.000000000 +0100
+++ b/net/ipv4/Kconfig 2006-07-14 10:17:50.000000000 +0100
@@ -581,3 +581,24 @@ config TCP_CONG_BIC
source "net/ipv4/ipvs/Kconfig"
+config IP_UDPLITE
+ bool "The UDP-Lite Protocol (EXPERIMENTAL)"
+ depends on INET && EXPERIMENTAL
+ default n
+ ---help---
+ The UDP-Lite Protocol <http://www.ietf.org/rfc/rfc3828.txt>
+
+ UDP-Lite is a Standards-Track IETF transport protocol (RFC 3828). It
+ features a variable-length checksum; which allows partially damaged
+ packets to be forwarded to media codecs, rather than being discarded
+ due to invalid (UDP) checksum values. This can have advantages for the
+ transport of multimedia (e.g. video/audio) over wireless networks.
+
+ The protocol runs on both IPv4 and IPv6. The socket API resembles that
+ of UDP. Applications must indicate their wish to utilise the partial
+ checksum coverage feature by setting a socket option; UDP-Lite will
+ otherwise run in (compatible) UDP mode.
+
+ Detailed documentation in <file:Documentation/networking/udplite.txt>.
+
+ If in doubt, say N.
diff -Nurp a/include/linux/udplite.h b/include/linux/udplite.h
--- a/include/linux/udplite.h 1970-01-01 01:00:00.000000000 +0100
+++ b/include/linux/udplite.h 2006-07-14 10:17:50.000000000 +0100
@@ -0,0 +1,96 @@
+/*
+ * Header file for UDP Lite (RFC 3828).
+ *
+ * Version: see net/ipv4/udplite.c
+ *
+ * Authors: Gerrit Renker, <[email protected]>
+ * William Stanislaus, <[email protected]>
+ *
+ * Fixes:
+ * Changes:
+ *
+ * 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.
+ */
+#ifndef _LINUX_UDPLITE_H
+#define _LINUX_UDPLITE_H
+#include <linux/types.h>
+
+/**
+ * struct udplitehdr - UDP-Lite header re-interpreting UDP (RFC 768) fields
+ *
+ * @source: source port number (as in UDP)
+ * @dest: destination port number (as in UDP)
+ * @checklen: checksum coverage length
+ * @check: checksum field (as in UDP)
+ *
+ * For the detailed semantics see RFC 3828.
+ */
+struct udplitehdr {
+ __u16 source;
+ __u16 dest;
+ __u16 checklen;
+ __u16 check;
+};
+
+/* UDP-Lite socket options */
+#define UDPLITE_SEND_CSCOV 10
+#define UDPLITE_RECV_CSCOV 11
+
+
+#ifdef __KERNEL__
+#include <linux/config.h>
+#include <net/sock.h>
+#include <linux/ip.h>
+
+/**
+ * struct udplite_sock - unreliable, connection-less UDP-Lite service
+ *
+ * @inet: has to be the first member
+ * @pending: any pending frames?
+ * @corkflag: when cork is required
+ * @encap_type: is this an encapsulation socket?
+ * @len: total length of pending frames
+ * @pcslen: partial checksum coverage length for sending socket
+ * @pcrlen: partial checksum coverage length for receiving socket
+ * @pcflag: partial checksum coverage flag
+ *
+ * NOTE: Checksum coverage length has different semantics for sending and
+ * receiving sockets.
+ *
+ */
+struct udplite_sock {
+ struct inet_sock inet;
+ int pending;
+ unsigned int corkflag;
+ __u16 encap_type;
+ /* The following members retain the information to create a
+ * UDP-Lite header when the socket is uncorked. */
+ __u16 len;
+ __u16 pcslen;
+ __u16 pcrlen;
+/* checksum coverage set indicators used by pcflag */
+#define UDPLITE_SEND_CC 0x1
+#define UDPLITE_RECV_CC 0x2
+ __u8 pcflag;
+};
+
+static inline struct udplite_sock *udplite_sk(const struct sock *sk)
+{
+ return (struct udplite_sock *)sk;
+}
+
+
+struct udplite6_sock {
+ struct udplite_sock udpl;
+ /*
+ * ipv6_pinfo has to be the last member of udplite6_sock,
+ * see inet6_sk_generic.
+ */
+ struct ipv6_pinfo inet6;
+};
+#endif /* __KERNEL__ */
+
+#endif /* _LINUX_UDPLITE_H */
diff -Nurp a/include/net/udplite.h b/include/net/udplite.h
--- a/include/net/udplite.h 1970-01-01 01:00:00.000000000 +0100
+++ b/include/net/udplite.h 2006-07-14 10:17:50.000000000 +0100
@@ -0,0 +1,196 @@
+/*
+ * Definitions for the UDP-Lite (RFC 3828) code.
+ *
+ * Version: see net/ipv4/udplite.c
+ *
+ * Authors: Gerrit Renker, <[email protected]>
+ * William Stanislaus, <[email protected]>
+ *
+ * Fixes:
+ * Changes:
+ *
+ * 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.
+ *
+ * NOTE: In UDP-Lite the checksum MUST always be computed, hence there is
+ * no UDPLITE_CSUM_DEFAULT and no UDPLITE_CSUM_NOXMIT here.
+ */
+
+#ifndef _UDPLITE_H
+#define _UDPLITE_H
+#ifdef CONFIG_IP_UDPLITE
+#include <linux/udplite.h>
+#include <net/snmp.h>
+#include <net/udp.h> /* for UDP_HTABLE_SIZE and proc structures */
+/*
+ * Global variables
+ */
+extern struct proto udplite_prot;
+extern struct hlist_head udplite_hash[UDP_HTABLE_SIZE];
+extern rwlock_t udplite_hash_lock;
+extern int udplite_port_rover;
+
+/**
+ * struct udp_lite_skb - UDP-Lite private variables
+ *
+ * @cscov: checksum coverage length
+ * @partial: flag, if set indicates partial csum coverage
+ * @header: private variables used by IPv4/v6 (thanks tcp.h!)
+ */
+struct udplite_skb_cb {
+ union {
+ struct inet_skb_parm h4;
+#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE)
+ struct inet6_skb_parm h6;
+#endif
+ } header;
+ __u16 cscov;
+ __u8 partial;
+};
+#define UDPLITE_SKB_CB(skb) ((struct udplite_skb_cb *)&((skb)->cb))
+
+
+/*
+ * Inlined functions shared between UDP-Litev4 and UDP-Litev6
+ */
+
+/* shared port space between ipv4/udplite.c and ipv6/udplite.c, cf. udp.c */
+static inline int udplite_lport_inuse(u16 num)
+{
+ struct sock *sk;
+ struct hlist_node *node;
+
+ sk_for_each(sk, node, &udplite_hash[num & (UDP_HTABLE_SIZE - 1)])
+ if (inet_sk(sk)->num == num)
+ return 1;
+ return 0;
+}
+
+/*
+ * Calculate / check variable-length UDP-Lite checksum
+ * skb->csum holds the sum of the IPv4 or IPv6 pseudo-header.
+ */
+static inline u16 __udplite_checksum_complete(struct sk_buff *skb)
+{
+ if (! UDPLITE_SKB_CB(skb)->partial)
+ return __skb_checksum_complete(skb);
+
+ return csum_fold(skb_checksum(skb, 0, UDPLITE_SKB_CB(skb)->cscov,
+ skb->csum));
+}
+
+static inline u16 udplite_checksum_complete(struct sk_buff *skb)
+{
+ return skb->ip_summed != CHECKSUM_UNNECESSARY &&
+ __udplite_checksum_complete(skb);
+}
+
+/*
+ * UDP-Lite checksum computation is all in software, hence simpler getfrag
+ */
+static inline int udplite_getfrag(void *from, char *to, int offset, int len,
+ int odd, struct sk_buff *skb)
+{
+ return memcpy_fromiovecend(to, (struct iovec *) from, offset, len);
+}
+
+
+/*
+ * net/ipv4/udplite.c
+ */
+extern void udplite4_register(void);
+extern int udplite4_mib_init(void);
+extern unsigned int udplite_poll(struct file *file, struct socket *sock,
+ poll_table * wait);
+extern int udplite_rcv(struct sk_buff *skb);
+extern void udplite_err(struct sk_buff *, u32);
+extern int udplite_disconnect(struct sock *sk, int flags);
+extern int udplite_ioctl(struct sock *sk, int cmd, unsigned long arg);
+extern int udplite_sendmsg(struct kiocb *iocb, struct sock *sk,
+ struct msghdr *msg, size_t len);
+#ifdef CONFIG_PROC_FS
+extern int udplite_proc_register(struct udp_seq_afinfo *afinfo);
+extern void udplite_proc_unregister(struct udp_seq_afinfo *afinfo);
+
+extern int udplite4_proc_init(void);
+extern void udplite4_proc_exit(void);
+#endif /* CONFIG_PROC_FS */
+
+
+/*
+ * net/ipv6/udplite.c
+ */
+#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE)
+extern struct proto udplitev6_prot;
+extern void udplitev6_init(void);
+extern void udplitev6_exit(void);
+#ifdef CONFIG_PROC_FS
+extern int udplite6_proc_init(void);
+extern void udplite6_proc_exit(void);
+#endif
+#endif /* CONFIG_IPV6 */
+
+
+/*
+ * MIB / runtime statistics for UDP-Litev4 and UDP-Litev6.
+ */
+enum {
+ UDPLITE_MIB_NUM = 0,
+ UDPLITE_MIB_INDATAGRAMS, /* total received datagramns */
+ UDPLITE_MIB_IN_PARTIALCOV, /* rcvd datagrams with partial coverage */
+ UDPLITE_MIB_NOPORTS, /* rcvd datagrams to wrong ports */
+ UDPLITE_MIB_INERRORS, /* total erroneous received datagrams */
+ UDPLITE_MIB_IN_BAD_COV, /* checksum coverage errors */
+ UDPLITE_MIB_IN_BAD_CSUM, /* checksum itself did not qualify */
+ UDPLITE_MIB_OUTDATAGRAMS, /* total sent datagrams */
+ UDPLITE_MIB_OUT_PARTIALCOV, /* sent datagrams with partial coverage */
+ __UDPLITE_MIB_MAX
+};
+
+struct udplite_mib {
+ unsigned long mibs[__UDPLITE_MIB_MAX];
+} __SNMP_MIB_ALIGN__;
+
+
+DECLARE_SNMP_STAT(struct udplite_mib, udplite_statistics);
+#define UDPLITE_INC_STATS(f) SNMP_INC_STATS(udplite_statistics, f)
+#define UDPLITE_INC_STATS_BH(f) SNMP_INC_STATS_BH(udplite_statistics, f)
+#define UDPLITE_INC_STATS_USER(f) SNMP_INC_STATS_USER(udplite_statistics, f)
+
+#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE)
+DECLARE_SNMP_STAT(struct udplite_mib, udplite_stats_in6);
+#define UDPLITE6_INC_STATS(f) SNMP_INC_STATS(udplite_stats_in6, f)
+#define UDPLITE6_INC_STATS_BH(f) SNMP_INC_STATS_BH(udplite_stats_in6, f)
+#define UDPLITE6_INC_STATS_USER(f) SNMP_INC_STATS_USER(udplite_stats_in6, f)
+#endif /* CONFIG_IPV6 */
+
+
+
+#else /* CONFIG_IP_UDPLITE is not enabled */
+
+
+/* empty v4 function definitions: */
+inline void udplite4_register(void) { }
+#define udplite4_mib_init(void) (1)
+#ifdef CONFIG_PROC_FS
+#define udplite4_proc_init(void) (0)
+inline void udplite4_proc_exit(void) { }
+#endif /* CONFIG_PROC_FS */
+
+
+/* empty v6 function definitions: */
+#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE)
+inline void udplitev6_init(void) { }
+inline void udplitev6_exit(void) { }
+#ifdef CONFIG_PROC_FS
+#define udplite6_proc_init(void) (0)
+inline void udplite6_proc_exit(void) { }
+#endif /* CONFIG_PROC_FS */
+#endif /* CONFIG_IPV6 */
+
+#endif /* CONFIG_IP_UDPLITE */
+
+
+#endif /* _UDPLITE_H */
diff -Nurp a/include/linux/socket.h b/include/linux/socket.h
--- a/include/linux/socket.h 2006-07-06 09:08:15.000000000 +0100
+++ b/include/linux/socket.h 2006-07-14 10:17:50.000000000 +0100
@@ -264,6 +264,7 @@ struct ucred {
#define SOL_IPV6 41
#define SOL_ICMPV6 58
#define SOL_SCTP 132
+#define SOL_UDPLITE 136 /* UDP-Lite (RFC 3828) */
#define SOL_RAW 255
#define SOL_IPX 256
#define SOL_AX25 257
diff -Nurp a/include/linux/in.h b/include/linux/in.h
--- a/include/linux/in.h 2006-06-19 08:45:25.000000000 +0100
+++ b/include/linux/in.h 2006-07-14 10:17:50.000000000 +0100
@@ -44,6 +44,7 @@ enum {
IPPROTO_COMP = 108, /* Compression Header protocol */
IPPROTO_SCTP = 132, /* Stream Control Transport Protocol */
+ IPPROTO_UDPLITE = 136, /* UDP-Lite Protocol (RFC 3828) */
IPPROTO_RAW = 255, /* Raw IP packets */
IPPROTO_MAX
diff -Nurp a/net/core/sock.c b/net/core/sock.c
--- a/net/core/sock.c 2006-07-06 09:08:24.000000000 +0100
+++ b/net/core/sock.c 2006-07-14 10:17:50.000000000 +0100
@@ -479,7 +479,12 @@ set_rcvbuf:
break;
case SO_NO_CHECK:
- sk->sk_no_check = valbool;
+ /* UDP-Lite (RFC 3828) mandates checksumming,
+ * hence user must not enable this option. */
+ if (sk->sk_protocol == IPPROTO_UDPLITE)
+ ret = -EOPNOTSUPP;
+ else
+ sk->sk_no_check = valbool;
break;
case SO_PRIORITY:
diff -Nurp a/Documentation/networking/udplite.txt b/Documentation/networking/udplite.txt
--- a/Documentation/networking/udplite.txt 1970-01-01 01:00:00.000000000 +0100
+++ b/Documentation/networking/udplite.txt 2006-07-14 10:17:50.000000000 +0100
@@ -0,0 +1,334 @@
+ ===========================================================================
+ The UDP-Lite protocol (RFC 3828)
+ ===========================================================================
+ last modified: Fri 14th July 2006
+
+
+ UDP-Lite is a Standards-Track IETF transport protocol whose characteristic
+ is a variable-length checksum. This has advantages for transport of multimedia
+ (video, VoIP) over wireless networks, as partly damaged packets can still be
+ fed into the codec instead of being discarded due to a failed checksum test.
+
+ This file briefly describes the existing kernel support and the socket API.
+ For in-depth information, you can consult:
+
+ o The UDP-Lite Homepage: http://www.erg.abdn.ac.uk/users/gerrit/udp-lite/
+ Fom here you can always also pull the latest patch for the stable
+ kernel tree and example application source code.
+
+ o The UDP-Lite HOWTO on
+ http://www.erg.abdn.ac.uk/users/gerrit/udp-lite/files/UDP-Lite-HOWTO.txt
+
+ o The Ethereal UDP-Lite WiKi (with capture files):
+ http://wiki.ethereal.com/Lightweight_User_Datagram_Protocol
+
+ o The Protocol Spec, RFC 3828: http://www.ietf.org/rfc/rfc3828.txt
+
+
+ I) APPLICATIONS
+
+ Several applications have been ported successfully to UDP-Lite. Ethereal
+ (now called wireshark) has UDP-Litev4/v6 support by default. The tarball on
+
+ http://www.erg.abdn.ac.uk/users/gerrit/udp-lite/files/udplite_linux.tar.gz
+
+ has source code for several v4/v6 client-server and network testing examples.
+
+ Porting applications to UDP-Lite is straightforward: only socket level and
+ IPPROTO need to be changed; senders additionally set the checksum coverage
+ length (default = header length = 8). Details are in the next section.
+ UDP-Lite is not enabled per default, set CONFIG_IP_UDPLITE=y to support it.
+
+
+ II) PROGRAMMING API
+
+ UDP-Lite provides a connectionless, unreliable datagram service and hence
+ uses the same socket type as UDP. In fact, porting from UDP to UDP-Lite is
+ dead easy: simply add `IPPROTO_UDPLITE' as the last argument of the socket(2)
+ call so that the statement looks like:
+
+ s = socket(PF_INET, SOCK_DGRAM, IPPROTO_UDPLITE);
+
+ or, respectively,
+
+ s = socket(PF_INET6, SOCK_DGRAM, IPPROTO_UDPLITE);
+
+ Since both UDP-Litev4 and UDP-Litev6 are supported, the porting process is the
+ same in both occasions. With just this change you are able to run UDP-Lite
+ services or connect to UDP-Lite servers. The kernel will assume that you are
+ not interested in using partial checksum coverage and so emulate UDP mode.
+
+ To make use of the partial checksum coverage facilities requires setting just
+ one socket option which takes an integer specifying the coverage length:
+
+ * Sender checksum coverage: UDPLITE_SEND_CSCOV
+
+ For example,
+
+ int val = 20;
+ setsockopt(s, SOL_UDPLITE, UDPLITE_SEND_CSCOV, &val, sizeof(int));
+
+ sets the checksum coverage length to 20 bytes (12b data + 8b header).
+ Of each packet only the first 20 bytes (plus the pseudo-header) will be
+ checksummed. This is useful for RTP applications which have a 12-byte
+ base header.
+
+
+ * Receiver checksum coverage: UDPLITE_RECV_CSCOV
+
+ This option is the receiver-side analogue. It is truly optional, i.e. not
+ required to enable traffic with partial checksum coverage. Its function is
+ that of a traffic filter: when enabled, it instructs the kernel to drop
+ all packets which have a coverage _less_ than this value. For example, if
+ RTP and UDP headers are to be protected, a receiver can enforce that only
+ packets with a minimum coverage of 20 are admitted:
+
+ int min = 20;
+ setsockopt(s, SOL_UDPLITE, UDPLITE_RECV_CSCOV, &min, sizeof(int));
+
+ The calls to getsockopt(2) are analogous. Being an extension and not a stand-
+ alone protocol, all socket options known from UDP can be used in exactly the
+ same manner as before, e.g. UDP_CORK or UDP_ENCAP.
+
+ A detailed discussion of UDP-Lite checksum coverage options is in section IV.
+
+
+
+ III) HEADER FILES
+
+ The socket API requires support through header files in /usr/include:
+
+ * /usr/include/netinet/in.h
+ to define IPPROTO_UDPLITE
+
+ * /usr/include/netinet/udplite.h
+ for UDP-Lite header fields and protocol constants
+
+ For testing purposes, the following can serve as a `mini' header file:
+
+ #define IPPROTO_UDPLITE 136
+ #define SOL_UDPLITE 136
+ #define UDPLITE_SEND_CSCOV 10
+ #define UDPLITE_RECV_CSCOV 11
+
+ Ready-made header files for various distros are in the UDP-Lite tarball.
+
+
+
+ IV) KERNEL BEHAVIOUR WITH REGARD TO THE VARIOUS SOCKET OPTIONS
+
+ To enable debugging messages, the log level must be set to 8, as most
+ messages use the KERN_DEBUG level (7).
+
+
+ 1) Sender Socket Options
+
+ If the sender specifies a value of 0 as coverage length, the module
+ assumes full coverage, transmits a packet with coverage length of 0
+ and according checksum. If the sender specifies a coverage < 8 and
+ different from 0, the kernel assumes 8 as default value. Finally,
+ if the specified coverage length exceeds the packet length, the packet
+ length is used instead as coverage length.
+
+
+ 2) Receiver Socket Options
+
+ The receiver specifies the minimum value of the coverage length it
+ is willing to accept. A value of 0 here indicates that the receiver
+ always wants the whole of the packet covered. In this case, all
+ partially covered packets are dropped and an error is logged.
+
+ It is not possible to specify illegal values (<0 and <8); in these
+ cases the default of 8 is assumed.
+
+ All packets arriving with a coverage value less than the specified
+ threshold are discarded, these events are also logged.
+
+
+ 3) Disabling the Checksum Computation
+
+ On both sender and receiver, trying to disable the UDP-Lite checksum
+ (option SO_NO_CHECK) in setsockopt(2) results in an error. Thus
+
+ setsockopt(sockfd, SOL_SOCKET, SO_NO_CHECK, ... );
+
+ will always result in an error, while
+
+ getsockopt(sockfd, SOL_SOCKET, SO_NO_CHECK, &value, ...);
+
+ will always return a value of 0 (meaning checksum enabled). Packets
+ with a zero checksum field are silently discarded by the receiver.
+
+
+ 4) Fragmentation
+
+ The checksum computation respects both buffersize and MTU. The size
+ of UDP-Lite packets is determined by the size of the send buffer. The
+ minimum size of the send buffer is 2048 (defined as SOCK_MIN_SNDBUF
+ in include/net/sock.h), the default value is configurable as
+ net.core.wmem_default or via setting the SO_SNDBUF socket(7)
+ option. The maximum upper bound for the send buffer is determined
+ by net.core.wmem_max.
+
+ Given a payload size larger than the send buffer size, UDP-Lite will
+ split the payload into several individual packets, filling up the
+ send buffer size in each case.
+
+ The precise value also depends on the interface MTU. The interface MTU,
+ in turn, may trigger IP fragmentation. In this case, the generated
+ UDP-Lite packet is split into several IP packets, of which only the
+ first one contains the L4 header.
+
+ The send buffer size has implications on the checksum coverage length.
+ Consider the following example:
+
+ Payload: 1536 bytes Send Buffer: 1024 bytes
+ MTU: 1500 bytes Coverage Length: 856 bytes
+
+ UDP-Lite will ship the 1536 bytes in two separate packets:
+
+ Packet 1: 1024 payload + 8 byte header + 20 byte IP header = 1052 bytes
+ Packet 2: 512 payload + 8 byte header + 20 byte IP header = 540 bytes
+
+ The coverage packet covers the UDP-Lite header and 848 bytes of the
+ payload in the first packet, the second packet is fully covered. Note
+ that for the second packet, the coverage length exceeds the packet
+ length. The kernel always re-adjusts the coverage length to the packet
+ length in such cases.
+
+ As an example of what happens when one UDP-Lite packet is split into
+ several tiny fragments, consider the following example.
+
+ Payload: 1024 bytes Send buffer size: 1024 bytes
+ MTU: 300 bytes Coverage length: 575 bytes
+
+ +-+-----------+--------------+--------------+--------------+
+ |8| 272 | 280 | 280 | 280 |
+ +-+-----------+--------------+--------------+--------------+
+ 280 560 840 1032
+ ^
+ *****checksum coverage*************
+
+ The UDP-Lite module generates one 1032 byte packet (1024 + 8 byte
+ header). According to the interface MTU, these are split into 4 IP
+ packets (280 byte IP payload + 20 byte IP header). The kernel module
+ sums the contents of the entire first two packets, plus 15 bytes of
+ the last packet before releasing the fragments to the IP module.
+
+ To see the analogous case for IPv6 fragmentation, consider a link
+ MTU of 1280 bytes and a write buffer of 3356 bytes. If the checksum
+ coverage is less than 1232 bytes (MTU minus IPv6/fragment header
+ lengths), only the first fragment needs to be considered. When using
+ larger checksum coverage lengths, each eligible fragment needs to be
+ checksummed. Suppose we have a checksum coverage of 3062. The buffer
+ of 3356 bytes will be split into the following fragments:
+
+ Fragment 1: 1280 bytes carrying 1232 bytes of UDP-Lite data
+ Fragment 2: 1280 bytes carrying 1232 bytes of UDP-Lite data
+ Fragment 3: 948 bytes carrying 900 bytes of UDP-Lite data
+
+ The first two fragments have to be checksummed in full, of the last
+ fragment only 598 (= 3062 - 2*1232) bytes are checksummed.
+
+ While it is important that such cases are dealt with correctly, they
+ are (annoyingly) rare: UDP-Lite is designed for optimising multimedia
+ performance over wireless (or generally noisy) links and thus smaller
+ coverage lenghts are likely to be expected.
+
+
+ V) UDP-LITE RUNTIME STATISTICS AND THEIR MEANING
+
+ Exceptional and error conditions are logged to syslog at the KERN_DEBUG
+ level. Live statistics about UDP-Lite are available in /proc/net/snmp
+ and can (with newer versions of netstat) be viewed using
+
+ netstat -svu
+
+ This displays UDP-Lite statistics variables, whose meaning is as follows.
+
+ InDatagrams: Total number of received datagrams (as in UDP).
+
+ InPartialCov: Number of received datagrams with csum coverage < length.
+
+ NoPorts: Number of packets received to an unknown port (as in UDP).
+ These cases are counted separately (not as InErrors).
+
+ InErrors: Number of erroneous UDP-Lite packets. Errors include:
+ * internal socket queue receive errors
+ * packet too short (less than 8 bytes or stated
+ coverage length exceeds received length)
+ * xfrm4_policy_check() returned with error
+ * application has specified larger min. coverage
+ length than that of incoming packet (cf. below)
+ * checksum coverage violated (InBadCoverage)
+ * bad checksum (InBadChecksum)
+
+ InBadCoverage: Datagrams with invalid checksum coverage (also InErrors):
+ * coverage length is less than the minimum 8
+ * coverage length exceeds actual datagram length
+
+ InBadChecksum: Datagrams with wrong checksum (also InErrors).
+
+ OutDatagrams: Total number of sent datagrams.
+
+ OutPartialCov: Number of sent datagrams with csum coverage < length.
+
+ If a receiving application has specified a minimum coverage length and
+ received a packet with a smaller coverage value than this, or if it has
+ specified full coverage (UDP mode) and received a partially covered packet,
+ this counts as error (under InErrors), and an error message is logged.
+
+ These statistics variables obey the following relations:
+
+ Total_received_datagrams = InDatagrams + InErrors + NoPorts
+
+ InErrors >= InBadCoverage + InBadChecksum
+
+ The `>' includes other errors such as socket queue errors (usually 0). For
+ IPv6, the same statistics variables are used, using the `UdpLite6' prefix,
+ and can be viewed using "grep ^UdpLite6 /proc/net/snmp6". Alternatively,
+ you can use the `nstat' utility found in the iproute2 package.
+
+
+
+ VI) OPEN ISSUES
+
+ 1) Sharing Code with UDP
+
+ On the mailing list there has been a suggestion to share code between
+ ipv?/udp.c and ipv?/udplite.c. There is indeed a potential for this, but the
+ challenge is not to mess up existing code. A line-by-line comparison between
+ ipv4/udp.c and ipv4/udplite.c revealed the following similarities:
+
+ * 45 functions appear in udp.c and modified in udplite.c
+ * 26 are used with trivial modifications (sed/perl could do this)
+ * 10 are used with minor changes (structure / sockopt names)
+ * 8 require real modifications (in control flow and algorithm)
+ * 1 function is missing in udplite.c (no equivalent of udp_check())
+
+ A summary of this analysis can be found on
+ http://www.erg.abdn.ac.uk/users/gerrit/udp-lite/udplite-comparison.html
+
+ Further similarities include structure identifiers, hence udp_seq_afinfo is
+ e.g. already reused in UDP-Lite; as are several UDP constants.
+
+ However, I doubt whether merging will make things better. In a lot of cases
+ the code is functionally identical but depends and operates on global data
+ structures and locks which are exported as kernel symbols:
+ * udp(lite)_hash
+ * udp(lite)_hash_lock
+ * udp(lite)_port_rover
+ * udp(lite)_statistics
+ Hence it would be necessary to rename these globals apart in both source code
+ files, which would lead to a lot of #ifdefs in udp.c and introduce a fragile
+ dependency between both. Any change made in udp.c would thus immediately
+ propagate to udplite.c - manual revision remains inevitable.
+
+ 2) MIB Standardisation
+
+ A MIB for UDP-Lite does not (yet) exist. For someone who is familiar with
+ SNMP/ASN.1 it would be an easy task to turn the above variable definitions
+ into a MIB - in the same manner as per e.g. RFC 2013. Anyone interested in
+ helping with this work should contact us at <[email protected]>.
+
+
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
[Index of Archives]
[Kernel Newbies]
[Netfilter]
[Bugtraq]
[Photo]
[Stuff]
[Gimp]
[Yosemite News]
[MIPS Linux]
[ARM Linux]
[Linux Security]
[Linux RAID]
[Video 4 Linux]
[Linux for the blind]
[Linux Resources]
