IPv6详解,第1卷,核心协议实现(英文影印版)(IPv6的权威参考书，好评如潮)

.2.2　ipv6 addresses　30
2.3　textual representation of ipv6 addresses　31
2.4　address scopes　33
2.4.1　scope zones　34
2.4.2　zone indices　36
2.4.3　textual representation of scoped addresses　38
2.4.4　deprecation of unicast site-local addresses　39
2.5　ipv6 address format　40
2.5.1　interface identifier generation　42
2.5.2　notes about address format　43
2.5.3　multicast address format　44
2.6　node address requirements　46
2.7　ipv6 address space management　47
2.8　code introduction　47
2.8.1　ipv6 address structures—in6_addr{} and sockaddr_in6{}　48
2.8.2　macros and variables　49
2.9　handling scope zones　51
2.9.1　initialization of scope zones　52
2.9.2　scope zone ids　53
2.9.3　zone ids in address structures　54
2.9.4　scope-related utility functions　57
2.10　interface address structures　64
2.10.1　ifaddr{} and in6_ifaddr{} structures　64
2.10.2　in6_ifreq{} and in6_aliasreq{} structures　67
2.10.3　multicast address structures　68
2.11　ipv6 prefix structure　70
2.12　overview of address manipulation routines　73
2.13　interface initialization for ipv6　76
2.13.1　in6_if_up() function　76
2.13.2　in6_ifattach() function　78
2.13.3　in6_ifattach_loopback() function　81
2.13.4　in6_ifattach_linklocal() function　82
2.13.5　get_ifid() function　87
2.13.6　get_hw_ifid() function　89
2.13.7　get_rand_ifid() function　93
2.13.8　in6if_do_dad() function　94
2.14　ipv6 interface address configuration　95
2.14.1　in6_control() function　95
2.14.2　in6_update_ifa() function　101
2.14.3　in6_joingroup() and in6_leavegroup() functions　115
2.14.4　in6_addmulti() and in6_delmulti() functions　116
2.14.5　in6_ifinit() function　118
2.14.6　in6_ifaddloop() and in6_ifloop_request() functions　120
2.15　deleting an ipv6 address　122
2.15.1　in6_purgeaddr() function　123
2.15.2　in6_ifremloop() function　124
2.15.3　in6_unlink_ifa() function　125
2.16　operation with address configuration utility　127
3　internet protocol version 6　131
3.1　introduction　131
3.2　ipv6 header format　132
3.2.1　comparison to the ipv4 header　133
3.3　ipv6 extension headers　134
3.3.1　order of extension headers　135
3.3.2　hop-by-hop options header　136
3.3.3　destination options header　137
3.3.4　routing header　137
3.3.5　fragment header　140
3.3.6　ipv6 options　142
3.4　source address selection　144
3.4.1　default address selection　144
3.4.2　source address selection　146
3.4.3　destination address selection　148
3.5　code introduction　149
3.5.1　statistics　151
3.5.2　header structures　151
3.5.3　ip6protosw{} structure　157
3.6　ipv6 packet address information in mbuf　160
3.6.1　ip6_setdstifaddr() function　162
3.6.2　ip6_getdstifaddr() function　162
3.6.3　ip6_setpktaddrs() function　163
3.6.4　ip6_getpktaddrs() function　164
3.7　input processing: ip6_input() function　164
3.8　processing hop-by-hop options header: ip6_hopopts_input() function　179
3.8.1　processing each option: ip6_process_hopopts() function　180
3.8.2　processing unknown option: ip6_unknown_opt() function　184
3.9　processing destination options header: dest6_input() function　185
3.10　reassembling fragmented packets　187
3.10.1　structures for packet reassembly　187
3.10.2　frag6_input() function　190
3.11　processing routing header: route6_input() function　204
3.12　forwarding: ip6_forward() function　209
3.13　output processing　219
3.13.1　source address selection—in6_selectsrc() function　219
3.13.2　route selection: ip6_selectroute() function　234
3.13.3　ip6_output() function　242
3.13.4　make extension headers: ip6_copyexthdr() function　276
3.13.5　split headers: ip6_splithdr() function　277
3.13.6　insert jumbo payload option: ip6_insert_jumboopt() function　278
3.13.7　fragmentation: ip6_insertfraghdr() function　281
3.13.8　path mtu determination: ip6_getpmtu() function　282
3.13.9　multicast loopback: ip6_mloopback() function　285
4　internet control message protocol for ipv6　287
4.1　introduction　287
4.2　icmpv6 message　288
4.2.1　destination unreachable message　289
4.2.2　packet too big message　291
4.2.3　time exceeded message　292
4.2.4　parameter problem message　293
4.2.5　echo request message　294
4.2.6　echo reply message　294
4.2.7　icmpv6 message processing rules　295
4.3　path mtu discovery mechanism　296
4.4　node information query　297
4.4.1　node information message format　299
4.4.2　noop query　301
4.4.3　supported qtypes query　301
4.4.4　node name query　301
4.4.5　node addresses query　303
4.4.6　ipv4 addresses query　304
4.5　code introduction　304
4.5.1　statistics　304
4.5.2　icmpv6 header　308
4.6　icmpv6 input processing　308
4.6.1　icmp6_input() function　310
4.6.2　notifying errors: icmp6_notify_error() function　320
4.7　path mtu discovery implementation　329
4.7.1　icmp6_mtudisc_update() function　329
4.8　icmpv6 output processing　332
4.8.1　sending error: icmp6_error() function　333
4.8.2　error rate limitation: icmp6_ratelimit() function　341
4.8.3　icmp6_reflect() function　342
4.9　node information query implementation　348
4.9.1　types and variables　348
4.9.2　ping6 command: send queries　350
4.9.3　ping6 command: receive replies　355
4.9.4　ping6 command: print supported qtypes　362
4.9.5　ping6 command: print node addresses　365
4.9.6　query processing: ni6_input() function　367
4.9.7　node name manipulation　376
4.9.8　create node addresses reply: ni6_store_addrs() function　383
4.10　node information operation　387
5　neighbor discovery and stateless address autoconfiguration　389
5.1　introduction　389
5.2　neighbor discovery protocol overview　390
5.3　stateless address autoconfiguration overview　391
5.4　nd protocol messages　392
5.5　example exchanges of nd protocol messages　.. 393
5.6　nd protocol packet types and formats　395
5.6.1　router solicitation message　396
5.6.2　router advertisement message　397
5.6.3　neighbor solicitation message　401
5.6.4　neighbor advertisement message　402
5.6.5　redirect message　403
5.7　neighbor discovery option types and formats　405
5.7.1　link-layer address options　405
5.7.2　prefix information option　406
5.7.3　redirected header option　407
5.7.4　mtu option　407
5.7.5　route information option　408
5.8　next-hop determination and address resolution　410
5.9　neighbor unreachability detection algorithm　411
5.10　stateless address autoconfiguration　412
5.10.1　address formation and address states　413
5.10.2　duplicate address detection algorithm　415
5.10.3　processing router advertisement　416
5.10.4　privacy extensions　417
5.11　router specific operation　419
5.11.1　sending unsolicited router advertisements　422
5.11.2　processing router solicitations　422
5.11.3　processing router advertisements　423
5.12　host specific operation　423
5.12.1　sending router solicitations　423
5.12.2　processing router advertisements　425
5.12.3　default router selection　426
5.13　code introduction　426
5.13.1　nd message definitions　427
5.13.2　neighbor cache—llinfo_nd6{} structure　429
5.13.3　operational variables—nd_ifinfo{} structure　432
5.13.4　default router—nd_defrouter{} structure　433
5.13.5　prefix—nd_prefix{} structure　434
5.13.6　prefix control—nd_prefixctl{} structure　436
5.13.7　nd message options—nd_opts{} structure　436
5.13.8　dad queue entry—dadq{} structure　437
5.13.9　ipv6 address—in6_ifaddr{} structure　438
5.13.10　destination cache　438
5.13.11　operation constants　438
5.14　initialization functions　439
5.14.1　nd6_init() function　439
5.14.2　nd6_ifattach() function　440
5.15　neighbor cache management functions　441
5.15.1　nd6_rtrequest() function　441
5.15.2　nd6_cache_lladdr() function　452
5.15.3　nd6_lookup() function　463
5.15.4　nd6_free() function　467
5.15.5　nd6_timer() function　470
5.16　nd protocol messages processing functions　478
5.16.1　nd6_ns_output() function　478
5.16.2　nd6_ns_input() function　484
5.16.3　nd6_na_input() function　493
5.16.4　nd6_na_output() function　504
5.16.5　nd6_rs_input() function　510
5.16.6　nd6_ra_input() function　513
5.16.7　icmp6_redirect_input() function　521
5.16.8　icmp6_redirect_output() function　528
5.17　nd protocol message options processing functions　536
5.17.1　nd6_option_init() function　536
5.17.2　nd6_option() function　536
5.17.3　nd6_options() function　538
5.18　default router management functions　540
5.18.1　defrouter_addreq() function　540
5.18.2　defrouter_delreq() function　541
5.18.3　defrouter_addifreq() function　542
5.18.4　defrouter_delifreq() function　544
5.18.5　defrouter_lookup() function　545
5.18.6　defrouter_select() function　546
5.18.7　defrtrlist_del() function　550
5.18.8　defrtrlist_update() function　553
5.19　prefix management functions　555
5.19.1　nd6_prelist_add() function　555
5.19.2　prelist_remove() function　557
5.19.3　prelist_update() function　558
5.19.4　find_pfxlist_reachable_router() function　568
5.19.5　prefix and address state about on-link condition　568
5.19.6　pfxlist_onlink_check() function　571
5.19.7　nd6_prefix_onlink() function　575
5.19.8　nd6_prefix_offlink() function　578
5.20　stateless address autoconfiguration functions　580
5.20.1　in6_ifadd() function　580
5.20.2　in6_tmpifadd() function　584
5.20.3　regen_tmpaddr() function　588
5.21　duplicate address detection functions　590
5.21.1　nd6_dad_find() function　590
5.21.2　nd6_dad_starttimer() function　590
5.21.3　nd6_dad_stoptimer() function　590
5.21.4　nd6_dad_start() function　591
5.21.5　nd6_dad_stop() function　594
5.21.6　nd6_dad_timer() function　594
5.21.7　nd6_dad_duplicated() function　598
5.21.8　nd6_dad_ns_output() function　599
5.21.9　nd6_dad_ns_input() function　600
5.21.10　nd6_dad_na_input() function　601
5.22　miscellaneous functions　602
5.22.1　nd6_is_addr_neighbor() function　602
5.22.2　nd6_output() function　604
5.22.3　rt6_flush() function　611
5.22.4　nd6_rtmsg() function　612
6　transport layer implications　615
6.1　introduction　615
6.2　tcp and udp over ipv6　616
6.3　pseudo header for ipv6　616
6.4　checksum difference between ipv4 and ipv6　617
6.5　ipv4-mapped ipv6 address usage　618
6.6　code introduction　618
6.6.1　protocol control blocks for ipv6　618
6.7　general operations on pcbs and sockets　627
6.7.1　ipv6 pcb allocation—in_pcballoc() function　627
6.7.2　bind local address—in6_pcbbind() function　630
6.7.3　fix remote address—in6_pcbconnect() function　639
6.7.4　function in6_pcbladdr()　642
6.7.5　search for a pcb entry—in6_pcblookup_local() function　644
6.7.6　search for ipv4-mapped pcb—in_pcblookup_local() function　647
6.7.7　search for a pcb entry—in6_pcblookup_hash() function　650
6.7.8　search for ipv4-mapped pcb—in_pcblookup_hash() function　651
6.7.9　detach an ipv6 pcb—in6_pcbdetach() function　653
6.7.10　control message signaling—in6_pcbnotify() function　655
6.7.11　flush pcb cached route—in6_rtchange() function　659
6.7.12　retrieve peer address—in6_setpeeraddr() function　660
6.7.13　retrieve local address—in6_setsockaddr() function　662
6.8　tcp-over-ipv6　663
6.8.1　tcp-over-ipv6 instance of ip6protosw{}　663
6.8.2　tcp output　663
6.8.3　initializing headers—tcp_fillheaders() function　669
6.8.4　tcp input—tcp6_input() and tcp_input() functions　670
6.8.5　tcp control input—tcp6_ctlinput() function　675
6.8.6　tcp user requests　678
6.9　udp-over-ipv6　685
6.9.1　udp-over-ipv6 instance of ip6protosw{}　685
6.9.2　udp output—udp6_output() function　685
6.9.3　udp input—udp6_input() function　692
6.9.4　udp control input—udp6_ctlinput() function　699
6.9.5　udp user requests handling　702
6.10　raw ipv6　709
6.10.1　raw ipv6 statistics　709
6.10.2　raw ipv6 output—rip6_output() function　710
6.10.3　raw ipv6 input—rip6_input() function　715
6.10.4　icmpv6 input—icmp6_rip6_input() function　719
6.10.5　raw ipv6 control input—rip6_ctlinput() function　724
6.10.6　raw ipv6 control output—rip6_ctloutput() function　725
6.10.7　raw ipv6 user requests handling　730
6.11　summary of operation with ipv4-mapped ipv6 addresses　738
6.12　viewing ipv6 connections with netstat　743
6.13　configuring ipv4-mapped ipv6 address support　745
7　socket api extensions　747
7.1　introduction　747
7.2　the basic socket api—[rfc3493]　748
7.2.1　basic definitions　748
7.2.2　interface identification　749
7.2.3　ipv4 communication over af_inet6 socket　750
7.2.4　address and name conversion functions　752
7.2.5　basic socket options　760
7.3　the advanced socket api—[rfc3542]　764
7.3.1　advanced definitions　764
7.3.2　ipv6 raw sockets　766
7.3.3　introduction to ancillary data　768
7.3.4　ipv6 packet information　770
7.3.5　manipulation of ipv6 extension headers　773
7.3.6　path mtu apis　778
7.3.7　socket extensions for the “r” commands　778
7.3.8　summary tables of socket options　780
7.4　kernel implementation of ipv6 socket apis　783
7.4.1　code introduction　783
7.4.2　ip6_pktopts{} structure　786
7.4.3　ipv6 socket option processing—ip6_ctloutput() function　790
7.4.4　getting socket options—ip6_getpcbopt() function　805
7.4.5　setting socket options and ancillary data　807
7.4.6　cleaning up—ip6_freepcbopts() function　823
7.4.7　ipv6 multicast socket options　824
7.4.8　ipv6 raw socket options—ip6_raw_ctloutput() function　835
7.4.9　icmpv6 socket options—icmp6_ctloutput() function　838
7.4.10　delivering incoming information—ip6_savecontrol() function　840
7.5　socket options and ancillary data examples　848
7.5.1　example of the send path　848
7.5.2　example of the receive path　850
7.6　implementation of library functions—libinet6　853
7.6.1　inet_pton() and inet_pton6() functions　853
7.6.2　inet_ntop() and inet_ntop6() functions　859
7.6.3　getaddrinfo() function　863
7.6.4　address ordering examples　888
7.6.5　freeaddrinfo() function　895
7.6.6　gai_strerror() function　897
7.6.7　getnameinfo() function　898
7.6.8　other library functions　906
references　909
index　915
about the authors ...　938

.　　·Chapter 6--IPv6 and IP Security
　　The figure below graphically depicts the coverage of the various aspects of IPv6 by these two books. We call this book "Book I" and the Advanced Protocols book "Book II."
　　These two books cover a number of protocol specifications, which are mainly published as Request for Comments (RFCs) and are mutually inter-related. The figure on the next page summarizes the relationship among the major RFCs covered in this book series and the chapters that describe the particular specification. The arrows show reference dependency between two particular specifications. For example, the arrow from RFC3041 to RFC2461 indicates the former refers to the latter.
　　Each chapter contains two main parts, each of which consists of orie or two consecutive sections. The first part provides a digest of the main protocol specifications published as RFCs that are covered in that chapter. For a protocol chapter such as Chapters 3, 4, and 5, the first part describes the basic operations of specific protocols. The first part defines the various protocol packets, explains the meaning and the purpose of each field in each packet, and describes how the related protocol works with the packets. This part also tries to Clarify ambiguous text in the RFCs where necessary.
　　The second part of each chapter concentrates on describing data structures and functions of KAME which implement the RFCs. Diagrams are provided to illustrate graphically the relation- ships among the various data structures. Function call graphs are provided to show code paths.In this part, code narrations may also clarify the ambiguities that exist in the specifications, and also identify areas in the RFCs that are difficult to implement. -
　　Some chapters have an additional trailing part. It describes useful utilities and demonstrates their usages, thereby highlighting how the implementation works in actual operation.
　　Intended Audience and How to Use this Book
　　This book can be used as a first book on IPv6 for those developers who are new to either IPv6 or KAME. This book will serve as a reference text for those developers who must either port or maintain KAME-based systems. This is the case for those BSD developers who must assume adoption of the KAME code and maintain and enhance KAME on different variants of BSD.
　　This book is also suitable as a textbook for an undergraduate senior level computing science course on IPv6 and related protocols. This book can also serve as a reference text of a project-based course on IPv6 for first-year graduate students. Instructors must set as prerequisites the fundamental computing science courses on computer networks, network programming, and a course on TCP/IP (optional). The students are also assumed to be fluent with the C programming language.
　　Readers are generally encouraged to follow the chapters in sequence, but extensive cross references throughout the book enable a reader to jump directly to the subject of interest from chapter to chapter without the concern of being lost in code mazes. For those readerswho are familiar with the existing TCP/IP technologies, Chapters 6 and 7 will provide a useful transition guide on how to port existing applications to IPv6 and how to write IPv6-enabled applications.
　　The detailed descriptions on the various design decisions reveal the deficiencies that exist in either the specification or the implementation. These insights will enable a researcher to seek out new research topics. The knowledge gained in KAME allows a researcher to establish' KAME as a platform on which to build experimental solutions..
　　Accompanying CD-ROMs
　　This book comes with 2 CD-ROMs. The first CD-ROM is an ISO image of FreeBSD4.8-RELEASE,which is the base operating system discussed in this book. It is a bootable CD-ROM and includes installation files. The installation procedure will start by turning on the computer with the CD-ROM. The detailed installation procedure can be found in the INSTALL. TXT 'file located in the root directory of the CD-ROM.
　　The second CD-ROM contains source files of FreeBSD4.8-RELEASE and the KAME snapshot release that are referred to in this book. There are two subdirectories in the root directory of the CD-ROM.
　　·FreeBSD4.8-RELEASE---contains the entire source tree of FreeBSD4.8-RELEASE
　　·kame-snap-2 0030421---contains the source tree of the KAME snapshot created on April 21, 2003
　　All source files described throughout this book can be found in the second CD-ROM.Readers can find the complete source code in the CD-ROM as they read this book. It' is also possi-ble to install the KAME implementation using the source tree under the kame -snap - 20030421 directory to check the actual behavior of the KAME stack. The installation procedure of the KAME snapshot is shown in Section 1.3.2.
　　Source Code Copyright
　　This book presents many parts of the source code developed by KAME and also refers to other source code distributed as part of the FreeBSD systems. All of the source code has Copyright notices, which are available in the copy of the source code contained in the second CD-ROM.
　　Reporting Errors and Errata Page
　　Although we tried hard to identify and fix errors in the review and update phase of this book, it is inevitable to miss some of the errors in this size of material. The authors are happy tO receive error reports on the content of this book, and plan to provide an error correction page on the Internet. It will be available at the following web page address: http://books.elsevier, com/companions/0124477518.
　　Acknowledgments
　　Personal Acknowledgment from Li Qing
　　In January of 2003, having just completed a three-year embedded IPv6 project at Wind River,I was collecting my notes and papers in response to a call-for-paper in ACM SIGCOMM on new communication protocols.
　　While reading through my thick notebook containing code paths, descriptions on various data structures and code fragments, numerous diagrams, and countless notes-to-self on the KAME implementation, I thought to myself how nice it would have been if there were a book on KAME. Then the idea struck me: I could put all of these notes into a book, which would help others gain understanding ofiPv6 and jump start their projects. This thought took a life of its own and grew to an ambitious attempt to cover the entire KAME implementation.
　　I presented this book idea to Jinmei-san, a Core KAME developer in March of 2003, and invited him to join me in this endeavor. Jinmei-san is a well-known and respected researcher in IPv6 and various important Internet technologies, such as Multicast Routing, DHCPv6, and DNS. He gladly accepted my offer.
　　Shima-san, another Core KAME developer joined us at the end of March. Shima-san is a well-known and respected researcher in both IPv6 and Mobility.
　　We agreed on a plan of action by May of 2003 and the book is complete nearly three years later.
　　Jinmei-san provided technical leadership throughout the entire book. His insighifut and meticulous attention to content, detail, and technical accuracy ensured the quality of this book.He helped me with content and has always been responsive to my technical questions and queries even with his busy 24-hour work schedule,
　　Shima-san helped me with both content and revision. Especially toward the end of the project when I was inundated by both my job and personal life, Shima-san and Jinmei-san carried me through to the finish line. Shima-san has certainly contributed more than his share of responsibility.
　　It has been a wonderful learning experience working with my co-authors. Their professional excellence, diligence, and work ethics have set a new standard for me to follow.
　　I would like to thank Rick Adams for sharing my vision of a book and then investing his patience and time over the past three years. I have always appreciated during our many conversations the fact that his frustrations were expressed in such a controlled manner, since the majority of times we spoke I was explaining why the'schedule had slipped yet again.
　　I want to thank my wife'Huaying Cheng for her understanding and support of my many late nights and long weekends spent in front of the computer.
　　Personal Acknowledgment from Jinmei, Tatuya
　　In addition to the contributors acknowledged below, I personally would like to thank the managers of Toshiba who have supported my activity in the KAME project and authorized this derivative work: Yukio Kamatani, Toshio Murai, Yasuhiro Katsube, and Atsushi Inoue. I would not have been a KAME developer without their understanding and support to begin with, much less written a book on KAME.
　　Similarly, I would not be what I am without Jun Murai and Hiroshi Esaki. They have always encouraged me in WIDE and KAME, provided new opportunities, and taught me the spirit of global contribution. Murai-san also kindly accepted our request of writing a foreword to this book.
　　Finally, I would like to thank my co-authors. Clearly, if Qing had not invited me, I would never have imagined writing a book on KAME. He has always generously accepted my delay in writing drafts, and kept the publishing contract valid while I was tweaking every detail of the book, causing another half a year of delay. Shima-san helped me when I desperately updated the draft chapters toward the deadline--far beyond the call of duty.
　　Persona!Acknowledgment from Shima, Keiichi
　　First of all, I would like to thank all the other KAME developers who had created the high-quality IPv6 reference code. This book would not exist without the code, that is, without their continuous effort to polish the code and their firm minds to deploy IPv6 technok)gy with their code. Fortunately, I could join the project, which was an honor to me, and I could contribute to this great project. It was a precious time and an irreplaceable experience for me. I was a student when I started IPv6-related work. Murai Jun-san and Yamamoto Kazuhiko-san gave me the chance to work on it at the WIDE project. I cannot imagine what I'would be if I never met them. Utashiro Kazumasa-san and Wada Eiiti-san supported my activity in the KAME project.All of my work could not have been done without the Internet technology. I thank all of the people who are making a continuous effort to operate the Internet.
　　Group Acknowledgment
　　The authors are indebted to all the KAME developers. This is also "their book" in that they wrote the KAME code, which is half of this book. In particular, Shinsuke Suzuki read every chapter and provided valuable comments and suggestions. Kenjiro Cho, Tsuyoshi Momose, and Ryuji Wakikawa carefully reviewed selected chapters, and helped us improve the book through their numerous detailed comments. We would also like to thank external reviewers, specifi-cally, Rob Austein, George Neville-Nell, Andre Oppermann, Shawn Routhier, Randall Stewart,Shinta Sugimoto, and Kazumasa Utashiro. They voluntarily spent a large amount of their private time. The cover image of this book was based on the well-known KAME image, which was used as a symbol of the KAME project designed by Manabu Higashida and Chizuru Higashida (atelier momonga).
　　We would also like to thank Gary R. Wright and W. Richard. Stevens whose book established our basic knowledge of TCP/IP. Their model for describing the BSD's networking code is so effective that we decided to borrow that model for describing KAME's IPv6 implementation in our books. Part of our dedication goes to them for their contribution to the networking engineering community. We hope to make the same level of contribution for the next generation of engineers in the IPv6 technology.
　　Finally, our sincere gratitude goes to our editors Rick Adams, Rachel Roumeli0tis, Dawn-marie Simpson, and the editorial staff at Morgan Kaufmann/Elsevier for their diligent efforts that materialized our manuscript into a professional book...