Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Fetching contributors…

Cannot retrieve contributors at this time

2742 lines (2197 sloc) 107.709 kB
CHANGELOG for KAME kit, 1998
$KAME: CHANGELOG.1998,v 1.3 2000/07/26 08:09:55 itojun Exp $
<199812>
Thu Dec 31 20:56:12 JST 1998 itojun@iijlab.net
* kit/ports/ucd-snmp: upgrade base version to ucd-snmp 3.5.3.
Thu Dec 31 03:37:03 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
* kit/ports/socks64:
made compilable on KAME FreeBSD 3.0.
(Include if_var.h. Should be removed in the future,
also with removal of in6_var.h)
Wed Dec 30 22:28:58 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
* kit/src
Made then compilable on KAME FreeBSD 3.0.
Especially many ifdef's are added to route6d/ifmcstat.c.
Tue Dec 29 18:07:52 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
* sys/net,netinet,netinet6
sync with FreeBSD3.0 as much as possible.(mainly netinet6)
Fri Dec 25 22:06:04 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
* sys/netinet6/nd6_rtr.c: Added consideration of ndpr_rrf_decrvalid
and ndpr_rrf_decrprefd for address lifetime initialization.
Without this, prefixes allocated by prefix command will be
IN6_IFF_DEPRECATED after some period of time.
Fri Dec 25 17:02:08 JST 1998 itojun@iijlab.net
* kit/ports/bind8 (FreeBSD): IPv6-ready bind8. named will accept
queries to IPv6 UDP/TCP port 53, dig/nslookup/whatever are able to
make queries toward IPv6 UDP/TCP port 53, and so forth.
1998-12-24 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* in6_proto.c,ip6_input.c,ip6_var.h: removed none_input().
Now a packet whose protocol is IPPROTO_NONE can safely be passed
to the userland.
netinet/in_proto.c was also modified.
Thu Dec 24 19:40:27 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
* kit/src/libinet6/resolv/res_debug.c
add ifdef of T_UINFO, T_UID, T_GID, to make it compilable on
FreeBSD 3.0.
1998-12-24 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* probe.c (probe_init): call shutdown() after opening the probe socket
to make the socket `send-only'.
Thu Dec 24 11:46:38 JST 1998 itojun@iijlab.net
* sys/netinet6/raw_ip6.c: setsockopt(IPV6_CHECSUM) sometimes caused
SEGV due to a bug in mbuf boundary checks. It is now fixed.
Thu Dec 24 03:11:19 JST 1998 itojun@iijlab.net
* kit/ports/apache13: updated to use new patch.
(fixed args for freeaddrinfo())
Reported by:
Florent Parent <Florent.Parent@viagenie.qc.ca>
Andreas Wrede <andreas@planix.com>
Tue Dec 22 20:10:40 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
* kit/src/faithd/tcp.c: Before terminating a relay process,
shutdown s_snd. This make opposite-direction relay process
to terminate also.
1998-12-22 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* mld6.c (mld6_input): Fixed a problem that zero divide occurs
when receiving a MLD query with Maximum Response Delay smaller
than 200(including zero).
Thanks to Niels Baggesen <Niels.Baggesen@uni-c.dk> for reporting the
problem and sending a patch.
1998-12-22 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* ip6_output.c (ip6_setmoptions): For link-local multicast
detection, use IN6_IS_ADDR_MC_LINKLOCAL instead of
IPV6_ADDR_INT16_MLL.
Thanks to: Tetsuya Isaki <isaki@se.hiroshima-u.ac.jp>
1998-12-22 Atsushi Onoe <onoe@sm.sony.co.jp>
* kit/src/libinet6/rcmd.c: fix declaration of iruserok() for
NetBSD.
Sat Dec 19 14:02:44 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
* kit/src/faithd/faithd.c, tcp.c
do closelog() and (re)openlog() for child after fork.
check EINTR for select() and read().
Fri Dec 18 12:25:49 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
* kit/src/faithd/tcp.c: BUG fix:
fixed select fds settting. add check of send result and retry.
clean-up'ed select routine.
(Thanks for jinmei-san for code review and comment, also thanks for
onoe-san for many background informations)
Thu Dec 17 00:26:18 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
* kit/src/faithd/tcp.c: BUG fix; use global integer rcvon and
writeon, to control the set/unset of readfds for s_rcv and
writefds for s_snd in select();
Wed Dec 16 13:30:46 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
* kit/src/faithd/faithd.c: BUG fix; give syslog() correct buffer
pointer. This fix the strange syslog() output problem on child
exit.
Wed Dec 16 12:48:53 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
* kit/src/faithd/tcp.c
Fork in tcp_relay() for going relay traffic and coming relay traffic.
And in those each process, do non-blocking write() so that OOB data
can be forwarded preferrably.
1998-12-15 Atsushi Onoe <onoe@sm.sony.co.jp>
* kit/src/libinet6/rcmd.c, rresvport_af.c: add compatible wrapper
functions to avoid conflict of symbols.
1998-12-14 Atsushi Onoe <onoe@sm.sony.co.jp>
* kit/usr.bin/telnet/commands.c: support source route for IPv4
and IPv6 (@gw1@gw2@dest).
1998-12-11 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* bgp.c (connect_process): modified some code fragments not to
call fatal even if {set,get}sockopt is failed. This is necessary
to interoperate with some(e.g. Cisco) implementations when the
peer is not listening to the BGP port.
1998-12-10 SUMIKAWA Munechika <sumikawa@ebina.hitachi.co.jp>
* syncronized netinet6/* codes of three OSs as much as possible
Thu Dec 10 04:14:59 JST 1998 shin@nd.net.fujitsu.co.jp
* sys/neitnet/
* sys/neitnet6/
changed IPPROTO_NONE as return value to IPPROTO_DONE.
and use IPPROTO_NONE only for protocl type value.
also, this fixes mbuf leak bug when received a packet with
IPPROTO_NONE.
1998-12-10 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/bgpd: supported `next hop self' when sending a BGP4+ UPDATE
message to an IBGP peer.
1998-12-10 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/bgpd/parse.c: changed restriction of using the `preference'
keyword for an EBGP peer only.
Sat Dec 5 04:53:05 JST 1998 sakane@ydc.co.jp
* kit/src/racoon:
Soft lifetime is set to 80% of hard lifetime.
This rate can be defined which you like by calling
pfkey_lifetime_rate().
Fri Dec 4 02:26:13 JST 1998 sakane@ydc.co.jp
* kit/src/racoon:
It's fixed to handle session for PF_KEY.
It's enable to display the entries on the negotiation of phase 2.
About address semantics for varius case is commented into isakmp.h.
Wed Dec 2 23:44:00 JST 1998 itojun@iijlab.net
* sys/netinet6: (NetBSD) fixed odd behavior in ND6. Now ND6 works
properly as expected.
Wed Dec 2 13:17:21 JST 1998 sakane@ydc.co.jp
* sys/netkey/key.c:
Fixed to hung up the kernel when running two of racoon.
<199811>
Sat Nov 28 00:54:28 JST 1998 itojun@iijlab.net
* kit/src/route6d: route tag support. route6d can advertise route tag
by "-t 0x1234". rip6query will show the advertised route tag,
if non-zero value is advertised.
Fri Nov 27 JST 1998 itojun@iijlab.net
* kit/lib/libutil: (FreeBSD) logwtmp() which takes care of IPv6 address
that does not fit UT_HOSTSIZE. (not really tested)
This was intended to replace original shared library by new libutil
to override logwtmp() used by /usr/bin/login. However,
/usr/bin/login records username/hostname by itself. Therefore,
the attempt was failed.
Fri Nov 27 01:54:10 JST 1998 itojun@iijlab.net
* kit/lib/libskey: (FreeBSD) S/Key library capable of handling
IPv6 hostnames listed in /etc/skey.access.
You can override standard /usr/lib/libskey.so.2.0 by doing
"ldconfig -m /usr/local/v6/lib". By doing so /usr/bin/login
will be able to handle IPv6 hostnames without re-compilation.
To test this add the followin entry to /etc/skey.access and try
a telnet session to ::1.
permit internet ::1 ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff
"prefixlen" syntax is added.
permit internet ::1/128
Thu Nov 26 10:15:35 JST 1998 k-sugyou@kame.net
- (FreeBSD) support IPv6 firewall
- kit/sbin/ip6fw: controll utility
Wed Nov 25 18:36:46 JST 1998 itojun@iijlab.net
* kit/libexec/telnetd: perform setsockopt(IP_TOS) only if the perr
is IPv4 host.
Tue Nov 24 18:53:38 JST 1998 sakane@ydc.co.jp
* kit/src/racoon:
A new Diffie-Hellman group as number 5 was supported.
Sat Nov 21 06:15:30 JST 1998 itojun@iijlab.net
* sys/netinet6/icmp6.c: ICMP6 redirect processing was wrong.
routing table update event was not properly propagated toward
upper-layers.
Sat Nov 21 03:55:17 JST 1998 itojun@iijlab.net
* KAME IPv6 on NetBSD-pmax is now confirmed to work.
From: Feico Dillema <dillema@acm.org>
Date: Fri, 20 Nov 1998 16:53:41 +0100
Subject: (KAME-snap 210) Status Report KAME and NetBSD-1.3.2-pmax
Fri Nov 20 17:00:35 JST 1998 itojun@iijlab.net
* sys/netinet6/nd6_rtr.c: Update prefix information option processing.
(experimental, FreeBSD/BSDI only)
Add more comments to RA processing code.
Separate prefix lifetime and address lifetime. Address lifetime
will be kept in struct in6_ifaddr. Implement "2 hour" rule for
address lifetime, which prevents DoS attack (hopefully).
TODO: on-link determination must be updated.
* kit/sbin/ifconfig: Add code to print out address lifetime.
It looks too noisy and commented out by default.
Thu Nov 19 23:28:43 JST 1998 shin@nd.net.fujitsu.co.jp
* kit/etc/rc.net6: don't assign prefix if $prefix is null
Thu Nov 19 19:03:53 JST 1998 shin@nd.net.fujitsu.co.jp
* sys/netinet6/nd6.h: removed reserved field of
the bit field, ndpr_stateflags, to also remove the necessity of
changing the size of reserved field at new bit member addition.
Wed Nov 18 20:48:07 JST 1998 itojun@iijlab.net
- Header chain chasing support for tcpdump.
To use this, user must specify "ip protochain x" or
"ip6 protochain x".
Since BPF code for header chasing cannot be optimized and is a bit
slow, this is separate from "ip proto x".
For example, "ip6 protochain 6" should capture any IPv6 packet
with TCP header (TCP with AH, or TCP with hop-by-hop option,
can be captured).
Wed Nov 18 10:05:42 JST 1998 itojun@iijlab.net
* kit/ports/lynx: fix security hole in "rlogin://" URL.
(obtained from bugtraq mailing list)
From: Artur Grabowski (art@STACKEN.KTH.SE)
Date: Tue, 17 Nov 1998 17:06:00 +0100
http://www.geek-girl.com/bugtraq/1998_4/0489.html
Wed Nov 18 04:00:32 JST 1998 shin@nd.net.fujitsu.co.jp
* kit/src/rtsol/rtsol.c
make compilable on FreeBSD 3.0
1998-11-17 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/pim6dd: added to support PIMv6 dense mode.
Pim6dd was based on pimdd developed at the University of Oregon.
Tue Nov 17 16:08:45 JST 1998 shin@nd.net.fujitsu.co.jp
* kit/src/gifconfig/gifconfig.8
manual update for gif multi dest extensions contributed from
dwiggins@bbn.com.
also descriptions about IPv6 support and enable switch of this
extension by link0 flag on/off is also added.
Tue Nov 17 14:57:51 JST 1998 shin@nd.net.fujitsu.co.jp
* sys/net/if.c,if.h,if_gif.c,route.c,route.h,rtsock.c
* sys/netinet/in_gif.c,in_gif.h
* sys/neitnet6/in6_gif.c,in6_gif.h
* kit/sbin/route/route.8,route.c
merged gif multi dest extensions contributed from dwiggins@bbn.com.
IPv6 support and enable switch of this extension by link0 flag on/off
is also added.
Tue Nov 17 01:51:12 JST 1998 shin@nd.net.fujitsu.co.jp
* kit/src/gifconfig.c:
support of printing physical IPv6 address value
Mon Nov 16 22:14:50 JST 1998 shin@nd.net.fujitsu.co.jp
* sys/net/if_gif.c: Bug Fix
added SIOC{S,G}IF{PHY,PSRC,PDST}*_IN6 ioctls to enable gif
tunneling over IPv6.
Mon Nov 16 17:57:26 JST 1998 itojun@iijlab.net
* kit/src/libinet6/rcmd.c: Update ahost (1st argument)
only if canonical hostname is available. This is to preserve
the original behavior.
Sat Nov 14 18:25:06 JST 1998 itojun@iijlab.net
* kit/usr.bin/telnet/commands.c (FreeBSD/NetBSD): preserve original
behavior.
- if the hostname to connect to is numeric, perform canonical name
lookup (look for PTR record, i.e. gethostbyaddr).
- if the hostname is non-numeric, do not perform canonical name
lookup.
* kit/usr.bin/telnet/commands.c (BSDI): preserve original behavior.
- never perform canonical name lookup.
Fri Nov 13 01:36:57 JST 1998 itojun@iijlab.net
* sys/crypto/sha1.c: fixed SHA-1 computation bug when the data source
is sized multiple of 64bytes. Thanks goes to Chris Winters
<winters@itd.nrl.navy.mil> for detailed bug report!
Thu Nov 12 20:14:19 JST 1998 itojun@iijlab.net
* sys/netinet6/in6.c: avoid hardcoding prefixlen == 64bit in
SIOC[ADG]LIFADDR processing.
bitwidth of interface id is always 64bit (defined in RFC2373),
but prefixlen may NOT be 64bit. RFC2373 allows non-RFC2374 address
encoding scheme. (see figure on page 8)
Thu Nov 12 19:53:34 JST 1998 shin@nd.net.fujitsu.co.jp
* kit/src/rrenumd/rrenumd.c
"AE" options added for rrenumd authentication and
encryption.
Thu Nov 12 16:13:53 JST 1998 itojun@iijlab.net
* kit/src/{libpcap,tcpdump}: support IPv6 address in pcap expression.
tcpdump host ::1
tcpdump net 3ffe:0501::/32
TODO: libpcap now requires getaddrinfo() if --enable-ipv6 is specified.
configure should check the existence and use alternatives
(missing/getaddrinfo.c?) if none found.
TODO: "gateway" syntax is not working in --enable-ipv6 setting.
Wed Nov 12 11:43:54 JST 1998 shin@nd.net.fujitsu.co.jp
* sys/net/if_dummy.c: correct if_type to IFT_DUMMY
Wed Nov 11 18:16:56 JST 1998 itojun@iijlab.net
* sys/netinet/in.c: implement SIOC[ADG]LIFADDR.
Wed Nov 11 14:16:00 JST 1998 itojun@iijlab.net
* kit/libexec/ftpd(FreeBSD): perform ioctl(IP_TOS) and
setsockopt(TCP_NOPUSH) to be performed only in IPv4 ftp connection.
login.cap is now supported (but never tested).
logwtmp() is fixed to log IPv6 numeric hostname as much as possible.
I dunno if it is right or not.
(previously logged as "invalid hostname")
Wed Nov 11 13:56:59 JST 1998 itojun@iijlab.net
* kit/src/rtsol: use SIOC[ADG]LIFADDR when possible. this is mainly
for test purposes, but looks nice.
Wed Nov 11 12:02:10 JST 1998 itojun@iijlab.net
* sys/netinet6/in6.c: support SIOC[ADG]LIFADDR for IPv6 address.
see ipngwg mailing list #6621 (October 10). IPv6 address support
will be added when IPv6 version is confirmed to be working right.
Tue Nov 10 16:23:50 JST 1998 itojun@iijlab.net
* sys/netinet/ip_output.c: (FreeBSD) prevent ipfw code from SEGV.
NOTE: we are still wondering whether ipfw code works right with
KAME or not. Your inputs and bug reports would be really helpful.
Tue Nov 10 16:10:26 JST 1998 itojun@iijlab.net
* kit/ports/ssh: update IPv6 patch to 1.4.
Tue Nov 10 12:31:53 JST 1998 itojun@iijlab.net
* kit/src/racoon: eliminate u_int{8,16,32}. use u_int{8,16,32}_t for
better portability.
* kit/src/racoon: support lifetime type "kb". NOTE: no kernel support
for expiration yet.
Tue Nov 10 00:21:04 JST 1998 itojun@iijlab.net
* sys/netinet6: accumulate bytes transferred over SA, so that we can
define lifetime by bytes (sadb_lifetime_bytes) in the future.
* kit/src/setkey: display bytes transferred over SA.
Mon Nov 9 15:51:28 JST 1998 itojun@iijlab.net
* sys/netkey/key.c: pass FQDN and USERFQDN identity extension
on ACQUIRE message. we need to check if it is allowed to pass
multiple identity extension to userland (racoon dislikes this).
Mon Nov 9 15:10:16 JST 1998 itojun@iijlab.net
* kit/src/setkey: changed the meaning of -h flag.
was: print usage and exit, now: display hexadecimal dump on -x.
Sat Nov 7 00:29:19 JST 1998 shin@nd.net.fujitsu.co.jp
* kit/src/rrenumd/Makefile, lexer.l, rrenumd.8, rrenumd.conf.5
man update, and fixed lexer file to support comment in conf file.
1998-11-06 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/bgpd/bgp_input.c (bgp_read): fixed a problem that bgpd
stopped when an ETIMEDOUT error occurred on a BGP socket.
Fri Nov 6 16:27:07 JST 1998 itojun@iijlab.net
* sys/crypto/cast128/cast128.c: speed up by replacing core functions
by macros.
Message-Id: <19981031020640N.suzuki@grelot.elec.ryukoku.ac.jp>
Date: Sat, 31 Oct 1998 02:06:40 +0900
From: Tomomi Suzuki <suzuki@grelot.elec.ryukoku.ac.jp>
Fri Nov 6 14:00:30 JST 1998 itojun@iijlab.net
* sys/netinet6: ND6 cleanups.
- remove old lladdr caching code. utilize nd6_cache_lladdr.
- if we got RS/RA/NS/redir packet without lladdr, make an neighbor
cache entry with NOSTATE state (= considered PASSIVE).
- checked relationship between neighbor cache and defrouter list. it
seems fine (there will be no defrouter list without neighbor cache).
Thu Nov 5 JST 1998 itojun@iijlab.net
* kit/src/route6d: ripng fix: update route lifetime only if the
advertisement is from same gw, with same metric. (see p13 of RFC2080)
Thanks to: ikuo@intec.co.jp
Thu Nov 5 21:14:01 JST 1998 shin@nd.net.fujitsu.co.jp
* kit/src/libinet6/resolv/res_debug.c
Made this really compilable on FreeBSD 3.0.
Thu Nov 5 20:33:02 JST 1998 shin@nd.net.fujitsu.co.jp
* kit/src/libinet6/ifname.c
* kit/src/libinet6/resolv/res_debug.c
* kit/src/ndp/ndp.c
* kit/src/prefix/prefix.c
Made compilable on FreeBSD 3.0.
Now prefix assignment seems to be successful.
Thu Nov 5 04:26:55 JST 1998 shin@nd.net.fujitsu.co.jp
* sys/netinet6/in6_prefix.c
BUG FIX: fixed matched prefix length validity check to comply
with spec.
Thu Nov 5 02:59:55 JST 1998 shin@nd.net.fujitsu.co.jp
* kit/src/Makefile
added rrenumd to SUBDIR.
Thu Nov 5 02:45:59 JST 1998 shin@nd.net.fujitsu.co.jp
* kit/src/rrenumd/rrenumd.c, parser.y, lexer.l, Makefile, rrenumd.8
fixed bugs and now it seems to be sending valid rrenum msgs.
* kit/src/rtadvd/rrenum.c,rtadv.c
fixed bugs and now seems to be successfully renumbering when
received rrenum msgs from rrenumd.
* kit/src/prefix/prefix.c
changed default value of use_prefix length.
Wed Nov 4 23:41:05 JST 1998 itojun@iijlab.net
* sys/netinet6/esp*: cleanup ESP pad length processing.
base spec requires 4n, cbc algorithms require 8n.
* sys/netinet6/ah_core.c: make sure to skip ifindex portion
in ip6 src/dst address.
Wed Nov 4 JST 1998 itojun@iijlab.net
* kit/src/racoon: fix IPv6 ID payload.
* sys/netkey/key.c: changed internal structure for SA management.
SA will be held into per-state linked list, not per-protocol
linked list.
Wed Nov 4 00:29:02 JST 1998 itojun@iijlab.net
* kit/src/tcpdump: try checking buggy implementation of CAST128.
SSLeay 0.9.0b has a bug in encryption round # on short keys -
rounds should be 12 for key <= 80bits.
Tue Nov 3 19:58:13 JST 1998 itojun@iijlab.net
* sys/crypto/blowfish: fixed cbc mode processing. now it should be
interoperable with other implementations (need testing).
Mon Nov 2 01:02:30 JST 1998 itojun@iijlab.net
* kit/ports/gated-ipv6: pathname of original distribution changed.
<199810>
Mon Oct 31 JST 1998 itojun@iijlab.net
* kit/src/racoon: AH algorithm must be determined by hash algorithm
type attribute, not the transform type.
* kit/src/racoon: sanity checker for config file improved.
* sys/netinet6/esp_output.c: fixed a serious bug in ESP tunnel output,
which mistakes policy determination and send packets in clear (simple
tunnel, not ESP tunnel) in some configuration.
* kit/src/tcpdump: ID payload now printed properly.
* kit/src/racoon: parser improvements. makefile improvements.
link print-isakmp.c from tcpdump so that packets can be monitored
after decryption, in debug mode.
Fri Oct 30 21:52:50 JST 1998 itojun@iijlab.net
* sys/netinet6, kit/usr.bin/netstat: gather more stats on
IPsec operations.
Thu Oct 29 JST 1998 itojun@iijlab.net
* kit/src/racoon: ignore notification payload on phase 1 negotiation
(responder-lifetime). this is necessary for interop with RedCreek
when responder-lifetime does not match.
* kit/src/racoon: compute long cipher key for phase 1 properly
(for example 3DES)
* kit/src/racoon: phase 2 quick mode: attach fake ID payload for
debugging (configurable)
* kit/src/racoon: ignore commit bit (we don't support this yet)
* kit/src/racoon: bug fix in DELETE payload processing.
TODO: handle it more properly, (i.e. remove SA if possible)
* kit/src/racoon: send and check Vendor ID. (does nothing tricky
at this moment)
* kit/src/racoon: phase 2 AH proposal must include authentication
method attribute. reject non-conforming proposal on config file,
and on the packet from the peer.
* kit/src/racoon: filter out phase 2 proposal that does not match
the SA type requested from the kernel. For example, AH proposals
will be filtered out when ESP SA is requested.
* kit/src/racoon: improve parser code.
Wed Oct 28 JST 1998 itojun@iijlab.net
* kit/src/racoon: SA payload fixes. (1) SAi_b must be the whole
SA payload sent from the initiator. (2) responder must send the
selected proposal only, not the whole payload.
* kit/src/racoon: phase 2 PFS fix. config file format has changed.
one must specify PFS DH group in phase 2 configuration, not per-
transform configuration.
* kit/src/racoon: ESP with authentication is now supported.
generate longer KEYMAT for this.
* kit/src/racoon: improve warnings on ATTR payload format.
* kit/src/racoon: bark if there's no "remote anonymous" section.
* kit/src/tcpdump: isakmp and ipsec improvements.
Wed Oct 28 13:54:48 JST 1998 itojun@iijlab.net
* kit/src/racoon: better PFS (Perfect Forward Secrecy) support.
RFC keyed MD5 support. ignore Vendor ID payload (we may check
content of Vendor ID payload in the future).
Tue Oct 27 23:28:45 GMT 1998 sakane@ydc.co.jp
* kit/src/racoon:
In phase 1, using real address as ID payload,
if ID was not specified in config file.
Tue Oct 27 22:37:30 GMT 1998 sakane@ydc.co.jp
* kit/src/racoon:
Applied t_id except hash_t when decision AH algorithm.
Wed Oct 28 07:05:57 JST 1998 itojun@iijlab.net
* kit/src/setkey: support keyed SHA1.
* sys/netkey: add more information about supported algorithms into
SADB_REGISTER message.
* sys/netinet6: cleanup AH/ESP algorithm table. add key length
information into the table.
Tue Oct 27 22:06:26 GMT 1998 sakane@ydc.co.jp
* kit/src/racoon:
Added Some comment about checking payload.
Implemented new SA payload parser.
Removed enc_t in ipsec_sa structure.
Supported to handle key length per algorithm.
Mon Oct 26 11:34:13 JST 1998 itojun@iijlab.net
* sys/netinet6: IPv4 options processing. not tested.
I believe that it will not work if there's source route option,
since ip_dooptions() rewrites the ip header.
Sun Oct 25 15:21:37 JST 1998 itojun@iijlab.net
* kit/src/rtsol: avoid kvm_read(). use ioctl() instead, to grab
interface information.
Sun Oct 25 JST 1998 itojun@iijlab.net
* sys/netinet6: add more sanity checks in esp{4,6}_input() and
ah{4,6}_input(), to avoid panic in heavy ipsec sessions.
Sat Oct 24 03:02:43 JST 1998 shin@nd.net.fujitsu.co.jp
Added parser to rrenumd. But not seems to be working yet.
Also man is not up to date.
Thu Oct 22 04:05:15 JST 1998 sakane@ydc.co.jp
* sys/netkey, kit/src/racoon:
Fixed the behavior about ACQUIRE, GETSPI, UPDATE and ADD.
There were some mistakes. Changed that kernel doesn't make a entry
for acquiring when SADB_ACQUIRE.
Wed Oct 21 22:57:25 JST 1998 itojun@iijlab.net
* made rfc AH work again.
* fix ipsec{4,6}_hdrsiz() (bug caused SEGV on AH tunnel case)
* wrap IF_ENQUEUE() by splimp()
Wed Oct 21 19:44:45 JST 1998 itojun@iijlab.net
* midway.c(en ATM driver on FreeBSD/BSDI): fix transmit buffer
management. in specific condition driver stops xmit'ing.
Wed Oct 21 15:52:23 JST 1998 sakane@ydc.co.jp
* kit/src/racoon:
Begin to handle Information Exchange. need more coding.
1998-10-21 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* if_gif.c (gif_input): put incoming packets to a network layer
queue instead of directly calling an input function to prevent
too many recursive function calls.
Wed Oct 21 13:16:39 JST 1998 itojun@iijlab.net
* kit/ports/gated-ipv6: port for famous routing daemon, GateDaemon IPv6.
Wed Oct 21 12:11:38 JST 1998 sakane@ydc.co.jp
* kit/src/racoon:
changed the way to compute KEYMAT.
changed the handling SPI and KEYMAT in pfkey_update() and pfkey_add().
NOTE: When SA expire, racoon will be strange behavior. To be fixed.
Wed Oct 21 01:19:41 JST 1998 itojun@iijab.net
* sys/netinet6: Update AH tunnel authenticity checking code.
Consider outer IP header authentic (if it gets authenticated),
and assume nothing (no authenticity) to inner IP header.
* sys/netinet6: more IPsec statistics.
Tue Oct 20 16:47:00 JST 1998
* sys/netinet{,6}: make AH tunnel mode working for IPv4.
* sys/netinet6: more statistics for AH.
* sys/netinet6: better sanity checks for IPv4 AH/ESP tunnel.
Tue Oct 20 13:54:27 JST 1998 itojun@iijlab.net
* sys/netinet6: make des-derived work. need interop tests.
Mon Oct 19 19:48:25 JST 1998 itojun@iijlab.net
* remove unused code/defines in ipsec.
* log() fixes.
* mark des-derived not working by rejecting it in esp_descbc_mature().
(iv management is not right)
1998-10-19 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* netstat/inet6.c (pim6_stats): added to print PIM for IPv6 statistics.
Mon Oct 19 17:39:48 JST 1998
* sys/netkey/key.c: variable "sab" was defined twice in key_checksab()
and it made all packets to be sent in clear. it is now fixed.
sorry for your troubles.
Sun Oct 18 JST 1998
* kit/src/tcpdump: add some code to dump isakmp packets,
on udp port 500. However, most part of the exchange is encrypted
(and that part cannot be decoded).
1998-10-17 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/bgpd/bgp.c (bgp_process_update): Several bugs were fixed.
The bugs were mostly about BGP4+ route reflector.
Thu Oct 15 16:17:09 JST 1998 itojun@iijlab.net
* sys/netinet6 and kit/usr.bin/netstat: added some ipsec statistics.
1998-10-14 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/bgpd/dump.c: added to dump bgpd status to a file. The status
includes various information such as bgpd internal routing table
and BGP4+ per peer status. Please do not forget to execute the
configure command before compiling.
Man pages were also updated.
Sat Oct 14 18:31:25 JST 1998 shin@nd.net.fujitsu.co.jp
* kit/etc/rc.net6
changed to use "prefix" command instead of "ifconfig" command
in router case.
Wed Oct 14 17:44:36 JST 1998 itojun@iijlab.net
* kit/sys/netkey, kit/sys/net/rtsock.c: PF_KEY and PF_ROUTE sockets
are stabilized. it should work fine against severe tests.
location of splnet() was wrong.
Sat Oct 14 16:06:16 JST 1998 shin@nd.net.fujitsu.co.jp
* kit/src/Makefile
added "prefix" command as to be installed by default.
Wed Oct 14 11:30:27 JST 1998 itojun@iijlab.net
* kit/sys/netkey: properly handle IPv6 address passed by SADB_ACQUIRE.
* kit/src/racoon: IPv6 support. guess IPv6 stack type, socket/bind
to IPv6 unspecified addr, and so forth. need more confirmation on
portability.
Sat Oct 14 11:05:48 JST 1998 shin@nd.net.fujitsu.co.jp
*sys/netinet6/in6_var.h
*sys/netinet6/in6_prefix.c
changed bit field structure member size from u_long to u_char,
because BSDI suppose the size differently between kernel
and userland.
And merged some diffs of in6_prefix.c between BSD variants.
now "prefix" command seems to work on BSDI.
Wed Oct 14 03:52:11 JST 1998 itojun@iijlab.net
* kit/ports/apache13: Port for apache 1.3.3. For non-FreeBSD OSes,
IPv6 patch is available from ftp://ftp.kame.net/pub/kame/misc/.
* kit/ports/apache12: renamed from kit/ports/apache (port for apache
1.2.6). 1.3.3 is highly recommended over 1.2.6.
Tue Oct 13 23:45:04 JST 1998 itojun@iijlab.net
* kit/src/setkey: add -x option, which dumps all the message
transmitted to PF_KEY socket. (uses SADB_X_PROMISC).
Tue Oct 13 23:27:36 JST 1998 itojun@iijlab.net
* sys/netkey: support SADB_X_PROMISC. maybe good for debuggin'.
Tue Oct 13 21:21:27 JST 1998 itojun@iijlab.net
* kit/src/racoon: make racoon code free from CPU endian.
now racoon works on KAME on NetBSD/sparc too.
(namely, eay_bn2v() and eay_v2bn() are updated)
Tue Oct 13 15:35:16 JST 1998 itojun@iijlab.net
* kit/src/racoon: be more strict about checking SSLeay's existence.
previously we checked md5.h, but some operating systems have md5.h
by default.
Tue Oct 13 14:22:32 JST 1998 itojun@iijlab.net
* sys/netkey: Add splnet() to prevent race condition.
* sys/netkey/keysock.c: Changed the way sadb_msg is sent to userland.
PF_KEY defines three ways to send sadb_msg to userland:
(1) to requesting process only, (2) to all listening processes, and
(3) to all registered processes. The implementation now conforms
to this.
1998-10-12 Atsushi Onoe <onoe@sm.sony.co.jp>
* kit/src/libinet6/name6.c
use res_query() for reverse lookup instead of res_search().
allow IPv4-compat address for getipnodebyaddr(), do not perform
any query for "::" to conform bsd-api-new-02.
allow misalign address for getipnodebyaddr().
Sun Oct 11 23:31:35 JST 1998 itojun@iijlab.net
* sys/net*: (NetBSD) IPsec is now working. Now we need to perform
bunch of tests...
Sun Oct 11 22:52:24 JST 1998 itojun@iijlab.net
* sys/netinet6/{esp,ah}_core.c: bark if no secret key is specified
for esp/ah algorithms that require secret key.
Sat Oct 11 22:35:59 JST 1998 shin@nd.net.fujitsu.co.jp
*kit/src/prefix/prefix.8
*kit/src/prefix/prefix.c
added several checking of missing args, and changed some default
behaviour.
Sun Oct 11 20:37:06 JST 1998 itojun@iijlab.net
* kit/{sbin,usr.sbin}/sysctl and sys/netkey (NetBSD and BSDI):
add net.key.* sysctl MIBs. for FreeBSD we already got net.key.*.
Sat Oct 11 01:03:17 JST 1998 shin@nd.net.fujitsu.co.jp
*sys/netinet6/in6_prefix.c
changed "panic" to "log(LOG_ERR...)" in bit_copy().
Sat Oct 11 00:42:15 JST 1998 shin@nd.net.fujitsu.co.jp
*kit/src/prefix/prefix.8
*kit/src/prefix/prefix.c
update usage description of man and program.
Sun Oct 10 JST 1998 itojun@iijlab.net
* sys/netinet6 and kit/sbin/ifconfig (NetBSD): fix ifconfig to
some extent, so that we can check status of if address flags (such as
"anycast"). there are some fixes necessary (ioctl API design issues).
Sat Oct 10 14:46:50 JST 1998 shin@nd.net.fujitsu.co.jp
*kit/src/prefix/prefix.c
fixed usage description.
removed unused function.
Sat Oct 10 03:27:44 JST 1998 shin@nd.net.fujitsu.co.jp
*kit/src/prefix/Makefile
Made compilable on NetBSD
also, this command seems to work on NetBSD
Sat Oct 10 02:59:02 JST 1998 shin@nd.net.fujitsu.co.jp
*kit/sbin/ifconfig.c
ifconfig.8
removed prefix related enhance(because they are moved to
new "prefix" command)
Sat Oct 10 02:42:00 JST 1998 shin@nd.net.fujitsu.co.jp
*kit/src/prefix/Makefile
prefix.c
prefix.8
Newly added these files.
Actually these are prefix related functions from
current KAME FreeBSD sbin/ifconfig.
Same functions in sbin/ifconfig will be removed.
Only working on FreeBSD now.
TODO: operational check on BSDI
compile check on NetBSD
complete man page
Sat Oct 10 01:54:58 JST 1998 shin@nd.net.fujitsu.co.jp
*sys/netinet6/in6_prefix.c
Bug Fix:
change ">>" to ">>=". (discovered by itojun)
Fri Oct 9 21:48:19 JST 1998 shin@nd.net.fujitsu.co.jp
*kit/sbin/ifconfig/ifconfig.c
enabled "-a" for prefix renumbering commands
print usage for prefix related commands
shorten long parameters.
Fri Oct 9 20:02:21 JST 1998 shin@nd.net.fujitsu.co.jp
*kit/src/rtadvd/rrenum.c
*kit/sbin/ifconfig/ifconfig.c
*sys/netinet6/in6_prefix.c
supported SIOCAIFPREFIX_IN6, SIOCCIFPREFIX_IN6,
SIOCSGIFPREFIX_IN6, by ifconfig.
And fixed several kernel bugs discovered using those commands.
Now prefix renumbering by ifconfig seems to be working well.
1998/10/09 17:06:51 JST shin@nd.net.fujitsu.co.jp
i386/conf Makefile.i386
separated SYSTEM_LD macro into 2 case, where "-g" is defined and not.
1998/10/09 13:32:16 JST shin@nd.net.fujitsu.co.jp
i386/conf Makefile.i386
add "ulimit" to SYSTEM_LD macro, not only to SYSTEM_LD_TAIL macro.
Fri Oct 9 11:52:33 JST 1998 itojun@iijlab.net
* kit/src/faithd: improve command/result parsing in ftp translation.
support EPSV ALL. reject PORT and PASV from client as it is bogus
for IPv6 ftp connection.
Thu Oct 8 21:09:30 JST 1998 itojun@iijlab.net
* kit/src/faithd: redesign ftp.d completely, to make the translator
code more context-free. Also, EPSV/EPRT is supported.
TODO: utilize "EPSV ALL" for improved performance,
better error recovery
1998-10-08 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* if_gif.c (gif_output): prevented infinite call of gif_output
by introducing a counter variable which is static in this
function. Note that this approach may introduce MUTEX problem
when using kernel thread.
Tue Oct 7 18:20:01 JST 1998 shin@nd.net.fujitsu.co.jp
implemented following cmd in kernel.
SIOCSGIFPREFIX_IN6
SIOCAIFPREFIX_IN6
SIOCCIFPREFIX_IN6
TODO: enhance ifconfig and rrenumd to utilize these cmds,
and test kernel behavior
1998-10-07 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* ip6_mroute.c: implemented kernel-level IPv6 multicast
forwarding. It can be compiled, but there have been no userland
routing daemon yet. So it will not effectively work for a while.
Wed Oct 7 13:04:01 JST 1998 itojun@iijlab.net
* take care of IPsec tunnel in computing MTU and TCP MSS.
ipsec{4,6}_hdrsiz is defined for this.
{esp,ah}*_hdrsiz_* are deprecated.
Wed Oct 7 1998 itojun@iijlab.net
* experimental ND6 code is enabled in KAME/BSDI and KAME/FreeBSD.
we are trying to figure out the following spec flaws:
- discovery-v2-03 talks almost nothing about how to manage neighbor
cache entry on reception of RA/RS/NS/redirect without link-layer
address option.
- IsRouter flag sometimes becomes out-of-sync, due to neighbor
cache expiration/creation rules.
we are still thinking about the spec, and changing nd6_cache_lladdr().
the experimental code works just fine so the change will not bite
you.
Wed Oct 7 00:33:03 JST 1998 itojun@iijlab.net
* kit/lib/libftpio: (FreeBSD only) Fixed IPv4 non-passive ftp.
(bind failed due to wrong argument)
Tue Oct 6 18:28:13 JST 1998 itojun@iijlab.net
If a packet is to be forwarded over IPsec tunnel, and it couldn't
due to "too big and don't fragment", report the correct tunnel MTU
toward the originator.
tunnel MTU = if MTU - sizeof(IP header) - ESP/AH headers/paddings
To test this, you may need
sysctl -w net.inet.ipsec.dfbit=1
to set DF bit on the outer IP header.
Tue Oct 6 13:48:42 JST 1998 itojun@iijlab.net
* sys/netinet6/ipsec.c: changed the way IPsec tunnel is created.
(see ipsec4_encapsulate() in sys/netinet6/ipsec.c)
* sys/netinet6/ipsec.c: define new sysctl MIB, net.inet.ipsec.dfbit,
to allow users to control DF bit treatment (copy/clear/set) on
ipsec tunnel encapsulation.
NOTE: this is per-host configuration, not a per-interface
configuration defined in draft-ietf-ipsec-arch-sec-07.txt.
Tue Oct 6 13:13:19 JST 1998 shin@nd.net.fujitsu.co.jp
fix for rtr renumbering related structure's member name and
order change at ifconfig, rrenumd, rtadvd
maybe minimum implementation of router renumbering at rtadvd completed
Tue Oct 6 12:57:50 JST 1998 shin@nd.net.fujitsu.co.jp
changed router renumbering related structure's member name and order.
added same interface check for SIOC*IFPREFIX_IN6 cmds.
added in6_rrenumreq structure for advanced ioctls for rtr renumbering
defined, SIOCAIFPREFIX_IN6, SIOCCIFPREFIX_IN6, SIOCSGIFPREFIX_IN6
TODO: implement new SIOC*PREFIX_IN6 cmds in kernel
Mon Oct 5 17:20:13 JST 1998
* Eliminate clause 3 from our KAME copyright notice, as we've heard
that 4-clause BSD copyright irritates people very much.
Mon Oct 5 10:46:05 JST 1998
* kit/ports/sendmail6: make it buildable, by removing -I/usr/src/sys
from site.config-v6.kame.
1998/10/03 00:59:54 JST shin@nd.net.fujitsu.co.jp
ports/mozilla Makefile
ports/mozilla/files md5
Patch level up.
-IPv6 hostname with AAAA record,
or numarical IPv6 address escaped by [ ],
can be specified as proxy server.
-adopted __res_state structure change.
1998/10/02 23:54:27 JST shin@nd.net.fujitsu.co.jp
src/ndp ndp.c
netinet6 nd6.h nd6.c nd6_nbr.c
Added "ln_expire" to llinfo_nd6 structure, and "expire" to
in6_nbrinfo structure.
NDP use them for state transition and rt_expire is no more used.
Also, ndp command is changed to use ln_expire to display each
entry's expire time.
Fri Oct 2 13:13:01 JST 1998 itojun@iijlab.net
* kit/ports/apache: distribute IPv6 patch separately, from
ftp://ftp.kame.net/pub/kame/misc/.
Thu Oct 1 22:50:38 JST 1998 itojun@iijlab.net
* kit/src/rtadvd: add capability "nolladdr" which controls
the presence of source link-layer address option on RA packets.
(mostly for debugging)
* kit/src/ndp: add option "-A" which tries "-a" (show NDP entries)
repeatedly.
Thu Oct 1 11:12:25 JST 1998 itojun@iijlab.net
* kit/ports/sendmail6: mark this port broken as it is not buildable
due to the change in resolver (see below).
<199809>
1998-09-30 Atsushi Onoe <onoe@sm.sony.co.jp>
* sys/net/if_atmsubr.c, sys/dev/en/midway.c:
fix bugs to allow "ifconfig up" for ATM-PVC interface without
assigning IPv4 address.
1998-09-30 Atsushi Onoe <onoe@sm.sony.co.jp>
* include/resolv.h, kit/src/libinet6/resolv/:
restore struct __res_state to original to keep binary
compatibility (avoid SEGV on NetBSD).
change default configuration options for resolver.
* kit/src/libinet6/name6.c:
change syntax for AI_ALL (now needs AI_V4MAPPED) to conform to
bsd-api-new-02(a).
* include/netdb.h, kit/src/libinet6/name6.c:
add AI_V4MAPPED_CFG to return conditional answer of IPv4-mapped
IPv6 address depends on whether kernel's mapped_addr flag is set.
Also change the definition of AI_DEFAULT to (AI_V4MAPPED_CFG|
AI_ADDRCONFIG).
* kit/src/libinet6/getaddrinfo.c:
support AI_NUMERICHOST flag to conform to bsd-api-new-02a.
replace CHECK_KERNPROTO by AI_ADDRCONFIG of getipnodebyname().
fix 'a.foo.bar and a.v6.foo.bar' problem in PF_UNSPEC case.
Wed Sep 30 13:26:22 JST 1998 sakane@ydc.co.jp
* sys/netkey/key.c:
check to be zero of acq_seq, because zero is reserved
as handling SADB_EXPIRE.
1998/09/30 12:48:37 JST shin@nd.net.fujitsu.co.jp
netinet6 ip6_input.c
Don't think packets destined to RTF_GATEWAY route as "goto ours".
1998/09/30 02:59:38 JST shin@nd.net.fujitsu.co.jp
netinet6 ip6_output.c
Copy m_flags(M_MCAST) to fragmented packets to disable
neighbor resolution procedure for them.
(because neighbor resolution waiting queue length is
for only one packet)
1998/09/30 02:24:11 JST shin@nd.net.fujitsu.co.jp
. USAGE
small grammar fix
1998/09/29 10:09:00 JST shin@nd.net.fujitsu.co.jp
usr.sbin Makefile
Commented out ppp and added suggestion to use kit/ports/ppp.
Still leave dir of kit/usr.sbin/ppp for a while.
Mon Sep 28 17:40:18 JST 1998 itojun@iijlab.net
* sys/netinet6/in6.h: Renamed IPV6_{JOIN,LEAVE}_MEMBERSHIP into
IPV6_{JOIN,LEAVE}_GROUP to conform to bsd-api-new-02a.
1998/09/26 14:13:27 JST shin@nd.net.fujitsu.co.jp
src/rtadvd rrenum.c
Adopted to router-renum-05.txt.
And still supporting, not finished.
1998/09/25 16:34:20 JST shin@nd.net.fujitsu.co.jp
src/rrenumd rrenumd.c
Change router renumbering packet formats as new draft
(router-renum-05.txt)
1998/09/25 16:31:44 JST shin@nd.net.fujitsu.co.jp
netinet6 icmp6.h
Change router renumbering packet formats as new draft
(router-renum-05.txt)
Also, changed values type to u_int{8,16,32}_t as same as
other icmp6 structure.
Thu Sep 24 21:51:19 JST 1998 itojun@iijlab.net
* sys/sys/socket.h: Changed CMSG_xxx macro defs. Previously, ALIGN()
was used (based on advanced API document). However, ALIGN() in
advanced API and ALIGN() in BSD unix variants (machine/param.h) have
very different meanings. So, now we've defined CMSG_ALIGN().
Thu Sep 24 19:42:22 JST 1998 itojun@iijlab.net
* kit/ports/sendmail: Update the IPv6 patch for sendmail to be used.
It is now 891+3.1W. (W means WIDE sendmail patchkit)
IPv6 patch has been distributed separately, but now it is merged
into WIDE sendmail patchkit.
1998-09-24 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/bgpd: Added new feature to restrict RIPng routes filtering
routes that do not match a list of prefixes.
Filtering function is also clarified and `default restrict'
attribute disappeared. Instead, `filterin default' and/or
`filterout default' attributes should be used.
Note that configuration files which used older filtering functions
should be rewritten.
The manpage was also updated, which should be read carefully.
Thu Sep 24 13:19:26 JST 1998 itojun@iijlab.net
* sys: (KAME/FreeBSD) updated ALTQ kernel part to 1.1.2.
* kit/ports/altq: (KAME/FreeBSD) use ALTQ 1.1.2.
From: Kenjiro Cho <kjc@csl.sony.co.jp>
Thu Sep 24 13:19:26 JST 1998 itojun@iijlab.net
* sys: (KAME/BSDI) Revamp PVC ATM code. It is now compatible with
KAME/FreeBSD code. Userland tools must be fetched from ALTQ
package. (ftp://ftp.csl.sony.co.jp/pub/kjc/)
* kit/src/apconfig: (KAME/BSDI) removed because it is not needed
any more.
From: Kenjiro Cho <kjc@csl.sony.co.jp>
1998/09/24 00:03:18 JST shin@nd.net.fujitsu.co.jp
src/rtadvd rrenum.h
Router Renumbering related header.
(Very sorry, I forgotten to add this)
1998/09/23 10:49:27 JST shin@nd.net.fujitsu.co.jp
netinet6 icmp6.c icmp6.h in6.c in6.h in6_proto.c
in6_var.h nd6.c nd6.h nd6_nbr.c nd6_rtr.c
Sync with FreeBSD/BSDI/NetBSD
Only for files and places where I changed recently or I thought I can do.
Wed Sep 23 01:43:54 JST 1998 sakane@ydc.co.jp
sys/netinet6/esp_core.c:
Modified to do check DES weak key.
Tue Sep 22 22:59:21 JST 1998 itojun@iijlab.net
* src/faith: clarify options. improve behavior on error.
usage: faith [-de] [-p prefix]
-d: disable
-e: enable
-p: configure prefix
1998/09/22 14:59:36 JST shin@nd.net.fujitsu.co.jp
src/rtadvd config.c rrenum.c rtadvd.c
When noticed new prefix addition, try to get the prefix information via
SIOCGIFPREFIX.
If can get it, and its origin is from RA, ignore it.
Also router renumbering support is proceeding.(still need more work)
1998/09/22 14:55:40 JST shin@nd.net.fujitsu.co.jp
netinet6 nd6_rtr.c
Change rt_request() to rtinit() in in6_ifadd() to let it send
rtmsg to rtsock. Thus rtadvd become able to know that new prefix
is added.
1998-09-22 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/ndp/ndp.c: option name clarification;
- replaced -F option with -H option
- replaced -Fp with -P
- replaced -Fr with -R
1998/09/22 00:51:14 JST shin@nd.net.fujitsu.co.jp
Modified files: (Branch: kame_227)
ports/mozilla Makefile
Removed files: (Branch: kame_227)
ports/mozilla/patches patch-bj patch-bk
Modification of sockaddr_in6 structure as kernel change.
Change of library linking order so that libinet6 is linked for
gethostbyname2(), not libc. (by sugyo-san)
Patches/patch-* are deleted and now the total patch file is retrieved
from ftp://ftp.kame.net/pub/kame/misc.
1998/09/21 22:19:26 JST shin@nd.net.fujitsu.co.jp
netinet6 in6_prefix.c nd6.h nd6_rtr.c
Bug FIX:
Correctly init setting ndpr_prefix's lifetimes.
Added prefix lifetime argument check.
Also, don't age nd_prefix entries whose ndpr_rrf_decrvalid and/or
ndpr_rrf_decrprefd is 0.
1998/09/21 22:14:49 JST shin@nd.net.fujitsu.co.jp
sbin/ifconfig ifconfig.c
Bug Fix
Don't break from arg check loop, when encountered an unknown arg,
even if "prefix" command was specified.
1998/09/21 20:54:10 JST shin@nd.net.fujitsu.co.jp
sbin/ifconfig ifconfig.8
Added description about prefix related add/delete extension.
Mon Sep 21 19:40:10 JST 1998 sakane@ydc.co.jp
sys/netinet6:
Fixed to check replay counter.
It's available to define variable window size.
1998/09/21 17:57:59 JST shin@nd.net.fujitsu.co.jp
netinet6 in6_prefix.c nd6_rtr.c
Bug Fix:
Point ndpr_prefix from ndpr_ifpr.ifpr_prefix in prelist_add()
where ndpr's memory is allocated.
Not in in6_prefix_ctl() which use stack to keep ndpr structure.
Now prefix deletion seems to work.
1998/09/20 13:06:00 JST shin@nd.net.fujitsu.co.jp
netinet6 in6_prefix.c nd6.h nd6_rtr.c
Put 2 cases in in6_prefix_ioctl into 1.
Let prelist_add(), prelist_update() return errno.
1998/09/20 11:29:12 JST shin@nd.net.fujitsu.co.jp
netinet6 in6_prefix.c nd6.h nd6_rtr.c
Moved if_prefixlist link/unlink into prelist_add/prelist_remove.
But prefix deletion via SIOCDIFPREFIX is not working yet.
1998/09/20 11:28:52 JST shin@nd.net.fujitsu.co.jp
netinet6 in6_prefix.c nd6.h nd6_rtr.c
Moved if_prefixlist link/unlink into prelist_add/prelist_remove.
But prefix deletion via SIOCDIFPREFIX is not working yet.
1998/09/20 11:21:46 JST shin@nd.net.fujitsu.co.jp
sbin/ifconfig ifconfig.c
Set in6_prereq's prefixlen in in6_getprefixlen()
Also changed the way of printing error message at prefix set/delete.
1998/09/19 17:57:21 JST shin@nd.net.fujitsu.co.jp
netinet6 ah_core.c ah_input.c ah_output.c dest6.c
esp_core.c esp_input.c esp_output.c
frag6.c icmp6.c in6_ifattach.c in6_pcb.c
in6_proto.c in6_rmx.c ip6_forward.c
ip6_input.c ip6_output.c ipsec.c mld6.c
nd6.c nd6_nbr.c raw_ip6.c tcp6_input.c
tcp6_subr.c tcp6_usrreq.c udp6_usrreq.c
Include in_var.h instead of in6_var.h
(Because just happen to found it)
1998/09/19 17:44:39 JST shin@nd.net.fujitsu.co.jp
netinet6 in6.c in6_prefix.c nd6.h nd6_rtr.c
Correct initialization of setting prefix contents.
Let prelist_update() and prelist_add() return nd_prefix structure.
Now prefix addition works but deletion is not yet.
1998/09/19 01:43:51 JST shin@nd.net.fujitsu.co.jp
sbin/ifconfig ifconfig.c
Added prefix set/delete operation. But not seems to be working now.
TODO debugging to make it work right.
1998/09/19 01:41:13 JST shin@nd.net.fujitsu.co.jp
netinet6 in6_var.h nd6.h nd6_rtr.c in6_prefix.c
Changed macro name NDPR_ORIG* to PR_ORIG*, and also moved its defs
from nd6.h to in6_var.h.
Changed the place of failsafe check in in6_prefix.c.
Fri Sep 18 13:54:11 JST 1998 itojun@iijlab.net
* kit/libexec/ftpd: add support for "EPSV num" and "EPSV ALL",
so that we can interoperate with NRL ftp client.
"EPSV ALL" support is a fake, it will just get accepted and do
nothing.
Fri Sep 18 04:57:37 JST 1998 itojun@iijlab.net
* sys/netinet6/ip6.h: router alert option type is now "5", not "14".
some of userland tools (such as tcpdump) need recompilation.
(ftp://playground.sun.com/pub/ipng/assignments/ipv6-parameters.txt)
1998/09/17 23:10:01 JST shin@nd.net.fujitsu.co.jp
conf files
net if.h
netinet6 in6.c in6_proto.c in6_var.h nd6.h
nd6_rtr.c tcp6_input.c
netinet6 in6_prefix.c
Support prefix SET,GET,DEL command over ioctl.
(implementation/confirmation from application is not yet)
Linked ifprefix(nd_prefix) entries from ifnet structure.
So if_prefixlist member is added to the end of ifnet structure.
All user-land applications which looks net/if.h should be recompiled.
And Bug Fix (discovered by jinmei san):
Still a sockaddr_in6 is used as an in6_addr in nd6_rtr.c
So Replaced temporal sockaddr_in6 by nd_prefix->ndpr_prefix.
1998/09/17 18:25:44 JST shin@nd.net.fujitsu.co.jp
netinet6 nd6_rtr.c
Bug Fix:
Copy only RA flags on RA receipt which already has correspondent
nd_prefix entry. (discovered by jinmei san)
Also, set ndpr_dynamic for nd_prefix entry created by receiving RA.
And only update dynamic nd_prefix entry when RA is received.
1998/09/17 15:04:03 JST shin@nd.net.fujitsu.co.jp
src/ndp ndp.c
As kernel structure definition change,
As kernel structure definition change, changed flags to bit field.
1998-09-17 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/traceroute6/traceroute6.c: Added -l option, which specifies
to print both numerical addresses and host names.
Because of this addition, the default format changed from printing both
to printing only host names.
Thu Sep 17 12:51:41 JST 1998 fujisawa@kame.net
Add SuMiRe nat/protocol router.
Thu Sep 17 04:05:03 JST 1998 sakane@ydc.co.jp
sys/netkey/key.c:
SAB do become mature when SA with proxy have become mature.
Tue Sep 15 09:45:35 JST 1998 sakane@ydc.co.jp
kit/src/setkey,sys/netkey/keyv2.h:
It's changed that the name of extension flags for padding.
Each are seq to seq-pad, zero to zero-pad, random to random-pad.
It's available to define cyclic sequence number for replay prevention.
This is not to be used.
Tue Sep 15 07:41:57 JST 1998 sakane@ydc.co.jp
sys/netkey/keydb.h:
mistaken key length was supplied by _KEYLEN().
1998/09/14 22:49:20 JST shin@nd.net.fujitsu.co.jp
net if.h
netinet6 nd6.c nd6.h nd6_rtr.c
Preparing for prefix addition over ioctl.
Modified nd_prefix structure to link it to protocol independent interface
prefix list.(which is not yet implemented)
Also, some nd_prefix member name is changed to avoid name collision.
Mon Sep 14 20:56:03 JST 1998 sakane@ydc.co.jp
kit/src/racoon:
It's available to handle expiration, but I don't know
to handle HARD lifetime. How should I do when key had been exchanged
by expiration of SOFT lifetime. Now, SADB_EXPIRE HARD is ignored.
Mon Sep 14 20:47:01 JST 1998 sakane@ydc.co.jp
sys/netkey/key.c:
Set expiration time when new SA generated as LARVAL by SADB_GETSPI.
Mon Sep 14 17:30:22 JST 1998 sakane@ydc.co.jp
kit/src/racoon,kit/src/setkey,sys/netkey/key.c:
Infinite loop had occurred in kdebug_sadb().
Mon Sep 14 14:10:59 JST 1998 sakane@ydc.co.jp
kit/src/racoon:
Fixed a bug that racoon generated many ISAKMP-SA
per a destination address.
Mon Sep 14 04:24:17 JST sakane@ydc.co.jp
kit/src/racoon:
Many items was fixed.
racoon can exchange KEY in limited environment.
see TODO.
Mon Sep 14 04:00:07 JST sakane@ydc.co.jp
kit/src/setkey:
Fixed calculation `diff time' when setkey -D.
1998/09/10 13:32:42 JST shin@nd.net.fujitsu.co.jp
src/rtadvd Makefile rtadvd.c
src/rtadvd rrenum.c
Supporting receiving rr message. Some message checking was added.
Still message processing need to be added.
1998/09/10 13:30:34 JST shin@nd.net.fujitsu.co.jp
netinet6 icmp6.c
Added message validity check for ICMP6_ROUTER_RENUMBERING message
1998-09-10 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* Set hop limit field for an outgoing packets to the value
advertised by a router, when user did not specify to use
a special value.
1998/09/09 22:31:38 JST shin@nd.net.fujitsu.co.jp
src/faithd faithd.c
Change waitpid() to wait3() to prevent child become zombie.
1998/09/09 15:36:29 JST shin@nd.net.fujitsu.co.jp
src/rrenumd rrenumd.c
Moved(and merged) router renumbering struct definition into
sys/netinet6/icmp6.h
And changes related to it.
1998/09/09 15:34:46 JST shin@nd.net.fujitsu.co.jp
netinet6 icmp6.h
Merged router renumbering struct definition in user-land into kernel.
1998/09/09 14:48:02 JST shin@nd.net.fujitsu.co.jp
i386/conf Makefile.i386
Added 'ulimit -d 65536'
Mon Sep 7 15:55:59 1998 sakane@ydc.co.jp
sys/netkey, kit/src/setkey, kit/src/racoon:
fixed that length in sadb_ext is a multiple of 64 bits.
Mon Sep 7 12:00:49 1998 sakane@ydc.co.jp
sys/netkey/key.c, kit/src/setkey:
change the rule of prefixlen.
key_check() doesn't take care of prefixlen.
you can specify 0.0.0.0/0 as wildcard.
1998/09/07 20:14:50 JST shin@nd.net.fujitsu.co.jp
src/tcpdump print-icmp6.c
Print router renumbering message.(just code type name)
1998/09/07 18:46:58 JST shin@nd.net.fujitsu.co.jp
src/rrenumd Makefile rrenumd.8 rrenumd.c
Initial add of rrenumd. Very ADHOC. Even more, receiving side is notyet.
1998-09-03 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* nd6_rtr.c (rt6_flush): newly defined to purge all the entries
of the routing table when their router is removed from the
Default Router List.
1998-09-03 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/ping6/ping6.c (main): fixed a problem that the -n option
didn't work.
Thu Sep 3 18:52:35 JST 1998 itojun@iijlab.net
* src/tcpdump/print-ip6.c: fixed flowlabel field.
* src/tcpdump/print-vjc.c: fix for NetBSD (anyway it won't be
used)
submitted by: Niels Baggesen <Niels.Baggesen@uni-c.dk>
1998-09-02 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/ndp: -F option can be specified with -r and/or -p option
to flush the default router list and the prefix list,
respectively. Note that both the kernel and the ndp command
should be up to date.
1998-09-02 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* nd6.c (nd6_ioctl): added 2 new ioctls; SIOCSRTRFLUSH_IN6 and
SIOCSPFXFLUSH_IN6. The former is to flush all default routers
in the default router list, and the latter is to flush all the
prefixes and the addresses derived from them in the prefix list.
1998-09-02 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/rtsold/rtsold.c (main): added -m option, which is for
mobile stations.
<199808>
1998-08-28 Atsushi Onoe <onoe@sm.sony.co.jp>
* change getnodeby{host,addr} to getipnodeby{host,addr}
to conform to bsd-api-new-02.
* support host resolver selection (dns, /etc/hosts) for each OS
(/etc/host.conf for FreeBSD, /etc/irs.conf for BSD/OS,
/etc/resolv.conf for NetBSD).
Fri Aug 28 14:39:27 JST 1998 sumikawa@ebina.hitachi.co.jp
* added mld6query command, which sent Multicast Listner
Discovery(draft-ietf-ipngwg-mld-00.txt) query and print
MLD report.
* ip6_input.c: fixed HBH options check bugs.
* ip6_output.c: enbale to join all multicast group(::) for
super user.
Fri Aug 28 13:29:58 JST 1998 sakane@ydc.co.jp
* sys/netkey:
Fixed that reference count about SAD and SPD.
1998-08-27 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/rtsold: Implemented a daemon version of rtsol. Many
imporvements were made such as;
- Auto solicitaion when re-attach to a link
- Source link-layer address option for Router Solicitations
1998-08-27 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* netstat/inet6.c (inet6print): Defined GETSERVBYPORT6, which
calls getservbyport replacing the second argument "tcp6" with
"tcp", "udp6" with "udp", resp.
- Replace getservbyport with the new macro for more friendly
outputs of netstat -f inet6.
1998-08-25 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* icmp6.c (icmp6_reflect): Changed the default way when a returned
ICMP packet exceeds the path MTU; RFC 1885 required to truncate
it, but it was canceled in the new spec.
Tue Aug 25 07:25:24 JST 1998 sakane@ydc.co.jp
* kit/src/setkey
It is available to define hard and soft lifetime extension,
and to display about lifetime.
Mon Aug 24 10:40:21 JST 1998 sakane@ydc.co.jp
* sys/netkey:
Begin to implement handling for SA expiration.
Now, if you define lifetime then SA will be removed on time.
But it is used sadb_lifetime_addtime as expiration time.
1998/08/22 00:41:35 JST shin@nd.net.fujitsu.co.jp
ports/ppp/patches patch-ad
Socks support. Don't use getaddrinfo for socks yet, because
wrap function of getaddrinfo is not supported by usuall socks library.
1998/08/21 19:53:21 JST shin@nd.net.fujitsu.co.jp
Modified files:
ports/ppp/pkg PLIST
Added files:
ports/ppp/patches patch-ab patch-ac
Removed files:
ports/ppp/patches patch-aa
V6 support for ports/ppp.
1998-08-20 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* sbin/ifconfig/ifconfig.c (in6_status): printed a new flag
`detached' for an IPv6 address, when it could be considered as
off-link(see below).
1998-08-20 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
Added some mechanism to make IPv6 router and prefix discovery be
compatible with mobile environment. If we move from one network
to another, this feature detects the fact automatically and treat
the older prefixes(and addresses derived from them) as off-link
even if their lifetime are not expired.
Note that the new feature is not available for FreeBSD 2.2.6.
Thu Aug 20 15:17:17 JST 1998 itojun@iijlab.net
* kit/ports/sendmail6: added a patch supplied by sendmail.org
that avoids duplicated "Content-transfer-encoding".
Wed Aug 19 17:24:20 JST 1998 sakane@ydc.co.jp
* sys/netkey/key.c:
Fixed to call key_kill().
(Thanks, chippo@cdsec.com.)
1998-08-19 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* nd6_rtr.c: Heavily changed to store advertising router list
in prefix list.
* kit/src/ndp/ndp.c (plist): changed to print the router list
stored in each prefix list entry.
Note that both the kernel and ndp command should be recompiled.
Wed Aug 19 16:33:55 JST 1998 sakane@ydc.co.jp
* kit/src/racoon:
racoon has been working on limited environment.
see TODO.
Wed Aug 19 16:30:48 JST 1998 itojun@iijlab.net
* kit/ports/wu-ftpd: redistribute IPv6 patch separately
from ftp://ftp.kame.net/pub/kame/misc, so that non-FreeBSD
users can enjoy wu-ftpd6 patch.
added EPRT/EPSV support.
Wed Aug 19 15:20:27 JST 1998 sakane@ydc.co.jp
* sys/netkey/key.c:
htonl(spi) when doing SADB_GETSPI.
Wed Aug 19 14:14:16 JST 1998 itojun@iijlab.net
* kit/lib/libftpio and kit/usr.bin/fetch:
support EPRT/EPSV (draft-ietf-ftpext-ftp-over-ipv6-02).
Wed Aug 19 12:48:55 JST 1998 itojun@iijlab.net
* kit/usr.bin/ftp and kit/libexec/ftpd:
support EPRT/EPSV (draft-ietf-ftpext-ftp-over-ipv6-02).
caveats:
- proxy mode is not working
- lacks AF negotiation
- ftp tries EPRT/EPSV first, then tries LPRT/LPSV. I dunno
if it is a good thing, or a bad thing.
Wed Aug 19 08:40:15 JST 1998 sakane@ydc.co.jp
* kit/src/racoon: Constants have been compliant with
both ipsec-doi-10 and isakmp-oakley-08.
Tue Aug 18 18:04:08 JST 1998 itojun@iijlab.net
* kit/ports/mrt: udpate base version to 1.4.8a-980728.
* kit/ports/popper: update base version to 2.53.
* kit/ports/fechmail: update base version to 4.5.5.
Also, the ipv6 support for the port was broken
(nobody tried this, it seems). It is now working properly.
1998/08/18 16:05:39 JST shin@nd.net.fujitsu.co.jp
netinet6 ip6_input.c
Removed link local addr ours check.
(To allow multiple link local addr. Also, its not necessary
at least in FreeBSD KAME, because it does more general ours check
by looking up routing table.)
Mon Aug 16 1998 itojun@iijlab.net
* kit and sys: To conform to bsd-api-new-02, "struct sockaddr_in6"
need to be changed. *Remove* sin6_ifindex and Reorder structure
member.
NOTE: you need a FULL recompilation of userland programs!!
Sun Aug 15 1998 itojun@iijlab.net
* kit and sys: changed IPV6_{ADD,DROP}_MEMBERSHIP to
IPV6_{JOIN,LEAVE}_MEMBERSHIP, to conform to bsd-api-new-02.
old def is kept for one week (one snapshot).
Thu Aug 13 16:43:30 JST 1998 itojun@iijlab.net
* kit/libexec/tftpd: address-family independent "tftpd" daemon.
Thu Aug 13 16:22:50 JST 1998 itojun@iijlab.net
* kit/usr.bin/tftp: address-family independent "tftp" command.
Thu Aug 13 11:23:58 JST 1998 itojun@iijlab.net
* kit/ports/sendmail6: port updated to support sendmail 8.9.1a.
port update submitted by: Issei Suzuki <issei@mikage.t-cnet.or.jp>
1998-08-12 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/bgpd: supported several new features;
- BGP routes can be advertised without sync with
RIPng(optional).
- RIPng default route generation
- RIPng route filter
Mon Aug 10 17:04:17 JST 1998 itojun@iijlab.net
* kit/ports/vnc: vnc 3.3.2 with VNC-over-IPv6 capability.
Sat Aug 8 17:27:10 JST 1998 itojun@iijlab.net
* kit/ports/altq, sys: update ALTQ to 1.1.1.
* kit/ports/heimdal: install sample configuration file.
1998-08-07 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* nd6.c (nd6_timer): proper handling for unresolved packets
in INCOMPLETE state; send ICMP6_DST_UNREACH to the originator.
Fri Aug 7 17:33:11 JST 1998 itojun@iijlab.net
* sys/netinet6/nd6_nbr.c: fix handling of anycast address in DAD
routine. previously anycast addresses are marked "tentative"
forever.
Fri Aug 7 17:07:32 JST 1998 itojun@iijlab.net
* kit/src/rtsol: add option "-v" (verbose).
more proper handling of interface, while DAD is performed.
RS retry behavior now conforms to internet-draft.
(try RS until we got an RA, or we've sent MAX_ROUTER_SOLICITATIONS
RS packets)
1998-08-06 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* icmp6.c (icmp6_error): Implemented rate limitation of ICMPv6
error messages. The limitation interval is configurable via sysctl
net.inet6.icmp6.errratelimit in micro seconds(note that 0 means no
limitation). Because of this change, sysctl(for BSDI and NetBSD)
and netstat should be re-installed.
Thu Aug 6 08:08:00 JST 1998 itojun@iijlab.net
* kit/src/tcpdump: flipped the meaning of -R flag. tcpdump now assumes
new AH/ESP packet format by default. (by adding -R it will handle
packets as old AH/ESP)
Wed, 5 Aug JST 1998 sakane@ydc.co.jp
* sys/netkey:
key_bbcmp() has replaced.
(thanks to wdk@squeak.zl2bkc.niwa.cri.nz)
Sun Aug 3 1998 itojun@iijlab.net
* sys/netinet6/esp_core.c: blowfish has a big key schedule struct,
which caused kernel stack overflow on call to esp_blowfish_cbc_xx().
made the key schedule variable a static var.
Sun Aug 2 01:35:59 JST 1998 itojun@iijlab.net
* kit/src/traceroute: LBL traceroute, with IPsec-related modification.
* kit/ports/traceroute: removed (since it is imported into
kit/src/traceroute).
Sat Aug 1 11:59:39 JST 1998 itojun@iijlab.net
* kit/src/traceroute6: disable IPsec in sender side.
"traceroute" protocol is not friendly with IPsec. If you try
a traceroute toward your IPsec peer, you will not be able to
get the expected result.
(we may have to think about enabling IPsec, later)
* kit/ports/traceroute: LBL traceroute, with the same care as above.
Sat Aug 1 10:10:12 JST 1998 itojun@iijlab.net
* kit/ports/ssh: updated to 1.2.26 IPv6 patchlevel 1.2.
ssh kit updated by: kick@wide.ad.jp
port update submitted by: Issei Suzuki <issei@mikage.t-cnet.or.jp>
Sat Aug 1 05:17:43 JST 1998 itojun@iijlab.net
* netkey/key.c: in key_mature(), perform more checks for the incoming
key association. xx_mature() in {ah,esp}_core.c implement algorithm-
dependent checks.
* IPsec: changed some of the kernel internal API for ESP/AH algorithms.
<199807>
1998-07-31 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* icmp6.c (icmp6_redirect_input): Supported received redirects
which redirect to on-link destinations. Some ND related functions
were also modified to accomplish this.
1998/07/30 17:57:15 JST shin@nd.net.fujitsu.co.jp
kit/ports/ppp - Imported sources
current source of user-level ppp command maintained at www.awfulhak.org.
Thu Jul 30 13:08:47 JST 1998 itojun@iijlab.net
* 3DES ivlen has fixed (must be 8)
Thu Jul 30 13:08:47 JST 1998 sakane@ydc.co.jp
* ipsec: support BYPASS policy
Thu Jul 30 11:09:06 JST 1998 itojun@iijlab.net
* ipsec interoperability fixes.
- use more proper ivlen for each of the algorithms
- padding rule fix (mainly for null encryption algorithm)
- loosen payload length check for ESP (must be multiple of 4)
1998/07/29 21:43:11 JST shin@nd.net.fujitsu.co.jp
netinet6 ipsec.c
Changed dport to sport(Though anyway it is inside of #if 0)
Wed Jul 29 21:20:34 JST 1998 sakane@ydc.co.jp
* sys/netinet6/esp_input.c: treat ESP NULL authentication alrogithm
specially. do not perform sequence number check.
Wed Jul 29 17:48:34 JST 1998 itojun@iijlab.net
* sys/netinet6/esp_core.c: 3DES code had a bug that caused SEGV.
(wrong pointer was passed as key buffer)
Wed Jul 29 1998 itojun@iijlab.net
* sys/netinet6/ipsec.c: upper-layer protocol check code in
ipsec[46]_getindex() was wrong. it needs to care about fragmentation
but it does not. commented out for safety. (to be fixed)
Wed Jul 29 1998 itojun@iijlab.net
* kit/src/libpcap: "tcpdump tcp" now works as expected.
Tue Jul 28 19:16:48 JST 1998 k-sugyou@kame.net
* kit/ports/newbie: updated to 0.21p7.
1998/07/28 18:26:26 JST shin@nd.net.fujitsu.co.jp
src/rtadvd rtadvd.conf
Added "addrs#1" for sample configuration which is necessary to
announce a prefix not assigned to the interface.
Tue Jul 28 1998 itojun@iijlab.net
* kit/ports/wu-ftpd: updated to 2.4.2beta18.
* kit/ports/sendmail6: updated to 8.9.1 IPv6 patchlevel 2.
1998-07-28 Atsushi Onoe <onoe@sm.sony.co.jp>
* kit/src/libinet6/name6.c:
re-support sethostent(), endhostent(), gethostent(), and
gethostbyname2 (RES_USE_INET6 stuffs) for compatibility.
h_aliases, h_addr_list of hostent structure must be a pointer
of array, not NULL pointer.
Tue Jul 28 07:09:50 JST 1998 sakane@ydc.co.jp
* sys/netinet6/ipsec.c
Remove the routines of the check to reject
when both the packet encrypted and policy was NONE.
Are they need ?
Tue Jul 28 04:25:48 JST 1998 sakane@ydc.co.jp
* kit/src/setkey
Changed the place in the command to specify the upper layer protocol,
and use `any' token as wildcard.
For examples,
add ::1 ::1 0x100 ...; -> add ::1 ::1 any 0x100 ...;
add ::1 ::1 0x100 tcp ...; -> add ::1 ::1 tcp 0x100 ...;
Sorry, it means that you must modify your configuration for IPsec.
Corrected to specify the IP address, port, prefix and SPI.
Modified the output to dump for the upper layer protocol of wildcard.
Tue Jul 28 02:41:56 JST 1998 sakane@ydc.co.jp
* netinet6/ipsec6.[ch]
Remove use_{in,out}bound_spd, spd_priority and outbound_call_ike.
It causes system to change the IPsec definition.
When SP entry found, IPsec processing will be worked regardless
of def_policy. If you want to do IPsec when no SP entry defined,
be set the def_policy to 2. At this time, if no SA entry found,
IKEd will be called by kernel.
TODO: The part of calling IKEd must be stable.
Thu Jul 23 19:08:57 JST 1998 sakane@ydc.co.jp
* sys/netkey/key.c
Modified the behavior of SADB_ACQIURE and SADB_REGISTER.
It's be possible for kernel to receive SADB_ACQUIRE message
from user land process, and send SADB_ACQUIRE message
to KMd registerd. Also, It's be possible to send algorithms
supported by kernel up to the KMd.
TODO: sadb_supprted should be from {ah,esp}_algorithms.
entry acquired should be free by lifetime.
1998-07-23 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* ip6_input.c (ip6_input,ip6_hopopts_input):
Router Alert Option support.
1998/07/23 07:32:30 JST shin@nd.net.fujitsu.co.jp
netinet6 in6.c nd6_nbr.c
Set Solicited Node Multicast Addr for p2p destaddr only on interface
with same IPv6 prefix.
Also, set proxy=1 when doing proxy NA.
1998/07/22 22:01:51 JST shin@nd.net.fujitsu.co.jp
netinet6 nd6_nbr.c in6.c
Support of Proxy NDP reply.
1998/07/22 19:41:52 JST shin@nd.net.fujitsu.co.jp
netinet6 in6.c
Join and leave Solicited Node Multicast addr for proxy NDPing
PPP dest addr.
And some other small fixes.
1998-07-22 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
Implemented MLD(Multicast Listener Discovery) host part.
1998-07-22 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/libinet6/ip6opt.c: inet6_option_init() and inet6_option_append()
were newly implemented.
Fri Jul 17 03:33:44 JST 1998 itojun@iijlab.net
* kit/src/hroute6d: install RTF_REJECT route for aggregated prefixes,
so that there will be no bogus packet loop at the border routers.
Thu Jul 16 23:11:45 JST 1998 sumikawa@ebina.hitachi.co.jp
* src/ndp, sys/netinet6/nd*
- added -i options.
* Neighbour Discovery
- supported multiple prefixes
- removing unreached router from default router list
- rewrote linked list handlings
- fixed pruning bug
Thu Jul 16 20:14:26 JST 1998 itojun@iijlab.net
* kit/ports/ucd-snmp: SNMP daemon/client. There are very few
IPv6 MIBs implemented at this moment. We'll implement more in
the future.
Thu Jul 16 00:46:47 JST 1998 sakane@ydc.co.jp
* sys/netkey/key.c
Fixed the format for debug message.
1998/07/15 19:36:03 JST shin@nd.net.fujitsu.co.jp
netinet6 in6_var.h
Added some in6_ifaddr operating macros.
And optimized IN6_ARE_MASKED_ADDR_EQUAL().
1998/07/15 19:00:59 JST shin@nd.net.fujitsu.co.jp
src/route6d ifmcstat.c
Only update mc from if6a.ia6_multiaddrs.lh_first when mc != 0.
(because now only top if6a has non zero ia6_multiaddrs.lh_first)
Also, added fe, fxp, cnw to havearpcom(p) list.
1998/07/15 17:59:36 JST shin@nd.net.fujitsu.co.jp
usr.sbin/ppp ip.c ipv6.c
Do SIOCDIFADDR_IN6 for those addresses on which SIOCAIFADDR_IN6 failed,
to remove possible gabage routes.
1998/07/15 16:18:15 JST shin@nd.net.fujitsu.co.jp
netinet6 in6.c
Bug Fix:
When deleting an IPv6 addr, many kinds of miserable things happens
by dual IFAFREE at SIOCAIFADDR_IN6 in in6_control().
1998-07-14 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* in6.c: implemented multicast kludge to save multicast addresses
linked from sockets when deleting all the unicast addresses.
Mon Jul 13 23:47:54 JST 1998 sakane@ydc.co.jp
Fixed the cause of memory leak in ipsec.c. There might be in.
Mon Jul 13 21:19:33 JST 1998 sakane@ydc.co.jp
* kit/src/setkey
Changed how to parse hexadecimal strings.
It's possible to specify a key by hexadecimal string with `0x'.
The quoted strings still too.
1998-07-13 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
Changed the way to manage multicast groups in a socket;
use list of multicast memberships instead of static array
to eliminate limitation of number of multicast groups per socket.
Mon Jul 13 10:27:20 JST 1998 itojun@iijlab.net
* kit/src/pluto: removed pluto IKE daemon. we'll concentrte on
racoon IKE daemon.
1998/07/13 03:38:32 JST shin@nd.net.fujitsu.co.jp
netinet6 icmp6.h in6_proto.c nd6.h nd6.c
Add sysctl variable "nd6_proxyall" to switch on/off Proxy Neighbor
Advertisement.
1998/07/13 03:35:39 JST shin@nd.net.fujitsu.co.jp
netinet6 in6.c
Add a check if IFF_LOOPBACK and/or IFF_LOOPBACK is set, to
in6_is_ifloop_auto().
Also, modified some comments.
1998/07/12 19:06:39 JST shin@nd.net.fujitsu.co.jp
netinet6 nd6_rtr.c
Compare received reachable timer with basereachable timer.
1998/07/12 18:45:39 JST shin@nd.net.fujitsu.co.jp
netinet6 nd6_rtr.c
Don't recompute reachable timer, if received RA's one is 0.
And then, if current reachable timer value is 0(may be by some error)
always recompute the reachable timer.
1998/07/12 18:29:04 JST shin@nd.net.fujitsu.co.jp
netinet6 nd6.h
Bug fix:
Kernel halts by 0 devide at ND_COMPUTE_RTIME(0).
So Changed "random() %" to "random() &". This may not give mathmatically
perfect random value, but I believe it is practically not a problem,
and also performance become better.
1998/07/10 17:22:27 JST shin@nd.net.fujitsu.co.jp
netinet6 in6.c
Bug fix:
Every local address of p2p link is same, so its loopback entry is
deleted when deleteing one of it, and consequently, became not able to
receive packets sent to the local address.
So only delete the p2p local addr loopback entry, when there is only
one of same addresses remains.
Fri Jul 10 19:55:49 JST 1998 itojun@iijlab.net
* kit/ports/heimdal: export control-free kerberos5 implementation.
Please be aware that it is not really tested very much.
Fri Jul 10 1998 itojun@iijlab.net
* kit/ports/vat6: famous videoconferencing system for IPv6.
Please be aware that it is not tested at all, so it is marked
as BROKEN.
1998-07-10 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/ping6/ping6.c (summary): added fflush at the end of the
function.
Submitted by: koba@flab.fujitsu.co.jp
1998-07-10 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* ip6_input.c (ip6_input): removed a quick hack for link-local
addresses on point-to-point interfaces. We believe there is no
problem.
1998-07-10 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* in6_proto.c: changed the default value of icmp6_rediraccept
from 0 to 1.
KAME kernel now safely handles received redirects if they redirect
to a better router. We'll soon support the redirects to on-link
destinations.
1998-07-09 Atsushi Onoe <onoe@sm.sony.co.jp>
* kit/src/libinet6/name6.c:
fixed handling IPv4 literal address and /etc/hosts.
IPv4-mapped address support is commented out until kernel will
accept it.
1998-07-08 Atsushi Onoe <onoe@sm.sony.co.jp>
* kit/src/libinet6/name6.c, include/netdb.h:
support getnodebyname(), getnodebyaddr(), freehostent().
old interfaces (ex. gethostbyname2) are left for compatibility.
Wed Jul 8 02:19:06 JST 1998 itojun@iijlab.net
* sys/netinet6/in6_ifattach.c: do not make interfaces IFF_UP on
assignment of link-local unicast address.
* sys/net/if.c: perform IPv6 DAD on interface state changes,
i.e. ioctl().
* kit/src/rtsol: make the interface up if it is not.
try to send RS repeatedly, if the interface is under DAD state.
1998/07/08 00:48:53 JST shin@nd.net.fujitsu.co.jp
net if_tun.c
Stop to up unused tun* interfaces at default. They are up'ed after they
are configured and began to used.(and address is assigned)
(suggested by itojun)
Mon Jul 6 12:51:52 JST 1998 itojun@iijlab.net
* kit/ports/sendmail6: update to sendmail 8.9.1+IPv6 patchlevel 1.
1998/07/02 23:54:19 JST shin@nd.net.fujitsu.co.jp
usr.sbin/ppp Makefile ipcp.c ipv6cp.c
Added compile option for connectiong with remote sppp.
1998/07/02 17:04:10 JST shin@nd.net.fujitsu.co.jp
netinet6 nd6.c
netinet6 in6_ifattach.c
When many continuous p2p interface exist, specified ifp's if_index
in nd6_ifattach() become so large that nd_ifinfo[if_index] could
point over the nd_ifinfo[] area, and break arbitrary memory value.
When if_index value is big, multiply new nd_ifinfo size several time
until the size become enough for specified ifp's if_index.
Thu Jul 2 16:50:13 JST 1998 itojun@iijlab.net
* kit/ports/sendmail6: update to sendmail 8.9.0+IPv6 patchlevel 4.
Thu Jul 2 12:46:08 JST 1998 itojun@iijlab.net
* kit/ports/popper: upgraded to use qpopper 2.52.
1998-07-01 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* icmp6.c (icmp6_rip6_input): Removed IPV6_STRIPHDR socket option
and related code.
* raw_ip6.c (rip6_input): ditto.
* src/icmp6dump,ping6, rtadvd, tcpdump6: ditto.
1998-07-01 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/bgpd: supported metric addition for incoming RIPng routes.
Wed Jul 1 16:30:48 JST 1998 itojun@iijlab.net
* kit/src/route6d: add -T, -N options (advertise default route only,
and restrict interfaces to handle)
Submitted by: kick@wide.ad.jp
Wed Jul 1 13:47:37 JST 1998 itojun@iijlab.net
* everywhere: ALTQ (alternate queueing framework) merged into
KAME/FreeBSD kit. (not for BSDI nor NetBSD)
Good things: It makes ATM driver more concrete, and you can play
with ALTQ.
Things to be fixed: IPv6-over-ATM support has to be checked.
Userland tool (apconfig) is now unusable due to difference in
the ATM driver. Userland tools must be installed via
kit/ports/altq.
For more information about ALTQ, visit the following URL:
http://www.csl.sony.co.jp/person/kjc/software.html
1998-07-01 Atsushi Onoe <onoe@sm.sony.co.jp>
* sys/net/if{,_mib}.c: fix index value of ifnet_addrs[] to 1 origin,
for sysctl net.link.generic.ifdata.<index> to work.
1998-07-01 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* ports/popper: Supported the qpopper2.5, in which buffer overrun
bugs were fixed.
Wed Jul 1 06:35:39 JST 1998 itojun@iijlab.net
* kit/libexec/rshd: bugfix on INET6 connection.
Submitted by: ume@calm.imasy.org
* kit/src/tcpdump: check AF field on CISCO HDLC packet.
Submitted from: fujiwara@rcac.tdi.co.jp
Wed Jul 1 01:06:08 JST 1998 itojun@iijlab.net
* sys/dev/en/midway*: moved "midway*" files from sys/pci to
sys/dev/en, to help ALTQ merge.
<199806>
1998-06-30 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/bgpd:
- supported synchronization of RIPng and BGP4+
- small bug fix of route aggregation
Tue Jun 30 13:35:15 JST 1998 itojun@iijlab.net
* sysctl MIB for net.inet6.ip6.xxx has changed. (previous def was bad)
past:
mib[0] = CTL_NET; mib[1] = PF_INET6; mib[2] = IPPROTO_IP;
now:
mib[0] = CTL_NET; mib[1] = PF_INET6; mib[2] = IPPROTO_IPV6;
1998-06-30 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/route6d/route6d.c (ripalarm): when aging routes,
check whether rrt list is empty or not after aging. If it
lacks, route6d may be core dumped.
1998/06/29 19:58:02 JST shin@nd.net.fujitsu.co.jp
usr.sbin/ppp command.c ipv6.c ipv6cp.c
net rtsock.c
netinet6 in6.c
Made it possible to assigne IPv6 link local addr to kernel, when
kernel have none.
Please update with kernel.
Mon Jun 29 13:15:05 JST 1998 itojun@iijlab.net
* sys/netinet6/{ah,esp}_core.c: renamed {ah,esp}.c into
{ah,esp}_core.c, to avoid file name collision in NetBSD kernel
source tree (with Sun ethernet driver).
Mon Jun 29 02:20:01 JST 1998 itojun@iijlab.net
* kit/ports/sendmail6: update to sendmail 8.9.0+IPv6 patchlevel 3.
Sun Jun 28 10:48:30 JST 1998 itojun@iijlab.net
* kit/ports/sendmail6: update to sendmail 8.9.0+IPv6 patchlevel 1.
Fri Jun 27 1998 itojun@iijlab.net
* sys: Kernel now compiles with "options MROUTING". (I have never
tried IPv4 multicast code with KAME kernel, though)
TODO: more cleaner way to deal with MBone stuff with gif interface.
Fri Jun 26 23:06:54 JST 1998 itojun@iijlab.net
* sys: TCP MSS computation takes care of the room occupied by
transport-mode IPsec (ESP/AH). This helps TCP to be performed
without packet fragmentation.
The functionality was implemented in WIDE Hydrangea days, and
removed by PFKEY-v2 merge. It is now become available again.
Fri Jun 26 20:44:23 JST 1998 itojun@iijlab.net
* sys/netinet/ip_output.c: If IPsec'ed packet is too big for the
interface, turn DF bit off and fragment it. This is the last
resort for TCP packet with DF bit set. THIS IS A QUICKHACK!
This applies only for non-AH, and non-forwarding (packet from myself)
case. In AH case, we can't modify IP header (DF bit).
In forwarding case, ICMP more fragment message should be sent to
the originating node.
TODO: more consideration to path MTU discovery, as gateway and as
originator/final destination
Fri Jun 26 15:55:12 JST 1998 itojun@iijlab.net
* sys/netinet6/esp_*.c: Some cleanup for ESP tunnel mode.
Now KAME host is able to serve as ESP tunnel gateway.
TODO: AH tunnel
TODO: better relationship with gif tunnel
1998-06-25 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* raw_ip6.c (rip6_input): striped all the extension headers
including IP6 header for a socket on which IN6P_STRIPHDR
was set. This feature will soon be default, not optional.
Thu Jun 25 02:23:06 JST 1998 itojun@iijlab.net
* kit/src/libinet6/getaddrinfo.c: cleanup hostname == NULL case.
If AI_PASSIVE is specified in "hint" structure, return anyaddr
(0.0.0.0 or ::). If not, return loopback address (127.0.0.1 or ::1).
Wed Jun 24 16:30:00 JST 1998 sumikawa@ebina.hitachi.co.jp
* src/ndp, sys/netinet6/nd*
- added -r, -p and -F options.
1998-06-24 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/rtadvd:
- Implemented dynamic prefix addition and deletion by watching
the routing table.
- Now rtadvd is completely independent of the configuration file.
It works even if the config file does not exist.
Wed Jun 24 13:39:59 JST 1998 sakane@ydc.co.jp
sys/netkey/, sys/netinet{,6}/
* Fixed the behavior when the SPD is refered.
* Modified
* Adjusted ipsecstat's members.
kit/src/setkey/
* Modified debug mode and verbose mode.
* Fixed man page and sample script in sample.cf.
1998-06-23 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/bgpd:
- fixed many bugs including infinite loop.
- route aggregation can be specified per interface base.
Tue Jun 23 18:25:12 JST 1998 itojun@iijlab.net
* kit/ports/socks64: updated socks64 patch from flab.
Now socks64 patch is fetchable, so you can build it yourself.
1998-06-23 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* icmp6.h (ICMP6_ROUTER_RENUMBERING): newly defined. Because of
this, the definition of FQDN query and reply was changed. So
you may have to update both kernel and userland applications.
1998-06-23 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* usr.sbin/portmap/from_local.c (find_local): rewrote the function
using sysctl interface. The older versions had the too strict
limitation on number of interface addresses. If you find some
trouble about /usr/sbin/portmap, we recommend you to replace it with
/usr/local/v6/sbin/portmap.
Tue Jun 23 02:58:40 JST 1998 itojun@iijlab.net
* kit/ports/v6tun: upgrade to 1.1.
* kit/ports/ssh: upgrade to 1.2.23 + IPv6 patch 1.1.
Submitted by: kick@wide.ad.jp
Tue Jun 23 1998 itojun@iijlab.net
* sys/netinet6/raw_ip6.c: upper-layer checksum was fixed. In the past
by invoking rtadvd kernel hangs up, due to stale pointers.
1998-06-18 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/rtadvd:
- Added RA consistency check.
- Added received RS/RA validation
- Implemented timer related functions.
- Implemented rate limitation of sending RAs.
- maxinterval and mininterval were configurable.
Thu Jun 18 04:36:01 JST 1998 sumikawa@ebina.hitachi.co.jp
* kit/ports/im: Email and NetNews user interface
Thu Jun 18 04:36:43 JST 1998 sakane@ydc.co.jp
sys/netkey/, sys/netinet{,6}/, kit/src/setkey/
* Restructed both SPD and SAD in kernel IPsec module.
There are two SPD. One is the based socket, and the another is
in the system mainly to handle the packet forwarding.
* Some systemwide IPsec's parameters has been replaced into ipsec.h,
and add new parameters to handle the SPD's functions.
* Add some commands for SPD management to `setkey'.
* Add a module for acquiring KEY, but not tested.
* Fixed when a PF_KEY socket was opened without PF_KEY_V2 as
its protocol.
Suggested by <suzuki@grelot.elec.ryukoku.ac.jp>.
Tue Jun 16 21:50:40 JST 1998 k-sugyou@kame.net
ports/mozilla
update to use mozilla-19980603
1998/06/16 13:32:45 JST shin@nd.net.fujitsu.co.jp
usr.sbin/ppp command.c filter.c ipv6cp.c
Log:
Several small fixes.
BCOPY->memcpy. Removed some func's obsolete arg. If_id initialization
for server.
1998/06/15 21:33:31 JST shin@nd.net.fujitsu.co.jp
net if_sppp.h if_spppsubr.c
IPV6CP of PPP support for sppp.
But only communication over serial cable is tested using slsppp driver
from fujiwara san. (At leaset, ping6 to remote link local addr is OK.)
Also, token negotiation is not supported yet.
1998/06/14 00:42:36 JST shin@nd.net.fujitsu.co.jp
netinet tcp_usrreq.c
netinet6 tcp6_usrreq.c udp6_usrreq.c
MAPPED ADDR support phase 3.
Now mapped addr is available for tcp.
1998-06-12 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/rtadvd/config.c (getconfig): Modified to use protocol
default values for unspecified parameters.
Now users may omit almost all parameters in rtadvd.conf if
they want to use protocol defaults. Users are also able to
specify more than 2 prefixes on an interface.
* src/rtadvd/if.c: Added to include source link layer address
option in RAs.
1998-06-12 Atsushi Onoe <onoe@sm.sony.co.jp>
* kit/src/libinet6/getaddrinfo.c, getnameinfo.c:
fix getaddrinfo() and getnameinfo() to use offsetof() to determine
offset of address in struct sockaddr_in{,6}.
(offsetof(struct sockaddr_in6, sin6_addr) has been changed
from 8 to 16).
Wed Jun 10 19:43:03 JST 1998 itojun@iijlab.net
* kit/src: To share kit/src among FreeBSD/BSDI/NetBSD kit, some changes
are added. All source code under kit/src must be portable across
these three platforms.
Submitted by: hiroshi ura
1998-06-10 Atsushi Onoe <onoe@sm.sony.co.jp>
* include/netdb.h: include <sys/types.h> to define size_t
for struct addrinfo.
1998-06-09 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* in6.c (in6_addmulti): added the 3rd argument to return
more specific errors to caller.
* in6.c (in6_addmulti):
(in6_control):
* in6_ifattach.c (in6_ifattach):
* ip6_output.c (ip6_setmoptions):
Reflected above change and used the error number if possible.
1998-06-09 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* esp_input.c (esp6_input):
* ah_input.c (ah6_input):
Set next header field of the previous header of ESP or AH
using ip6_get_prevhdr().
* icmp6.c (icmp6_error): modified to send ICMP error for packets
including unknown header. Older versions discarded them, but
spec required to return error for such packets.
* raw_ip6.c (rip6_input): called icmp6_error with
ICMP6_PARAMPROB_NEXTHEADER when there is no listening socket for
the incoming packet.
* frag6.c (frag6_input): replaced a block to store new next
header value with calling ip6_get_prevhdr.
* ip6_input.c (ip6_get_prevhdr): was newly defined. The function
is not so efficent so it will be not used when we provide a
better way.
Tue Jun 9 21:32:47 JST 1998 itojun@itojun.org
* sys/netinet6/ipsec.c: fixed a critical bug when compiling kernel
with IPSEC option. No data was coming to IPv6 sockets.
Suggested by: Atsushi Onoe <onoe@sm.sony.co.jp>
1998-06-09 Atsushi Onoe <onoe@sm.sony.co.jp>
* kit/src/route6d: ignore interfaces not marked IFF_UP at
initialization, to avoid failure in IPV6_ADD_MEMBERSHIP.
Tue Jun 9 13:43:33 JST 1998 itojun
* kit/ports/v6tun: a handy tool for establishing IPv6 tunnel over
IPv4 dialup/ssh session.
Mon Jun 8 05:48:54 JST 1998 itojun@itojun.org
* ipsec: ESP payload authentication algorithm is now switchable
via PF-KEY v2. (more checks need to be done)
* ipsec: Changed SPI handling method in kernel ipsec module.
SPI is now always kept in scalar variables in network byteorder.
This is to support new netkey module.
Sun Jun 7 JST 1998 itojun@itojun.org
* sys/crypto: Fixed SHA-1 checksum length
(wrong: 16bytes correct: 20bytes)
Pointed out by: Francis Dupont
Sun Jun 7 14:17:52 JST 1998 itojun@itojun.org
* Incorporate REVISED FreeBSD TTCP security hole fix (FreeBSD security
advisory 98:03).
Pointed out by: kenji rikitake
Sun Jun 7 JST 1998 sakane@ydc.co.jp
* ipsec code is now start working, but may need some more time to
stabilize.
Sat Jun 6 JST 1998 itojun@itojun.org
* kit/ports/mrt: Multi-threaded Routing Toolkit from Merit.
* sys/i386/isa/if_ed.c: location to call in6_ifattach() is fixed.
Submitted by: Kouki Higuchi <luke@fct.kgc.co.jp>
* sys: ipsecstat is now visible via sysctl MIB.
* kit/usr.bin/netstat: show ipsec statistics on netstat -s.
Sat Jun 6 JST 1998 sakane@ydc.co.jp
* sys: pfkey v2 and SPD is merged into.
NOTE: This breaks IPsec code, for the moment.
Also, some of userland tools (for example kit/src/pluto) cannot
be compiled.
* kit/src/setkey: revamp for pfkey v2.
NOTE: this is not compatible with past version.
Fri Jun 5 JST 1998 itojun@itojun.org
* kit/ports/zebra: update to use zebra-980604.
* kit/ports/socks64: modified socks5 from Fujitsu Laboratories Ltd.
(not compile-able at this moment, since socks64 patch is yet to be
released)
Thu Jun 4 JST 1998 itojun@itojun.org
* kit/src/libinet6: getaddrinfo() and getnameinfo() are cleaned up.
Thu Jun 4 00:27:42 JST 1998 itojun@itojun.org
* Endian-ness of outbound (kernel->userland) sin6_ifindex has fixed.
We temporarily use host endian as draft is yet to be released.
1998/06/03 23:57:26 JST shin@nd.net.fujitsu.co.jp
kern uipc_socket.c
netinet udp_usrreq.c
netinet6 udp6_usrreq.c in6.h in6.c in6_proto.c
ip6_var.h
Mapped addr support phase2. Now UDP is aple to use mapped addr
by doing below.
sysctl -w net.inet6.ip6.mapped_addr=1
Wed Jun 3 01:44:50 JST 1998 itojun@itojun.org
* sys/i386/isa/if_zp.c: multicast support added. it is essential
for using IPv6 on zp interface.
1998/06/02 00:20:29 JST shin@nd.net.fujitsu.co.jp
netinet6 udp6_usrreq.c
PRU_DISCONNECT was denied when in6p_faddr is not UNSPEC addr.
Fixed it to be denied when in6p_faddr is UNSPEC addr.
Also, changed the macro to BSD API conformed one.
1998/06/01 14:05:16 JST shin@nd.net.fujitsu.co.jp
src/faith faith.c
Changed the option argument spec of faith command.
1998/06/01 12:49:31 JST shin@nd.net.fujitsu.co.jp
src/libinet6 getaddrinfo.c
Changed the way of specifing faith prefix for getaddrinfo().
Now it is specified in environmental variable GAI.
<199805>
Sat May 30 10:42:32 JST 1998 itojun@itojun.org
* kit/ports/ssh: ssh port updated to use ssh 1.2.23, and IPv6 patchkit
supplied by KIKUCHI Takahiro <kick@rcac.tdi.co.jp> (Thanks!).
sshd and ssh are now totally AF independent, and listens both v4 and
v6 sockets!
1998/05/29 21:16:00 JST shin@nd.net.fujitsu.co.jp
i386/conf GENERIC.v6 PAO.v6.diff
Added compile option to switch mapped addr usage.
1998/05/29 21:09:10 JST shin@nd.net.fujitsu.co.jp
kern uipc_socket.c
Step1 for mapped addr support.(actual support is not yet)
Added so_pcb2 to struct socket, and let inet6 domain use it.
Other domains(inet, unix, etc) continue to use so_pcb.
1998/05/29 14:06:46 JST shin@nd.net.fujitsu.co.jp
netinet6 in6.h in6_proto.c
Addition of a sysctl for switching mapped addr support. (actual mapped
addr support in kernel is not yet)
Also changed some part of sysctl definition in in6.h, for IPSEC defined
sysctls and MAPPED_ADDR_ENABLED defined sysctls could easily coexist.
1998/05/28 22:30:33 JST shin@nd.net.fujitsu.co.jp
ports/mozilla/patches patch-bg
Patch for af switching hack's bug fix and the other.(from sugyou)
Now mozilla can access v4 and v6 pages concurrently.
Ftp access is also available.
1998/05/28 19:52:51 JST shin@nd.net.fujitsu.co.jp
libexec/ftpd ftpd.c
Fixed ftpd's ftp data connection from 21 to 20.
Thu May 28 20:11:08 JST 1998 itojun@itojun.org
* sin6_ifindex support for (most of) connectionless inbound sockets
and raw ip6 outbound packets.
kit/src/ping6 now uses sin6_ifindex to control outgoing interface,
as a trial case.
May 27-28, 1998 itojun@itojun.org
* cast128 bugfixes.
By Tomomi Suzuki <suzuki@grelot.elec.ryukoku.ac.jp>.
May 27-28, 1998 itojun@itojun.org
* getaddrinfo() fix for PF_UNSPEC case. Return IPv4/IPv6 anyaddr
(0.0.0.0 or ::) or localaddr (127.0.0.1 or ::1) in AI_PASSVE cases.
Suggested by KIKUCHI Takahiro <kick@rcac.tdi.co.jp>.
Tue May 26 18:41:38 JST 1998 itojun@itojun.org
enable sockaddr_in6 new members: sin6_scope_id and sin6_ifindex.
NOTE!!! please make sure make sure that all userland code, libraries
and port binaries are based on new definition.
You may want to perform the following before making userland code,
to make it sure:
% rm /usr/local/v6/lib/*
% find kit -name '*.o' -print | perl -nle unlink
% find kit -name '*.a' -print | perl -nle unlink
Tue May 26 04:05:02 JST 1998 itojun@itojun.org
* preparation for sockaddr_in6 change.
kernel: bzero() for sockaddr_in6 are clarified.
userland: updated getaddrinfo() and getnameinfo() to make them
independent of the definition of sockaddr_in6.
* IPV6_CHECKSUM setsockopt (advanced api)
* inet6_rthdr_foobaa(): routing header manipulation functions
(advanced api)
Mon May 25 01:15:50 JST 1998 itojun@itojun.org
* RC5 cbc mode added. (thanks goes to tomomi suzuki@ryukoku-u)
1998/05/22 16:06:51 JST shin@nd.net.fujitsu.co.jp
ports/mozilla
now mozilla can only access IPv6 www home pages.
Wed May 20 16:12:13 JST 1998 itojun@itojun.org
* kit/src/*: Makefile cleanups.
1998-05-20 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* ipv6_icmp.c (icmp6_rip6_input):
* raw_ip6.c (rip6_input):
Maked sure to 0-clear the pointer opts after each iteration
of for loop. Older kernel causes kernel crash when more than 1
raw IP6 sockets exist and optional information is expected
on each socket.
Tue May 19 22:23:03 JST 1998 itojun@itojun.org
* "options DES" is obsoleted. You just need to say "options IPSEC".
Tue May 19 17:33:00 1998 sumikawa@ebina.hitachi.co.jp
* kit/hroute6d: changed name from route6d to avoid duplication.
1998-05-19 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/v6test/v6test.c (main):
Added -f option to specify the path of configuration file.
Document file was also modified.
shin 1998/05/18 21:21:32 JST
src/route6d route6d.c
* Added ageout of each entry to ripalarm(), which is called every
SUPPLY_INTERVAL6 interval
* Added one condition to the check if a route is loopback route or
not
Sun May 17 12:08:54 JST 1998 itojun@itojun.org
* sys/netinet6/nd6.c: implemented NDP upper-layer reachability hint.
Fri May 15 12:06:03 JST 1998 itojun@itojun.org
* TTCP security hole fix merged in.
* ATM related fix. Now we can configure kernel with ATM support.
(could experienced person try this?)
* kit/ports/mozilla: inet6-capable mozilla. You'll need Motif
as well as some CPU power for compilation. (some more work
needs to be done)
Fri May 14 1998 itojun@itojun.org
* some cleanups on NDP (neighbor discovery protocol).
Wed May 13 18:22:09 JST 1998 itojun@itojun.org
* kit/ports/popper: inet6-capable qpopper. Use with inet6d.
* kit/ports/fetchmail: inet6-capable fetchmail.
shin 1998/05/13 17:55:59 JST
kit/usr.sbin/ppp
modifications to use SOCKS5.
shin 1998/05/13 16:02:21 JST
kit/src/faithd
Changed to refer sysctl configured value for FAITH prefix.
Tue May 12 12:38:21 JST 1998 itojun@itojun.org
* kit/ports/newbie: updated to 0.21p3.
* kit/ports/XFree86: imported a easy-installation for IPv6-ready
XFree86 (IPv6 patch by inria forks)
* sys/netinet6/icmp6.c: fixed false badcode counter increments.
1998-05-12 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
Imported toshiba-origin bgpd, which supports RIPng and
BGP4+. Thanks to Toru Kon'no <toru@isl.rdc.toshiba.co.jp>
Mon May 11 01:59:54 JST 1998 itojun@itojun.org
* kit/src/tcpdump: imported LBL's tcpdump for more portability.
kit/usr.sbin/tcpdump (FreeBSD origin) was removed.
* kit/src/libpcap: imported LBL's tcpdump for more portability.
kit/src/libpcap6{,.src} (FreeBSD origin) was removed.
Sat May 9 00:30:34 JST 1998 itojun@itojun.org
* kit/src/hroute6d: hitachi-origin route6d, with route filtering/
aggregation functionalities. (quickhack-ed version imported)
Fri May 8 14:37:00 JST 1998 itojun@itojun.org
* sys/netinet6: sanity checks for ifindex (passed via msghdr) added.
Wed May 6 18:43:36 JST 1998 itojun@itojun.org
* kit/ports/newbie added. This is an IPv6-aware nameserver/resolver.
Wed May 6 15:43:22 JST 1998 itojun@itojun.org
* Makefile fixes so that "kit" can be made without modifying
/usr/src/sys.
shin 1998/05/05 01:49:58 JST
kit/src/faith
Added command for setting FAITH prefix value.
shin 1998/05/04 20:20:13 JST
sys/netinet6/in6.h in6_proto.c ip6_input.c ip6_var.h
Enabled FAITH prefix value change from sysctl().
itojun 5/5
snapshot name changed from "hydrangea" to "kame"
1998-05-01 Atsushi Onoe <onoe@sm.sony.co.jp>
* kit/usr.bin/rsh/rsh.c
Search "rlogin" from $PATH to use IPv6 capable rlogin command.
1998-05-01 Atsushi Onoe <onoe@sm.sony.co.jp>
* kit/src/libinet6/rcmd.c
Call freeaddrinfo() in rcmd() to fix memory leak.
Try ai_next even if ECONNREFUSED case, not to wait 31 seconds
when only rshd4 is available on the remote host.
<199804>
1998-04-28 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* frag6.c (frag6_input):
Changed the way of deletion of the fragment header after
reassembling; use ovbcopy instead of copying each 4 byte
in a while loop.
shin 1998/04/27 11:39:32 JST
kit/src/resolv/res_init.c
Bug fix.
Bus error happens at res_init()
when nameserver directive is at the end of the resolv.conf file
without "\n" terminated.
1998-04-20 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* dest6.c (dest6_input): Fixed a bug of returned value; from 0 to
the next header identifier.
1998-04-17 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* in6_cksum.c (in6_cksum):
Changed the definition of the pseudo header union not to include
IP6 source and destination addresses.
Changed the way to calculate the sum of IP6 pseudo header
not to copy IP6 source and destination addresses to a separate
buffer.
1998-04-10 Atsushi Onoe <onoe@sm.sony.co.jp>
* kit/usr.bin/telnet/commands.c
As original telnet, try the next available address
if connect() fails.
Thu Apr 2 23:08:07 JST 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
sys/netinet6/in6.c
Added solicited node multicast address deletion at if addr deletion time.
And modified comment.
1998-04-02 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* src/v6test/getconfig.c (make_frghdr): Changed not to use HTONS for
IP6F_MORE_FRAG, since byte order consideration for the flag
is now in its definition.
* src/v6test/ext.conf:
Added a hop by hop option.
Changed the configuration for too big fragment offset test case2;
use the correct IPv6 payload length value.
* frag6.c (frag6_input): Specification conformance.
- Changed the check when searching a reassemble queue for an
incoming fragment; don't take into account the next header field
of the fragment header. It was necessary for IPv4 fragments, but
is not required for IPv6 fragments any more.
- Changed to use unfragmentable part length and the next header
value of the fragment header in the 1st(offsset = 0) fragment.
The 1st fragment is not necessarily the '1st-reached' fragment.
- Added a check if the reassembled packet would not exceed 65535
bytes.
--> Now Hydrangea passes sumikawa & jinmei's devilish tests
about fragment headers:-)
* icmp6.c (icmp6_error): Removed the receiving interface index
from the errornous packet if it had link-local scope source or
destination.
* icmp6.h (ICMP6_TIME_EXCEED_REASSEMBLY, ICMP6_TIME_EXCEED_TRANSIT):
The latest advanced API conformance; EXCEEDED->EXCEED.
* src/ping6/ping6.c (pr_icmph):
* src/traceroute6/traceroute6.c (packet_ok):
* usr.bin/tcpdump/tcpdump/print-icmp6.c (icmp6_print):
* src/icmp6dump/icmp6dump.c (dump):
* icmp6.c (icmp6_input):
* ip6_forward.c (ip6_forward):
Reflected the above changes.
* frag6.c (frag6_freef): Added a bunch of code to return ICMP error for
the 1st fragment.
* ip6_var.h: added ip6q_unfrglen member to ip6q{}.
Thu Apr 2 17:34:50 JST 1998 Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
sys/net/if_tun.c
Removed SIOCSIFPHYADDR (which I added before) from tunifioctl because
it turned out not to be necessary for tun interface.
1998-04-01 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* ip6_var.h: ip6q_nxtp member was escaped by 'notyet.' The
location of the member caused segmentation faults.
* ip6_input.c (ip6_input): put 'return' outside of the while loop
at the end of the function. It was a very careless mistake, sorry.
Wed Apr 1 02:53:52 JST 1998 itojun
- ESP 3DES support merged in. (thanks goes to tomomi suzuki@ryukoku-u)
<199803>
1998/03/31 20:24:37 JST Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
usr.sbin Makefile Makefile.inc
Added ppp to SUBDIR.
Moved 'NOMAN= yes' from kit/Makefile.inc to portmap/Makefile.
1998/03/31 20:22:50 JST Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
usr.sbin/portmap Makefile
Moved 'NOMAN= yes' from ../Makefile.inc to ./Makefile.
1998/03/31 20:17:23 JST Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>p
usr.sbin/ppp
Several bug fixes, improvements and small changes.
-Re added 'set ifaddrinet6' command, because it is useful
when assigning p2p style addr with dest addr on ppp links.
-Added `set v6prefix` command to define default IPv6 prefix.
-Added address macros MYV6PREFIX and HISV6PREFIX. This needs
prefix value defined by `set v6prefix`.
-Enabled deletion of addrs assigned by ppp at link down time.
-Added japanese man.
1998/03/31 11:31:42 JST Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
netinet6/in6.c
When an IPv6 addr is removed, also remove the ownaddr loopback rtentry
if it exists. (an addtion to the change of 1998/03/31 00:38:33)
1998/03/31 00:38:33 JST Yoshinobu Inoue <shin@nd.net.fujitsu.co.jp>
netinet6/in6.c nd6.c nd6.h nd6_rtr.c
P2p interface support extension. (Because, hydrangea requires rtentry
for ownaddr to be set to loopback.)
-added nd6_p2p_rtrequest.(set ownaddr rtentry to loopback)
-added a function which set ownaddr rtentry to loopback, in the case
nd6_p2p_rtrequest doesn't work.
Enabled the place where ln->ln_state is set to ND6_LLINFO_REACHABLE
when rt->rt_expire == 0, in nd6_rtrequest.
This seems to be necessary for setting permanent ndp entry.
1998-03-28 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* route6.c (ip6_rthdr0): Conformance with new spec drafts; remove check
if ip6r0_len <= 46.
* icmp6.c (icmp6_input): when delivering ICMPv6 error to upper
layers, take into account of intermediate extension headers.
XXX: To accomplish this, the prototype of ctlinput functions is
modified. It's inconsistent with the prototype of generic ctlinput
functions.
* udp6_usrreq.c (udp6_ctlinput):
if there are extension headers between ip6 header and udp header,
skip them.
* tcp6_subr.c (tcp6_ctlinput):
if there are extension headers between ip6 header and tcp header,
skip them.
* nd6_nbr.c (nd6_ns_input): if the target address is an anycast,
clear override flag when calling nd6_na_output().
(nd6_na_output): remove a line where override flag is
automatically set. The flag is now set in nd6_ns_input if necessary.
* tcp6_input.c (tcp6_input): if the packet is anycasted, drop it.
(syn_cache_add6): if the packet is anycasted, it's not cached.
* icmp6.c (icmp6_reflect):
add check if the destination of the original packet is an anycast
when choosing the source of the reflected packet.
* in6.c (in6_ifawithscope): add check if each address is an
anycast. If so, it's not selected.
* ip6_input.c (ip6_input): set M_ANYCAST6 flag to the mbuf header
when receiving packets with anycast destinations.
* nd6.c (nd6_rtrequest): for loopbacked packets, set rt_ifa to
the in6_ifaddr which has the same address of the destination.
This hack is for receiving anycast addresses.
* sbin/ifconfig/ifconfig.c
support 'anycast' parameter.
You can configure anycast address by ifconfig with this parameter.
e.g. ifconfig ep0 3ffe:: alias anycast
Update of the kernel is also required.
* mbuf.h (M_ANYCAST6): added the flag to set to incoming anycasted
packets.
1998-03-27 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* in6_var.h:
anycast support:
add ifra_flags member to in6_aliasreq{}.
add ia6_flags member to in6_ifaddr{}.
define IN6_IFF_ANYCAST flag.
* in6.c (in6_control):
in SIOCAIFADDR_IN6 case, set ia6_flags from in6_aliasreq.
* in6_proto.c: add sysctl interface for ip6_hdrnestlimit.
* usr.bin/netstat/inet6.c (ip6_stats):
print # of dropped packets due to too many extension headers.
* ip6_input.c (ip6_input):
introduce 'ip6_hdrnestlimit', which controls the upper limit of
number of extension headers. default value is set in in6_proto.c.
* in6_pcb.c (in6_pcbnotify):
change type of the 4th argument from in6_addr to 'in6_addr *',
which is more efficient.
* tcp6_subr.c (tcp6_ctlinput):
* udp6_usrreq.c (udp6_ctlinput):
reflect the above change when calling in6_pcbnotify().
* icmp6.c (icmp6_error): remove garbage.
Thu Mar 26 13:51:48 1998 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* icmp6.c (icmp6_error):
improve the algorithm to check if the original packet is an ICMP6
error or not; do IP6_EXTHDR_CHECK for each intermediate extension header,
so the check will be rarely failed.
Thu Mar 26 13:47:36 1998 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* ip6_input.c (ip6_input):
change algorithm of processing IPv6 header and Hop-by-Hop header.
1st check if the packet is destined to ourself, then process
hop-by-hop options header(if any). So if we are not a router
and the packet are not destined to ourself, HBH headers aren't processed.
Thu Mar 26 16:40:05 JST 1998 itojun
* bugfix to route6d: multicast address overwritten by mistake
Wed Mar 25 1998 itojun
* kernel/ip6_output.c: multicast group management ADVAPI fix
(forgot to embed ifindex into the multicast address)
now rtadvd should work correctly
Tue Mar 24 00:16:06 JST 1998 shin <shin@nd.net.fujitsu.co.jp>
* usr.sbin/ppp: several bug fix, man page add, and a little
enhancement(save interface id to file).
Mon Mar 23 11:10:14 JST 1998
* ports/apache: famous http server, with a patch for inet/inet6 capability.
Sun Mar 22 10:42:49 JST 1998
* cast128 support (from suzuki@ryukoku-u)
* ah bug fix (new-ah/old-ah header size)
Sun Mar 22 09:14:08 1998 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* ip6_input.c (ip6_input):
use upper layer protocol identifier for faith check instead of
the next header value of IP6 header.
Fri Mar 20 08:13:49 1998 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* usr.bin/tcpdump/tcpdump/print-icmp6.c (icmp6_print):
ICMPv6 error parameters are 'ntohl'ed.
Fri Mar 20 08:13:49 1998 JINMEI, Tatuya <jinmei@isl.rdc.toshiba.co.jp>
* usr.bin/tcpdump/tcpdump/print-icmp6.c (icmp6_print):
ICMPv6 error parameters are 'ntohl'ed.
* sys: Bunch of advanced-api modifications
Mon Mar 09 22:56:02 JST
* sys/net/if_tun: IPv6 support
Thu Mar 05 16:20:31 JST 1998 shin <shin@nd.net.fujitsu.co.jp>
* usr.sbin/ppp: IPv6 support
<199802>
* sys: discovery-v2 code is now default, "options INET6_NDV2" is not needed
any more.
<19980101-19980111>
* Happy new year!
* src/rtadvd: bug fix, was generated garbled packet
(enbug'ed by advanced-api-04 related changes)
* sbin/{route,ifconfig}: bug fix
was not able to handle explicit prefixlen less than 64
* usr.bin/netstat: -s fixed
* now we are able to grab media addr for adaptec 5940 ATM card.
* ip6 linklocal address assignment improved.
- p2p/tunnel interfaces now borrow interface id from 1st ethernet card
* use of int{8,16,32}m_t are gone
Jump to Line
Something went wrong with that request. Please try again.