summaryrefslogtreecommitdiff
path: root/sys/netinet
AgeCommit message (Collapse)Author
1999-04-23dont accept packets with the destination address of a down interface;Niels Provos
proff@netbsd.org.
1999-04-22Heh, this was definitely not tested. cpp sytax errors.Niklas Hallqvist
1999-04-21From Tom Henderson <tomh@cs.berkeley.edu>:Niels Provos
Fixed a sequence wraparound bug in the snd_recover variable discovered in very large (multiple GByte) transfers (in loss free conditions, snd_recover was not sufficiently tracking snd_una). Thanks to Mark Smith for finding this. Fixed a bug in tcp_newreno that was preventing retransmission of data due to partial acks. (Discovered by Jayanth Vijayaraghavan)
1999-04-20messup, learn to test *all* variants of compile options whenNiklas Hallqvist
altering the logic round such.
1999-04-20use open() with O_EXCL.. but also move to /var/run to avoid the 2nd-level ↵Theo de Raadt
/tmp race which darren obviously does not think important
1999-04-20Merge MROUTING and IPSEC wrt handling of IP-in-IP tunnelled packets.Niklas Hallqvist
Fix a panic case in the MROUTING code too. Drop M_TUNNEL support, nothing ever uses it.
1999-04-16fix ipf return-rst panic bug; beckTheo de Raadt
1999-04-12move encdebug to a useful placeTheo de Raadt
1999-04-11Introduce net.inet.{ah,esp}.enable sysctl controls that are off by default.Niklas Hallqvist
If you are going to use either of AH or ESP or both, enable these in /etc/sysctl.conf. Also correct the IPSec debugging sysctl code, it is now named net.inet.ip.encdebug. Some corrected function signatures too.
1999-04-09The kernel parts of a sysctl that can switch on/off IP-in-IP (protocol 4)Niklas Hallqvist
support, when IPSEC is compiled in. The default is disabled. Turn on with: sysctl -w net.inet.ip4.allow=1 ***Only*** do this if you are really knowing what you do! This control does not control the tunnel modes of ESP and AH.
1999-04-09Make the tdbi handling more robust, removes a panic caseNiklas Hallqvist
1999-04-09Check for local address spoofing on encapsulated packets.Angelos D. Keromytis
1999-04-04fix tunnelling; provosTheo de Raadt
1999-03-27add SADB_X_BINDSA to pfkey allowing incoming SAs to refer to an outgoingNiels Provos
SA to be used, use this SA in ip_output if available. allow mobile road warriors for bind SAs with wildcard dst and src addresses. check IPSEC AUTH and ESP level when receiving packets, drop them if protection is insufficient. add stats to show dropped packets because of insufficient IPSEC protection. -- phew. this was all done in canada. dugsong and linh provided the ride and company.
1999-03-24Implement lifetime expiration notifications. Fix some typos. Remove statics.Niklas Hallqvist
1999-03-24Reworked udp_output() to minimize the number of if() statements needed to getcmetz
packets out. Also had the nice side effect of fewer blocks now move around by ifdefs, which makes it more readable.
1999-03-24Removed inclusion of netinet6/in6.h. This was an artifact of when the corecmetz
IPv6 symbols were there rather than in netinet/in.h, and now not only is unnecessary but also could create problems (see PR library/781).
1999-03-24Removed old NRL convention in6a_words virtual-field from struct in6_addr.cmetz
Let's not encourage the use of an obsolete convention.
1999-03-24Replace 'in6a_words' (old NRL convention) with 's6_addr32' (new BSDI et al.cmetz
convention that is more common and more specific as to the access size)
1999-03-13indentTheo de Raadt
1999-03-06Update IP pointer, when doing multiple transforms.Angelos D. Keromytis
1999-03-04be more careful with freeing TDBsTheo de Raadt
1999-03-04more paranoid maintenanceTheo de Raadt
1999-02-25Prettier reporting, report number of flows in use per SA.Angelos D. Keromytis
1999-02-25Skipjack key bzero'ing case.Angelos D. Keromytis
1999-02-25Print more information in /kern/ipsecAngelos D. Keromytis
1999-02-25Bzero key information before free'ing it, just for paranoia's sake.Angelos D. Keromytis
1999-02-25Use M_XDATA for the key schedule, instead of M_TEMP.Angelos D. Keromytis
1999-02-25Clear the DF bit, so packets don't get dropped inside a tunnel.Angelos D. Keromytis
The real solution is probably to keep soft state; however, it's not entirely clear what a tunnel's properties with regards to fragmentation are (it may be considered a infinitely-large MTU pipe).
1999-02-25Be a bit more paranoid before free'ing memory.Angelos D. Keromytis
1999-02-25Move union sockaddr_union to ip_ipsp.hAngelos D. Keromytis
1999-02-24Update copyright; remove a few annoying debugging printfs.Angelos D. Keromytis
Btw, OpenBSD hit 25000 commits a couple commits ago.
1999-02-24add skipjack support backTheo de Raadt
1999-02-24Remove encap.h include; saner debugging printfs; fix buglets; work withAngelos D. Keromytis
pfkeyv2.
1999-02-24Added in.h definitions required by IPv6 Basic API and Advanced API specs, andcmetz
a few extras that are just plain useful. Note that I used u_intxx_t instead of the POSIX p1003.1g uintxx_t that those specs mandate, so as to not increase the number of outside symbol definitions that in.h depends on.
1999-02-23Forgot this.Angelos D. Keromytis
1999-02-23This one seems to work, at least in the direction Sparc->i386.Angelos D. Keromytis
Btw, for all the curious souls, I am in Canada at the moment (land of free commits).
1999-02-23Turns out the patch didn't fix the problem. Reverting to old code.Angelos D. Keromytis
1999-02-23Fix an endianness problem (not an issue in libc).Angelos D. Keromytis
1999-02-21split ipintr() to create new ipv4_input() for tunnels; NRLTheo de Raadt
1999-02-19fix definition of plen for ICMP (ping -s73 problem)Kjell Wooding
1999-02-19ipq lockingTheo de Raadt
1999-02-19when checking for weak keys compare sizeof(des_cblock) bytes insetad of ↵Artur Grabowski
sizeof(key) bytes
1999-02-17add fragment flood protection; configureable using sysctl ip.maxqueueTheo de Raadt
1999-02-17ipsec skipjack, based on free .fi code (some .gov type will test this for me)Theo de Raadt
1999-02-17indentTheo de Raadt
1999-02-17indent; remove local typesTheo de Raadt
1999-02-17inet6 indentTheo de Raadt
1999-02-16Fix a padding check bug in the experimental-padding. iliya@ericom.comAngelos D. Keromytis
1999-02-15when allocating sack blocks check for failing malloc(), if it failsNiels Provos
just ignore the sack block + fix misleading comment. tomh@CS.Berkeley.EDU