Skip to content


Subversion checkout URL

You can clone with
Download ZIP
Commits on Dec 15, 2008
  1. This main goals of this project are:

    qingli committed
    1. separating L2 tables (ARP, NDP) from the L3 routing tables
    2. removing as much locking dependencies among these layers as
       possible to allow for some parallelism in the search operations
    3. simplify the logic in the routing code,
    The most notable end result is the obsolescent of the route
    cloning (RTF_CLONING) concept, which translated into code reduction
    in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
    struct rtentry{}. The change in design obsoletes the semantics of
    RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
    applications such as "arp" and "ndp" have been modified to reflect
    those changes. The output from "netstat -r" shows only the routing
    Quite a few developers have contributed to this project in the
    past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
    Andre Oppermann. And most recently:
    - Kip Macy revised the locking code completely, thus completing
      the last piece of the puzzle, Kip has also been conducting
      active functional testing
    - Sam Leffler has helped me improving/refactoring the code, and
      provided valuable reviews
    - Julian Elischer setup the perforce tree for me and has helped
      me maintaining that branch before the svn conversion
  2. @kientzle

    Spell "substitution" correctly.

    kientzle committed
    PR:		bin/128616
    Submitted by:	Jaakko Heinonen
    Pointy hat:	Tim Kientzle
    MFC after:	30 days
  3. back out unintended change

    sam committed
  4. Fix definition of IEEE80211_CHAN_MAX; it was defined as 255 but

    sam committed
    really was meant to be 256.  Adjust usage accordingly and replace
    bogus usage of this value in checking IEEE channel #'s.
    NB: this causes an ABI change; ifconfig must be recompiled
  5. o fix DEBUG sku definition

    sam committed
    o add definitions for 900MHz radios: Ubiquiti SR9 and XR9, and ZComax GZ-901
  6. move channel parsing to a getchannel routine so it can be reused to c…

    sam committed
    the channel argument supplied to chanswitch
  7. fix handling of unknown country codes; atoi doesn't return -1

    sam committed
    for an invalid string as I thought; so use strtol instead
  8. 0 is a potential ISO CC; use new NO_COUNTRY #define to identify

    sam committed
    when the CC is not set.  Note NO_COUNTRY is set to 0xffff for now
    (must be 16 bits as ieee80211_regdomain struct defines sku's and
    cc's as uint16_t which may need fixing).
  9. o distinguish between adhoc and ahdemo modes

    sam committed
    o do not require 1/2 and 1/4 rate channels be present in the
      calibration list when doing a gsm regulatory change; the
      existing 900MHz cards are not self-identifying so there is
      no way (using the calibration channel list) to check
  10. Improve regdomain.xml parser:

    sam committed
    o store XML_Parser in the state block so we can report line numbers for errors
    o complain about netband w/o mode
    o complain about unknown modes
    o complain about band w/o enclosing netband
    o complain about duplicate freqband
    o complain about unknown channel flags
    o complain about band w/o freqband's
    o complain about band w/o maxpower
    o complain about country w/o ISO cc
    o complain about country w/o regdomain reference
  11. Replace adhoc checks in ieee80211_start with a per-node flag that

    sam committed
    indicates if an association id is required before outbound traffic
    is permitted.  This cleans up the previous change that broke mcast
    traffic "to the stack" in ap mode as a side effect.
    Reviewed by:	sephe, thompsa, weongyo
  12. fix ini setup

    sam committed
    Submitted by:	Jiri Fojtasek <>
Commits on Dec 14, 2008
  1. @bsdphk
  2. Add __packed to the part sub-structure to compensate for

    marcel committed
    the 32-bit default structure alignment on ARM.
  3. Add an ENVIRONMENT section to note that the RSH variable can be set to

    murray committed
    override the use of the rsh(1) program in rcmd(3).
    PR:		docs/51133
    Obtained from:	NetBSD, partially
    MFC after:	3 weeks
  4. Combine ath rate control modules with ath after bringing ath_hal

    sam committed
    src into the tree.  The old split was balanced on module dependencies
    and symbol exposure that no longer exists.  Users that want a module
    setup with rate control algorithm other than sample must override
    ATH_RATE in the ath module Makefile.
    Reviewed by:	imp
  5. @amotin

    To avoid one doubtless netgraph SMP scalability limitation point, switch

    amotin committed
    node queues processing from single swi:net thread to several specialized
    Reviewed by:	julian
    Tested with:	Netperf Cluster
  6. @bsdphk
  7. @bsdphk

    Move the code that injects received characters into the tty system into

    bsdphk committed
    a separate public function ucomrxchars(), to avoid requirement of
    simple metadata prefixing on the USB data stream.
  8. Merge the resolver part of BIND 9.4.3 into HEAD.

    ume committed
    It includes the following fix:
      2426.   [bug]           libbind: inet_net_pton() can sometimes return the
                              wrong value if excessively large netmasks are
                              supplied. [RT #18512]
    Reported by:	Maksymilian Arciemowicz <>
  9. prop change.

    ume committed
  10. delete svn:keywords.

    ume committed
  11. prop change

    ume committed
  12. Add a check, that is currently under discussion for 8 but that we need

    bz committed
    to keep for 7-STABLE when MFCing in_pcbladdr() to not change the
    behaviour there.
    With this a destination route via a loopback interface is treated as
    a valid and reachable thing for IPv4 source address selection, even
    though nothing of that network is ever directly reachable, but it is
    more like a blackhole route.
    With this the source address will be selected and IPsec can grab the
    packets before we would discard them at a later point, encapsulate them
    and send them out from a different tunnel endpoint IP.
    Discussed on:	net
    Reported by:	Frank Behrens <>
    Tested by:	Frank Behrens <>
    MFC after:	4 weeks (just so that I get the mail)
  13. Make sure that the direct jls invocations prints something

    bz committed
    reasonable close to and in the same format as it had always.
    r185435 said it would try that but I had been living with jail
    patches for too long to actually remember the single-line format
    when adding backwards compatibility back in p4.
    Reported by:	Philipp Wuensche <>
    Tested by:	Philipp Wuensche <>
    MFC after:	4 weeks (just for me to get the mail)
  14. Bootstrapping merge history for resolver.

    ume committed
  15. Fix the fallouts from r146267:

    ru committed
    - Add the forgotten "mode" argument to the "mode" command.
    - Move the description of "info" to where it belongs.
  16. @bsdphk

    Send all debug to stderr.

    bsdphk committed
  17. @bsdphk
  18. @jkoshy
  19. Clean out the remaining alpha-isms.

    kensmith committed
Commits on Dec 13, 2008
  1. @thompsa

    Dont leak the rnh lock on error.

    thompsa committed
  2. @amotin

    Revert rev. 183277:

    amotin committed
    Remove ng_rmnode_flags() function.
    ng_rmnode_self() was made to be called only while having node locked.
    When node is properly locked, any function call sent to it will always be
    queued. So turning ng_rmnode_self() into the ng_rmnode_flags() is not just
    meaningless, but incorrent, as it violates node locking when called outside.
    No objections:	julian, thompsa
Something went wrong with that request. Please try again.