Commit Graph

100926 Commits

Author SHA1 Message Date
Sam Leffler
6f3f5a1170 always free node reference when processing a probe request frame; was
conditionally doing an unref if operating in adhoc mode which turns
out to be wrong
2004-04-02 23:12:45 +00:00
Sam Leffler
694dca643b o change ieee80211_dup_bss to inherit explicit data from ic_bss instead of
blindy copying the node contents; this turns out to be a bad idea as we
  add more state in the node for things like WPA
o track node allocation failures in ieee80211_dup_bss instead of the callers

Obtained from:	madwifi
2004-04-02 23:09:24 +00:00
Sam Leffler
c64bfa0f8f track node allocation failure stats in ieee80211_alloc_node instead
of each caller

Obtained from:	madwifi
2004-04-02 23:06:41 +00:00
Sam Leffler
410ca74bb8 replace explicit malloc/free with MALLOC/FREE for portability
Obtained from:	madwifi
2004-04-02 23:02:24 +00:00
Sam Leffler
b4c5a90fee extract node matching logic into new ieee80211_match_bss routine for use
in background/incremental scanning

Obtained from:	madwifi
2004-04-02 23:00:30 +00:00
Sam Leffler
849b898054 diff reduction against madwifi 2004-04-02 22:56:09 +00:00
Sam Leffler
a6381c54bb when processing beacon/probe response frames capture "is probe" in a
variable to improve code legibility since we're going to use it more
in forthcoming mods

Obtained from:	madwifi
2004-04-02 22:54:07 +00:00
Sam Leffler
4bd067c5a1 add a case of IEEE80211_ELEMID_IBSSPARMS so the statistic on unknown
elements is more meaningful (will add more with WPA merge)

Obtained from:	madwifi
2004-04-02 22:50:31 +00:00
Sam Leffler
7aa402898f when doing internal bridging free the node instead of just dropping the
reference for consistency (explicit refcounting is discouraged)

Obtained from:	madwifi
2004-04-02 22:48:52 +00:00
Sam Leffler
9089c48c13 when receiving a frame w/ a bad version number don't unref the node; the
driver is responsible for that

Obtained from:	madwifi
2004-04-02 22:47:11 +00:00
Sam Leffler
fce2da8b66 insure basic rate bit is set in derived rate set; this
works around a protocol/firmware problem with some versions found
in hermes/prism cards
2004-04-02 20:22:25 +00:00
Sam Leffler
63beab8306 diff reduction against madwifi/p4 2004-04-02 20:19:20 +00:00
Max Laier
6cd9ebedd7 Add rc.d script to start pflogd and add rcvars etc. Also document vars in
rc.conf(5) and put a sample entry to newsyslog.conf

Reviewed by:	-current
Approved by:	bms(mentor)
2004-04-02 19:25:27 +00:00
Vinod Kashyap
87b320bb72 Moved 3ware 9000 driver (twa) stuff from sys/conf/NOTES to /sys/i386/conf/NOTES. 2004-04-02 18:50:56 +00:00
Daniel Eischen
ab39bc9a92 Unbreak natd.
Reported and submitted by:	Sean McNeil (sean at mcneil.com)
2004-04-02 17:57:57 +00:00
Roman Kurakin
67de8ecbf8 Add Cronyx-Tau entry. 2004-04-02 17:35:53 +00:00
Roman Kurakin
8135893765 Add Cronyx-Tau entry. 2004-04-02 17:19:59 +00:00
Dag-Erling Smørgrav
c959700c77 style(9): return foo -> return (foo)
also fix a continuation indent I missed in the previous commit.
2004-04-02 16:41:16 +00:00
Dag-Erling Smørgrav
81f58729a1 Clean up whitespace, fix continuation indents, wrap some long lines. 2004-04-02 16:39:12 +00:00
Dag-Erling Smørgrav
4cdc9a643e Unbreak LINT on 64-bit platforms. Note that this code is not style(9)-
compliant, but I'll leave that for someone else.

Noticed by:	tinderbox
Pointy hat to:	the usual suspects
2004-04-02 15:09:57 +00:00
Ken Smith
34b678a695 Rearrangements needed for syscons(4) to be used as a console device
on architectures that need to call cninit() before the machine is
ready to support mutexes (required by make_dev()).

	- Remove make_dev() call from scinit() when flags indicate
	  unit is the system console, rely on sc_attach_unit() to
	  handle it.
	- When trying to access current screen's status (scr_stat
	  structure) use the static one provided for the initial
	  system console if no dev_t is available.
	- When calling make_dev() in sc_attach_unit() catch special
	  case of system's initial console and set up dev_t structure
	  to include pointer to console's scr_stat struct.

Reviewed by:	marcel
Tested by:	marcel, grehan (ppc), others on current@
Approved by:	rwatson (mentor)
2004-04-02 15:02:44 +00:00
Kirill Ponomarev
0a25d7a9ea - Use "/bin/hostname" explicitly instead of "hostname".
Approved by:	tobez
MFC after:	1 week
2004-04-02 12:18:40 +00:00
Marcel Moolenaar
4e55f7230a In ns8250_putc() insert a barrier between writing the character and
checking for transmitter empty.
2004-04-02 07:37:28 +00:00
Marcel Moolenaar
16283d130f Allow the selection of a debug port with hw.uart.dbgport. Unlike
other architectures (like ia64), the variable has to be set to
an OpenFirmware device name.
2004-04-02 07:33:35 +00:00
Marcel Moolenaar
af81ff3f49 Call kbd_attach() only when KBD_INSTALL_CDEV is enabled as the function
is only defined in that case.
2004-04-02 05:59:06 +00:00
Kris Kennaway
c5af600675 Add missing comment terminator. 2004-04-02 04:57:40 +00:00
Julian Elischer
4f73277a35 The comment complained about not having a thread_unlink()
and did the work itself, but thread_unink() has existed for a while... use it.
2004-04-02 01:01:34 +00:00
Robert Watson
445a8f0348 For now, restore an splx(s) I removed when introducing slisunitfree(). 2004-04-01 23:54:49 +00:00
Andrey A. Chernov
f853699a55 Simplify one condition in prev. commit:
short_too already assumes FLAG_LONGONLY
2004-04-01 22:32:28 +00:00
Andrey A. Chernov
ed4fbbd5e3 Fix parsing of ambiguous options, whole loop must be processed 2004-04-01 22:09:07 +00:00
Alfred Perlstein
e5ff40f0ca Fix booting with ps2 keyboards. 2004-04-01 21:48:31 +00:00
John Baldwin
e43257aa7d Finish fixing up Alpha to work with an MP safe ptrace():
- ptrace_single_step() is no longer called with the proc lock held, so
  don't try to unlock it and then relock it.
- Push Giant down into proc_rwmem() instead of forcing all the consumers
  (including Alpha breakpoint support) to explicitly wrap calls to
  proc_rwmem() with Giant.

Tested by:	kensmith
2004-04-01 20:56:44 +00:00
Johan Karlsson
72673aaaa2 Remove these MAINTAINER lines since the maintainers has had their
commit bits retired for safe keeping.
2004-04-01 20:31:49 +00:00
Kirill Ponomarev
c6e88cbf92 - Use absolute path for hostname(1)
Approved by:	green
2004-04-01 19:28:00 +00:00
Jacques Vidrine
8074e24dce The previous commit changed the behavior of nsdispatch() in the
case where an /etc/nsswitch.conf file was present, but could not
be opened (e.g. due to permissions).  Previously, the open failure
condition was suppressed, and the built-in defaults were used.  In
revision 1.11, however, propagated the open failure causing all
nsdispatch() invocations to return NS_UNAVAIL, and thus many APIs
including getpwnam and gethostbyname unconditionally failed.

This commit restores the previous behavior.

Pointy hat:	nectar  (+1 for obstinance; ache had to use clue bat)
Reported by:	ache
2004-04-01 19:12:45 +00:00
Julian Elischer
10030054ac Do the looping retry trick in the first operation to try to talk
with the device, not the second..

Submitted by: ticso@cicely12.cicely.de
2004-04-01 18:55:28 +00:00
Dag-Erling Smørgrav
9a02a2f5ed Fix style issues in twa lines added in rev 1.878. 2004-04-01 17:55:50 +00:00
Dag-Erling Smørgrav
7cfcb873b5 Comment out the au88x0 line which was inadvertantly included in the
previous commit.
2004-04-01 17:53:04 +00:00
David Malone
de06a3982c If the user doesn't specify a device and we can't open any of the defaults,
don't print an error using a uninitialised devbuf.
2004-04-01 16:37:46 +00:00
Yoshihiro Takahashi
a8c6f0ac9c The twa device and related options are not needed. 2004-04-01 14:23:41 +00:00
SUZUKI Shinsuke
b5676acff4 UDP checksum is mandatory in IPv6 (RFC2460 p.28)
Obtained from: KAME
2004-04-01 13:48:23 +00:00
Garance A Drosehn
396e20d19f Add spaces after flag names on .Fl macros.
Noticed by:	ru
MFC after:	4 days
2004-04-01 13:28:36 +00:00
Ruslan Ermilov
ba7bc837d0 Moved the statistical counter under hw.ste.rxsyncs.
Suggested by:	njl
2004-04-01 12:55:38 +00:00
Dag-Erling Smørgrav
7eb17244fa Move twa from files.i386 to files. This unbreaks LINT on !i386.
Pointy hat to:	vkashyap, ps
2004-04-01 10:02:50 +00:00
Dag-Erling Smørgrav
387a06e15b Back out 1.188 (major number for twa). The twa driver does not need it.
Pointy hat to:	ps
2004-04-01 10:00:04 +00:00
Ruslan Ermilov
f48cbcdecc Assorted mdoc(7) fixes. 2004-04-01 07:54:27 +00:00
Mike Silbersack
bdf861859a Fix last commit to conform to mdoc style.
Submitted by:	hmp
2004-04-01 07:36:56 +00:00
Peter Grehan
6611d6692f Match the specific MPC106 host bridge PCI ID rather than all
generic host bridges: this avoids a race with the UniNorth
generic match.
2004-04-01 07:34:36 +00:00
Scott Long
cd587b1397 Don't print out 'GIANT-LOCKED' for INTR_FAST drivers. 2004-04-01 07:18:42 +00:00
Mike Silbersack
a61bc764ce Document the m_defrag function. (Mostly copied from the description
in uipc_mbuf.c)
2004-04-01 06:50:21 +00:00