Commit Graph

116671 Commits

Author SHA1 Message Date
Pawel Jakub Dawidek
7be67fe3d1 Reimplement ggatec/ggated applications.
Change communication protocol to be much more resistant on network
problems and to allow for much better performance.

Better performance is achieved by creating two connections between
ggatec and ggated one for sending the data and one for receiving it.
Every connection is handled by separeted thread, so there is no more
synchronous data flow (send and wait for response), now one threads
sends all requests and another receives the data.

Use two threads in ggatec(8):
- sendtd, which takes I/O requests from the kernel and sends them to the
  ggated daemon on the other end;
- recvtd, which waits for ggated responses and forwards them to the kernel.

Use three threads in ggated(8):
- recvtd, which waits for I/O requests and puts them onto incoming queue;
- disktd, which takes requests from the incoming queue, does disk operations
  and puts finished requests onto outgoing queue;
- sendtd, which takes finished requests from the outgoing queue and sends
  responses back to ggatec.

Because there were major changes in communication protocol, there is no
backward compatibility, from now on, both client and server has to run
on 5.x or 6.x (or at least ggated should be from the same FreeBSD version
on which ggatec is running).

For Gbit networks some buffers need to be increased. I use those settings:
kern.ipc.maxsockbuf=16777216
net.inet.tcp.sendspace=8388608
net.inet.tcp.recvspace=8388608
and I use '-S 4194304 -R 4194304' options for both, ggatec and ggated.

Approved by:	re (scottl)
2005-07-08 21:28:26 +00:00
Pawel Jakub Dawidek
84436f14c4 Add CANCEL command which allows to remove one request from the queue or
all requests from the queue if request number is not given.

Bump version number.

Approved by:	re (scottl)
2005-07-08 21:08:53 +00:00
Damien Bergamini
bd6875fb78 oops ... forgot to declare sc in my previous commit. unbreak build.
Approved by:	re (scottl)
2005-07-08 19:33:42 +00:00
Damien Bergamini
69e2ffe63e fix allocation of rx dma buffers. buffers must be contiguous.
Approved by:	re (scottl)
2005-07-08 19:30:29 +00:00
Sam Leffler
e7af1f56e9 document ENABLE_WPA_SUPPLICANT_EAPOL
Approved by:	re (scottl)
2005-07-08 19:26:55 +00:00
Damien Bergamini
df00c70d2e Make sure that all the necessary USB endpoints are found during device
attachment.

Reviewed by:	silby (mentor)
Approved by:	re (scottl)
2005-07-08 19:19:06 +00:00
Damien Bergamini
11c70aecee ral_init() takes a struct ral_softc as parameter not a struct ifnet.
Fixes a panic that occured when setting the interface parameters while
the interface was associated.

Approved by:	re (scottl)
2005-07-08 19:13:49 +00:00
Roman Kurakin
fbb7165a4b Use implicit type cast for ->k_lock to fix compilation of ndis
as a part of the GENERIC kernel with INVARIANT* and WITNESS*
turned off.
(For non GENERIC kernel KTR and MUTEX_PROFILING should be also
off).

Submitted by:	Eygene A. Ryabinkin <rea at rea dot mbslab dot kiae dot ru>
Approved by:	re (scottl)
PR:		81767
2005-07-08 18:36:59 +00:00
Sam Leffler
c52dab628a fix another instance of the MORE_DATA bit handling for frames on the
power save queue (missed in previous commit)

Submitted by:	Bruno Randolf
Approved by:	re (scottl)
2005-07-08 16:36:02 +00:00
Sam Leffler
5ca4a14b24 handle RTM_IFINFO events so we terminate when an interface is
marked down

Reviewed by:	brooks, avatar
Approved by:	re (scottl)
2005-07-08 16:33:52 +00:00
Sam Leffler
0a5ae7ccb8 reclaim node reference when ieee80211_encap fails
Reviewed by:	avatar
Approved by:	re (scottl)
2005-07-08 16:31:19 +00:00
Sam Leffler
025afcf2d5 update ural ids
Reviewed by:	damien
Approved by:	re (scottl)
Obtained from:	openbsd
2005-07-08 16:28:24 +00:00
John Baldwin
522ccb2381 Regen.
Approved by:	re (scottl)
2005-07-08 15:06:58 +00:00
John Baldwin
4acd2e73e5 Mark second instance of lchown() MP safe just like the first.
Approved by:	re (scottl)
2005-07-08 15:01:13 +00:00
Ken Smith
5adb21a681 Missed one piece of the cluster's quirk. Need to override WARNS because
if _FREEFALL_CONFIG is set gcc bails since pam_sm_setcred() in pam_krb5.c
no longer uses any of its parameters.

Pointy hat:	kensmith
Approved by:	re (scottl)
2005-07-08 14:53:45 +00:00
Denis Peplin
774812956c Merge the following from the English version:
1.879 -> 1.880 new.sgml

Obtained from:  The FreeBSD Russian Documentation Project
Approved by:    re (scottl)
2005-07-08 14:34:10 +00:00
Max Laier
ade6e491fc ALTQify ste(4).
Requested and tested by:	<nike_d at cytexbg dot com>
Approved by:			re (scottl)
2005-07-08 13:05:59 +00:00
David E. O'Brien
43c1498efa This commit was generated by cvs2svn to compensate for changes in r147824,
which included commits to RCS files with non-trunk default branches.
2005-07-07 22:06:54 +00:00
David E. O'Brien
362c19dcb0 Fix bogus fmov* SPARC opcodes.
(Note this makes the vendor branch not represent Binutils in the vendor's
CVS repository at any point in time.  Portmgr did not like the state of
Binutils on Sparc that represented the point in time the vendor fixed this
issue.  I'd rather have fixed this on RELENG_6.)

Approved by:	re
2005-07-07 22:06:54 +00:00
Scott Long
98e104cbc9 Patch the freebsd spec file for PowerPC so that PIC code builds correctly.
This allows FreeBSD/PPC to build and run out of stock CVS sources.  This
also takes the file off of the vendor branch.

Submitted by: kan, grehan
Approved by: re, kan
2005-07-07 21:40:37 +00:00
John Baldwin
a3b324f003 Lock Giant around a call to userland_sysctl() in ibcs2_sysi86().
Approved by:	re (scottl)
2005-07-07 19:30:30 +00:00
John Baldwin
e9ccc53c63 Remove an extra call to read() in ibcs2_read() that can never be reached.
(It's already called a few lines earlier.)

Approved by:	re (scottl)
2005-07-07 19:29:42 +00:00
John Baldwin
92df1b8f6f Conditionally acquire Giant around the ISC and Xenix system calls based
on the SYF_MPSAFE flag in the wrapper system calls for the ISC and Xenix
system call vectors.

Approved by:	re (scottl)
2005-07-07 19:28:55 +00:00
John Baldwin
55522478e6 Lock Giant in svr4_add_socket() so that the various svr4_*stat() calls
can be marked MP safe as this is the only part of them that is not
already MP safe.

Approved by:	re (scottl)
2005-07-07 19:27:29 +00:00
John Baldwin
03badf38ab Remove an unused syscallarg() macro leftover from this code's origins in
NetBSD.

Approved by:	re (scottl)
2005-07-07 19:26:43 +00:00
John Baldwin
07fac65b15 Rototill this file so that it actually compiles. It doesn't do anything
in the build still due to some #undef's in svr4.h, but if you hack around
that and add some missing entries to syscalls.master, then this file will
now compile.  The changes involved proc -> thread, using FreeBSD syscall
names instead of NetBSD, and axeing syscallarg() and retval arguments.

Approved by:	re (scottl)
2005-07-07 19:25:47 +00:00
John Baldwin
8d948cd1ec Fix the computation of uptime for linux_sysinfo(). Before it was returning
the uptime in seconds mod 60 which wasn't very useful.

Approved by:	re (scottl)
2005-07-07 19:17:55 +00:00
John Baldwin
a146c8c620 - Remove spl's from osf1_usleep_thread().
- Fix the calculation of the 'slept' timeval returned by
  osf1_usleep_thread() which has been broken since 1.1.

Approved by:	re (scottl)
2005-07-07 19:16:30 +00:00
John Baldwin
9f3157a254 Regenerate.
Approved by:	re (scottl)
2005-07-07 18:20:38 +00:00
John Baldwin
bcd9e0dd20 - Add two new system calls: preadv() and pwritev() which are like readv()
and writev() except that they take an additional offset argument and do
  not change the current file position.  In SAT speak:
  preadv:readv::pread:read and pwritev:writev::pwrite:write.
- Try to reduce code duplication some by merging most of the old
  kern_foov() and dofilefoo() functions into new dofilefoo() functions
  that are called by kern_foov() and kern_pfoov().  The non-v functions
  now all generate a simple uio on the stack from the passed in arguments
  and then call kern_foov().  For example, read() now just builds a uio and
  calls kern_readv() and pwrite() just builds a uio and calls kern_pwritev().

PR:		kern/80362
Submitted by:	Marc Olzheim marcolz at stack dot nl (1)
Approved by:	re (scottl)
MFC after:	1 week
2005-07-07 18:17:55 +00:00
Xin LI
0e3e87bd43 Fix a bug when shell expansion is done against dangling symlinks, by
converting the stat() call to a lstat() call, which will cover the
situation.  One can exercise this bug by referring a dangling link with
something like */the-link.

Approved by:	re (scottl)
Submitted by:	Simon 'corecode' Schubert [corecode fs ei tum de]
Obtained from:	NetBSD via DragonFlyBSD (NetBSD rev. 1.51 and DragonFly
		rev. 1.6)
MFC After:	3 days
2005-07-07 18:10:33 +00:00
Joe Marcus Clarke
a617a18a23 Fix ptsname(3) by converting it to use devname(3) to obtain the name of
a tty device instead of the legacy minor number approach.  This is known to
fix gnome-vfs' sftp module as well as kio_sftp and kdesu on -CURRENT.

Thanks to scottl for the snprintf() approach idea.

Reviewed by:	phk
Tested by:	pav
		mich
Approved by:	re (scottl)
2005-07-07 17:48:40 +00:00
Ken Smith
2672e71736 This is sort of an MFS. Peter made these changes to the RELENG_*
branches but missed HEAD.  This patch extends his a little bit,
setting it up via the Makefiles so that adding _FREEFALL_CONFIG
to /etc/make.conf is the only thing needed to cluster-ize things
(current setup also requires overriding CFLAGS).

From Peter's commit to the RELENG_* branches:
> Add the freebsd.org custer's source modifications under #ifdefs to aid
> keeping things in sync.  For ksu:
> * install suid-root by default
> * don't fall back to asking for a unix password (ie: be pure kerberos)
> * allow custom user instances for things like www and not just root

The Makefile tweaks will be MFC-ed, the rest is already done.

MFC after:      3 days
Approved by:    re (dwhite)
2005-07-07 14:16:38 +00:00
Jeff Roberson
8b3676f1a1 - Since we don't hold a usecount in pfs_exit we have to get a holdcnt
prior to calling vgone() to prevent any races.

Sponsored by:	Isilon Systems, Inc.
Approved by:	re (vfs blanket)
2005-07-07 07:33:10 +00:00
Jung-uk Kim
c687e6de5b `net.inet.ipf.fr_running' can be a negative value, which was introduced by
recent ipfilter import.

Approved by:	re (scottl), anholt (mentor)
2005-07-07 05:59:44 +00:00
Jung-uk Kim
f8b12dd293 add myself as a new committer
Reviewed by:	anholt (mentor)
Approved by:	re (scottl)
2005-07-07 05:48:37 +00:00
Sam Leffler
82145663e3 reclaim mbuf when ieee80211_crypto_encap fails
Approved by:	re (scottl)
Obtained from:	netbsd
2005-07-07 02:55:04 +00:00
Scott Long
8225bf4c36 Fix ifnet fallout in if_ti.
Reviewed by: brooks
Approved by: re
2005-07-07 01:05:45 +00:00
Alexander Kabaev
2b9fc5e637 Exclude build32 target from the list of toolchain and kernel-toolchain
dependencies. We do not need to build 32bit compat tools to compile
a kernel.

Approved by:	re (scottl)
Reported by:	jhb
2005-07-07 00:58:41 +00:00
Sam Leffler
8f40943151 only invoke ath_rate_tx_complete to update rate control state when the
frame being sent is to be ack'd and hasn't been filtered by the h/w;
this insures we don't pass in tx descriptors that have no meaningful
state (e.g. mcast/bcast frames are not acked and so have no tx retry
counts)

Approved by:	re (scottl)
Obtained from:	Atheros
2005-07-07 00:04:50 +00:00
Sam Leffler
1729745e80 This commit was generated by cvs2svn to compensate for changes in r147801,
which included commits to RCS files with non-trunk default branches.
2005-07-06 22:06:27 +00:00
Sam Leffler
a593614af4 clr ptr after freeing state to avoid subsequent null ptr deref
(on vendor branch since this change already exists in Jouni's cvs).

Approved by:	re (scottl)
2005-07-06 22:06:27 +00:00
Warner Losh
dc1206ea0f Add support for Farallon EtherMAC PC Card.
Move ethernet MAC address setting into pccard attachment
Fix panic from IFP2ENADDR() use prior to if_alloc
Remove OLDCARD compat support.  This should work still on oldcard, however.
sn_attach now requires that the resources be activated now, so adjust.

Approved by: re (scottl)
2005-07-06 15:59:47 +00:00
Warner Losh
ce44283887 Add device entry for a Farallon EtherMac PC Card that I won on ebay. This
one is supported by the sn driver.

Approved by: re (scottl)
2005-07-06 15:56:16 +00:00
Sam Leffler
fa56f74570 add pureg/-pureg parameter for controlling pureg ap mode support
Reviewed by:	avatar
Approved by:	re (scottl)
2005-07-06 15:39:11 +00:00
Sam Leffler
c4f040c3f3 add "pureg" mode for ap operation: reject association requests from
11b-only stations when operating in 11g

Reviewed by:	avatar
Approved by:	re (scottl)
2005-07-06 15:38:27 +00:00
Colin Percival
8d284299a9 Correct a buffer overflow which occurred in the handling of some
particularly corrupt deflated data streams.

Security:	FreeBSD-SA-05:16.zlib
Approved by:	re (security blanket)
2005-07-06 14:00:52 +00:00
Sam Leffler
bc5627d91f Fix handling of data frames queued for a station in power save mode:
don't mark the MORE_DATA bit when taking it off the ps queue, there's
no 802.11 header then; we must wait to do this at encap time so
mark the mbuf instead.

Reviewed by:	avatar
Approved by:	re (scottl)
Obtained from:	Atheros
2005-07-06 01:55:17 +00:00
Sam Leffler
ebdda46c84 Fix race condition in handling node reference counts for authenticating
stations in ap mode.  Track when a node's first auth frame is
received and use this to decide whether or not to bump the refcnt.
This insures we only ever bump the refcnt once.

Reviewed by:	avatar
Approved by:	re (scottl)
2005-07-06 01:51:44 +00:00
Tai-hwa Liang
720bd46e9f Only update the scan entry state based on newly received frames.
This fixes duplicative BSS entries(memory leaks as well) listed in
"ifconfig dev list scan" when a station fails to associate with an AP.

Reviewed by:	sam
Approved by:	re (scottl)
2005-07-06 01:31:34 +00:00