GitHub/moto-9609/android_kernel_motorola_exynos9610.git
16 years agomac80211: support for mesh interfaces in mac80211 data path
Luis Carlos Cobo [Sat, 23 Feb 2008 14:17:10 +0000 (15:17 +0100)]
mac80211: support for mesh interfaces in mac80211 data path

This changes the TX/RX paths in mac80211 to support mesh interfaces.
This code will be cleaned up later again before being enabled.

Signed-off-by: Luis Carlos Cobo <luisca@cozybit.com>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agomac80211: support functions for mesh
Luis Carlos Cobo [Sat, 23 Feb 2008 14:17:09 +0000 (15:17 +0100)]
mac80211: support functions for mesh

The two important features coded in mesh.c are:

Recently Multicast Cache: in on-demand HWMP, multicast traffic is retransmitted
by every receiving node. Even though a mesh TTL counter avoids infinite loops,
it is also necessary to avoid traffic explosion by keeping a cache of multicast
mesh frame that have been received recently. With this feature, maximum number
of retransmissions of a multicast frame for the case of N nodes within the range
of each other would be N. Without it, the maximum number of retransmissions
would be in the order of N^(MESH_TTL - 1).

Code to support mesh tables.

Signed-off-by: Luis Carlos Cobo <luisca@cozybit.com>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agomac80211: mesh function and data structures definitions
Luis Carlos Cobo [Sat, 23 Feb 2008 14:17:08 +0000 (15:17 +0100)]
mac80211: mesh function and data structures definitions

Signed-off-by: Luis Carlos Cobo <luisca@cozybit.com>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agomac80211: add mesh interface type
Johannes Berg [Sat, 23 Feb 2008 14:17:07 +0000 (15:17 +0100)]
mac80211: add mesh interface type

This adds the mesh interface type.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agonl80211/cfg80211: support for mesh, sta dumping
Luis Carlos Cobo [Sat, 23 Feb 2008 14:17:06 +0000 (15:17 +0100)]
nl80211/cfg80211: support for mesh, sta dumping

Added support for mesh id and mesh path operation as well as
station structure dumping.

Signed-off-by: Luis Carlos Cobo <luisca@cozybit.com>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agoWEXT: add mesh interface type
Johannes Berg [Sat, 23 Feb 2008 14:17:05 +0000 (15:17 +0100)]
WEXT: add mesh interface type

This introduces a new WEXT type IW_MODE_MESH for mesh networks,
used for scan results.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agowireless: various definitions for mesh networking
Luis Carlos Cobo [Sat, 23 Feb 2008 14:17:04 +0000 (15:17 +0100)]
wireless: various definitions for mesh networking

Signed-off-by: Luis Carlos Cobo <luisca@cozybit.com>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years ago[IPV6]: Fix powerpc allmodconfig build warnings.
David S. Miller [Thu, 6 Mar 2008 04:58:10 +0000 (20:58 -0800)]
[IPV6]: Fix powerpc allmodconfig build warnings.

Introduced by changeset 95e41e93e18d8e1e272ce23d96bae4f17ce11d42
("[IPV6]: Make ndisc_flow_init() common for later use.")

Reported by Stephen Rothwell.

In file included from net/ipv6/netfilter/ip6_tables.c:21:
include/linux/icmpv6.h:192: warning: 'struct in6_addr' declared inside parameter list
include/linux/icmpv6.h:192: warning: its scope is only this definition or declaration, which is probably not what you want

Signed-off-by: David S. Miller <davem@davemloft.net>
16 years agonet: replace remaining __FUNCTION__ occurrences
Harvey Harrison [Thu, 6 Mar 2008 04:47:47 +0000 (20:47 -0800)]
net: replace remaining __FUNCTION__ occurrences

__FUNCTION__ is gcc-specific, use __func__

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years agoCOSA/SRP: convert channel_data.rsem to mutex
Matthias Kaehlcke [Thu, 6 Mar 2008 02:53:01 +0000 (18:53 -0800)]
COSA/SRP: convert channel_data.rsem to mutex

COSA/SRP driver: The semaphore channel_data.rsem is used as a mutex,
convert it to the mutex API

Signed-off-by: Matthias Kaehlcke <matthias@kaehlcke.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[IPV4]: Add 'rtable' field in struct sk_buff to alias 'dst' and avoid casts
Eric Dumazet [Thu, 6 Mar 2008 02:30:47 +0000 (18:30 -0800)]
[IPV4]: Add 'rtable' field in struct sk_buff to alias 'dst' and avoid casts

(Anonymous) unions can help us to avoid ugly casts.

A common cast it the (struct rtable *)skb->dst one.

Defining an union like  :
union {
     struct dst_entry *dst;
     struct rtable *rtable;
};
permits to use skb->rtable in place.

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[IPV6]: Remove commented lines.
Daniel Lezcano [Wed, 5 Mar 2008 20:37:29 +0000 (12:37 -0800)]
[IPV6]: Remove commented lines.

Remove commented lines from netns patchset.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years agoMerge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
David S. Miller [Wed, 5 Mar 2008 20:26:41 +0000 (12:26 -0800)]
Merge branch 'master' of /linux/kernel/git/davem/net-2.6

Conflicts:

net/mac80211/rc80211_pid_algo.c

16 years ago[NETNS][IPV6] icmp6 - make icmpv6_socket per namespace
Benjamin Thery [Wed, 5 Mar 2008 18:49:18 +0000 (10:49 -0800)]
[NETNS][IPV6] icmp6 - make icmpv6_socket per namespace

This patch make the changes necessary to support network namespaces in
ICMPv6.

Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] ip6_input - enable ipv6_rcv to handle several network namespace
Daniel Lezcano [Wed, 5 Mar 2008 18:48:56 +0000 (10:48 -0800)]
[NETNS][IPV6] ip6_input - enable ipv6_rcv to handle several network namespace

The different subsystem of ipv6 are ready for namespaces, so let's
activate it for ipv6_rcv.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] route6 - pass always a valid socket to ip6_dst_lookup
Daniel Lezcano [Wed, 5 Mar 2008 18:48:35 +0000 (10:48 -0800)]
[NETNS][IPV6] route6 - pass always a valid socket to ip6_dst_lookup

The ip6_dst_lookup receive a socket as parameter. In some part of the code
it is called with a NULL socket parameter. We want to rely on the socket
to retrieve the network namespace, so we always pass a valid socket in all
cases.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] route6 - add netns parameter to ip6_route_output
Daniel Lezcano [Wed, 5 Mar 2008 18:48:10 +0000 (10:48 -0800)]
[NETNS][IPV6] route6 - add netns parameter to ip6_route_output

Add an netns parameter to ip6_route_output. That will allow to access
to the right routing table for outgoing traffic.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] addrconf - make addrconf per namespace
Benjamin Thery [Wed, 5 Mar 2008 18:47:47 +0000 (10:47 -0800)]
[NETNS][IPV6] addrconf - make addrconf per namespace

All the infrastructure to propagate the network namespace information
is ready. Make use of it.

There is a special case here between the initial network namespace and
the other namespaces:

* When ipv6 is initialized at boot time (aka in the init_net), it
registers to the notifier callback. So addrconf_notify will be called
as many time as there are network devices setup on the system and the
function will add ipv6 addresses to the network devices. But the first
device which needs to have its ipv6 address setup is the loopback,
unfortunatly this is not the case. So the loopback address is setup
manually in the ipv6 init function.

* With the network namespace, this ordering problem does not appears
because notifier is already setup and active, so as soon as we
register the loopback the ipv6 address is setup and it will be the
first device.

Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] addrconf - Pass the proper network namespace parameters to addrconf
Daniel Lezcano [Wed, 5 Mar 2008 18:46:57 +0000 (10:46 -0800)]
[NETNS][IPV6] addrconf - Pass the proper network namespace parameters to addrconf

This patch propagates the network namespace pointer to the address
configuration routines which need it, which means adding a new
parameter to these functions, and make them use it instead of using
the initial network namespace.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] proc - protect snmp6 from non-init_net calls
Daniel Lezcano [Wed, 5 Mar 2008 18:46:31 +0000 (10:46 -0800)]
[NETNS][IPV6] proc - protect snmp6 from non-init_net calls

This patchset avoids creation of the /proc entry for snmp6 when
the call is made from a network namespace different from the init_net.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] af_inet6 - allow socket creation per namespace
Benjamin Thery [Wed, 5 Mar 2008 18:45:59 +0000 (10:45 -0800)]
[NETNS][IPV6] af_inet6 - allow socket creation per namespace

Allow creation of IPv6 raw and datagram sockets in network namespaces
other than init_net.

Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] Move sysctl initialization later on in the IPv6 init sequence
Benjamin Thery [Wed, 5 Mar 2008 18:45:36 +0000 (10:45 -0800)]
[NETNS][IPV6] Move sysctl initialization later on in the IPv6 init sequence

This patch moves initialization of IPv6 sysctl stuff at the end of
IPv6 initialization.

This will be helpful for network namespaces where some sysctl entries
depend on per-namespace variables, that need to be allocated and
initialized before they are referenced by sysctl.

Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years agomake s2io.c:init_tti() static
Adrian Bunk [Tue, 4 Mar 2008 23:19:22 +0000 (15:19 -0800)]
make s2io.c:init_tti() static

Make the needlessly global init_tti() static.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Acked-by: "Ramkrishna Vepa" <Ramkrishna.Vepa@neterion.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
16 years agoFIXED_PHY must depend on PHYLIB=y
Adrian Bunk [Wed, 27 Feb 2008 15:07:54 +0000 (17:07 +0200)]
FIXED_PHY must depend on PHYLIB=y

This patch fixes the following build error introduced by commit
a79d8e93d300adb84cccc38ac396cfb118c238ad and reported by Olaf Hering:

<--  snip  -->

...
  LD      .tmp_vmlinux1
arch/powerpc/sysdev/built-in.o: In function `of_add_fixed_phys':
fsl_soc.c:(.init.text+0xd34): undefined reference to `fixed_phy_add'
make: *** [.tmp_vmlinux1] Error 1

<--   snip  -->

Signed-off-by: Adrian Bunk <adrian.bunk@movial.fi>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
16 years agofix drivers/net/atarilance.c compilation
Adrian Bunk [Mon, 3 Mar 2008 21:34:13 +0000 (23:34 +0200)]
fix drivers/net/atarilance.c compilation

This patch fixes the following build error:

<--  snip  -->

...
  CC [M]  drivers/net/atarilance.o
{standard input}: Assembler messages:
{standard input}:406: Error: symbol `Lberr' is already defined
{standard input}:460: Error: symbol `Lberr' is already defined
make[3]: *** [drivers/net/atarilance.o] Error 1

<--  snip  -->

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
16 years agoIntel ethernet adapter: Update MAINTAINERS
Auke Kok [Mon, 3 Mar 2008 22:37:47 +0000 (14:37 -0800)]
Intel ethernet adapter: Update MAINTAINERS

Another team member unfortunately left: update MAINTAINERS.

Condense the 3 lists down to a single list for all our drivers.

Point to our new sourceforge index page which is slightly
better navigateable than the sf.net project page.

Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
16 years agoixgbe: fix typo in speed mesage
Emil Tantilov [Mon, 3 Mar 2008 22:37:42 +0000 (14:37 -0800)]
ixgbe: fix typo in speed mesage

Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
16 years agoe100: Do suspend/shutdown like e1000
Auke Kok [Mon, 3 Mar 2008 22:37:31 +0000 (14:37 -0800)]
e100: Do suspend/shutdown like e1000

This fixes a "trying to free already free IRQ" message and simplifies
the shutdown/suspend code by re-using already existing code when going
to suspend. The code is now symmetric with e100_resume.

Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
16 years agoehea: Fix missing Kconfig dependency
Thomas Klein [Mon, 3 Mar 2008 12:52:06 +0000 (13:52 +0100)]
ehea: Fix missing Kconfig dependency

Fixed Kconfig: ehea driver requires sparse mem

Signed-off-by: Thomas Klein <tklein@de.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
16 years agoaxnet_cs: change debugging level for "Too much work at interrupt" message.
Komuro [Sat, 1 Mar 2008 01:52:03 +0000 (10:52 +0900)]
axnet_cs: change debugging level for "Too much work at interrupt" message.

This message is frequently displayed even if normal file-transfer.

Signed-off-by: Komuro <komurojun-mbn@nifty.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
16 years agomv643xx_eth: Define module alias for platform device
Martin Michlmayr [Thu, 28 Feb 2008 20:11:48 +0000 (21:11 +0100)]
mv643xx_eth: Define module alias for platform device

The mv643xx_eth driver can be loaded as a platform device, as is done by
various Orion (ARM) based devices.  The driver needs to define a module
alias for the platform driver so udev will load it automatically.
Tested with Debian on a QNAP TS-209.

Signed-off-by: Martin Michlmayr <tbm@cyrius.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
16 years agode2104x: remove BUG_ON() when changing media type
Ondrej Zary [Mon, 25 Feb 2008 17:45:46 +0000 (18:45 +0100)]
de2104x: remove BUG_ON() when changing media type

When the chip dies (probably because of a bug somewhere in the driver),
de_stop_rxtx() fails and changing the media type crashes the whole machine.
Replace BUG_ON() in de_set_media() with a warning.

Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
16 years agodm9161: add configuration for MII/RMII
frederic Rodo [Wed, 27 Feb 2008 11:58:37 +0000 (12:58 +0100)]
dm9161: add configuration for MII/RMII

Signed-off-by: Frederic Rodo <f.rodo@til-technologies.fr>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
16 years ago[IPCONFIG]: The kernel gets no IP from some DHCP servers
Stephen Hemminger [Wed, 5 Mar 2008 01:03:49 +0000 (17:03 -0800)]
[IPCONFIG]: The kernel gets no IP from some DHCP servers

From: Stephen Hemminger <shemminger@linux-foundation.org>

Based upon a patch by Marcel Wappler:

   This patch fixes a DHCP issue of the kernel: some DHCP servers
   (i.e.  in the Linksys WRT54Gv5) are very strict about the contents
   of the DHCPDISCOVER packet they receive from clients.

   Table 5 in RFC2131 page 36 requests the fields 'ciaddr' and
   'siaddr' MUST be set to '0'.  These DHCP servers ignore Linux
   kernel's DHCP discovery packets with these two fields set to
   '255.255.255.255' (in contrast to popular DHCP clients, such as
   'dhclient' or 'udhcpc').  This leads to a not booting system.

Signed-off-by: David S. Miller <davem@davemloft.net>
16 years agoMerge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/linville/wireles...
David S. Miller [Wed, 5 Mar 2008 00:44:01 +0000 (16:44 -0800)]
Merge branch 'master' of /linux/kernel/git/linville/wireless-2.6

16 years agob43legacy: Fix module init message
Michael Buesch [Tue, 4 Mar 2008 19:31:13 +0000 (20:31 +0100)]
b43legacy: Fix module init message

This fixes the module init message to tell that the legacy
driver loaded. This makes it less confusing, in case both drivers are loaded.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agorndis_wlan: fix broken data copy
Jussi Kivilinna [Tue, 4 Mar 2008 18:05:27 +0000 (20:05 +0200)]
rndis_wlan: fix broken data copy

Replace broken code that attempted to copy 6 byte array to 64-bit
integer. Due to missing cast to 64-bit integer, left shift operation
were 32-bit and lead to bytes been copied over each other. New code
uses simple memcpy, for greater readability and efficiency.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agolibertas: compare the current command with response
Sebastian Siewior [Tue, 4 Mar 2008 17:22:27 +0000 (18:22 +0100)]
libertas: compare the current command with response

instead of with itself.

Signed-off-by: Sebastian Siewior <bigeasy@linutronix.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agolibertas: fix sanity check on sequence number in command response
David Woodhouse [Mon, 3 Mar 2008 11:20:12 +0000 (12:20 +0100)]
libertas: fix sanity check on sequence number in command response

Slightly more useful if we compare it against the sequence number of the
command we have outstanding, rather than comparing the reply with itself.

Doh. Pointed out by Sebastian Siewior

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Acked-by: Dan Williams <dcbw@redhat.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agop54: fix eeprom parser length sanity checks
Johannes Berg [Fri, 29 Feb 2008 22:28:25 +0000 (23:28 +0100)]
p54: fix eeprom parser length sanity checks

When I called p54_parse_eeprom() on a hand-coded structure
I managed to make a small mistake with wrap->len which caused
a segfault a few lines down when trying to read entry->len.
This patch changes the validation code to avoid such problems.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Tested-by: Florian Fainelli <florian.fainelli@telecomint.eu>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agop54: fix EEPROM structure endianness
Johannes Berg [Fri, 29 Feb 2008 12:56:33 +0000 (13:56 +0100)]
p54: fix EEPROM structure endianness

Since the EEPROM structure is read from hardware, it is
always little endian, annotate that in the struct and
make sure to convert where applicable.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Cc: Michael Wu <flamingice@sourmilk.net>
Tested-by: Florian Fainelli <florian.fainelli@telecomint.eu>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agossb: Add pcibios_enable_device() return value check
Yoichi Yuasa [Fri, 29 Feb 2008 10:09:31 +0000 (11:09 +0100)]
ssb: Add pcibios_enable_device() return value check

This patch has added pcibios_enable_device() return value check.

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years agorc80211-pid: fix rate adjustment
Stefano Brivio [Sun, 2 Mar 2008 10:33:10 +0000 (11:33 +0100)]
rc80211-pid: fix rate adjustment

Merge rate_control_pid_shift_adjust() to rate_control_pid_adjust_rate()
in order to make the learning algorithm aware of constraints on rates. Also
add some comments and rename variables.

This fixes a bug which prevented 802.11b/g non-AP STAs from working with
802.11b only AP STAs.

This patch was originally destined for 2.6.26, and is being backported
to fix a user reported problem in post-2.6.24 kernels.

Signed-off-by: Stefano Brivio <stefano.brivio@polimi.it>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
16 years ago[ESP]: Add select on AUTHENC
Herbert Xu [Tue, 4 Mar 2008 22:29:21 +0000 (14:29 -0800)]
[ESP]: Add select on AUTHENC

Now the ESP uses the AEAD interface even for algorithms which are
not combined mode, we need to select CONFIG_CRYPTO_AUTHENC as
otherwise only combined mode algorithms will work.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[TCP]: Improve ipv4 established hash function.
David S. Miller [Tue, 4 Mar 2008 22:28:41 +0000 (14:28 -0800)]
[TCP]: Improve ipv4 established hash function.

If all of the entropy is in the local and foreign addresses,
but xor'ing together would cancel out that entropy, the
current hash performs poorly.

Suggested by Cosmin Ratiu:

Basically, the situation is as follows: There is a client
machine and a server machine. Both create 15000 virtual
interfaces, open up a socket for each pair of interfaces and
do SIP traffic. By profiling I noticed that there is a lot of
time spent walking the established hash chains with this
particular setup.

The addresses were distributed like this: client interfaces
were 198.18.0.1/16 with increments of 1 and server interfaces
were 198.18.128.1/16 with increments of 1. As I said, there
were 15000 interfaces. Source and destination ports were 5060
for each connection.  So in this case, ports don't matter for
hashing purposes, and the bits from the address pairs used
cancel each other, meaning there are no differences in the
whole lot of pairs, so they all end up in the same hash chain.

Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[TCP]: TCP cubic v2.2
Sangtae Ha [Tue, 4 Mar 2008 22:17:41 +0000 (14:17 -0800)]
[TCP]: TCP cubic v2.2

We have updated CUBIC to fix some issues with slow increase in large
BDP networks. We also improved its convergence speed. The fix is in
fact very simple -- the window increase limit of smax during the
window probing phase (i.e., convex growth phase) is removed. We found
that this does not affect TCP friendliness, but only improves its
scalability. We have run some tests in our lab and also over the
Internet path from NCSU to Japan. These results can be seen from the
following page:

http://netsrv.csc.ncsu.edu/wiki/index.php/Intra_protocol_fairness_testing_with_linux-2.6.23.9
http://netsrv.csc.ncsu.edu/wiki/index.php/RTT_fairness_testing_with_linux-2.6.23.9
http://netsrv.csc.ncsu.edu/wiki/index.php/TCP_friendliness_testing_with_linux-2.6.23.9

Signed-off-by: Sangtae Ha <sha2@ncsu.edu>
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] route6 - Make ip6_dst_gc simpler
Daniel Lezcano [Tue, 4 Mar 2008 21:50:14 +0000 (13:50 -0800)]
[NETNS][IPV6] route6 - Make ip6_dst_gc simpler

This patches improves the readibility of the ip6_dst_gc() routine.
It simplifies long lines which grow a lot due to the introduction
of network namespaces support.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Acked-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] route6 - make garbage collection work with multiple network namespaces
Benjamin Thery [Tue, 4 Mar 2008 21:49:47 +0000 (13:49 -0800)]
[NETNS][IPV6] route6 - make garbage collection work with multiple network namespaces

This patch makes the necessary changes to make IPv6 dst_entry garbage
collection work with multiple network namespaces.

In ip6_dst_gc(), static local variables are now declared
per-namespace.

Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] route6 - move ip6_dst_ops inside the network namespace
Benjamin Thery [Tue, 4 Mar 2008 21:49:23 +0000 (13:49 -0800)]
[NETNS][IPV6] route6 - move ip6_dst_ops inside the network namespace

The ip6_dst_ops is moved inside the network namespace structure.  All
references to this structure are now relative to the initial network
namespace.

Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] route6 - dynamically allocate ip6_dst_ops
Daniel Lezcano [Tue, 4 Mar 2008 21:48:53 +0000 (13:48 -0800)]
[NETNS][IPV6] route6 - dynamically allocate ip6_dst_ops

ip6_dst_ops is dynamically allocated in init and exit functions.  That
provides the ability to do multiple instanciations of this structure.

This will be needed for network namespaces, indeed dst_ops stores data
that are required to be per namespace: entries and gc_thresh.

Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] rt6_info - move rt6_info structure inside the namespace
Daniel Lezcano [Tue, 4 Mar 2008 21:48:30 +0000 (13:48 -0800)]
[NETNS][IPV6] rt6_info - move rt6_info structure inside the namespace

The rt6_info structures are moved inside the network namespace
structure. All references to these structures are now relative to the
initial network namespace.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] rt6_info - make rt6_info accessed as a pointer
Daniel Lezcano [Tue, 4 Mar 2008 21:48:10 +0000 (13:48 -0800)]
[NETNS][IPV6] rt6_info - make rt6_info accessed as a pointer

This patch make mindless changes and prepares the code to use dynamic
allocation for rt6_info structure. The code accesses the rt6_info
structure as a pointer instead of a global static variable.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] route6 - make route6 per namespace
Daniel Lezcano [Tue, 4 Mar 2008 21:47:47 +0000 (13:47 -0800)]
[NETNS][IPV6] route6 - make route6 per namespace

This patch makes the routing engine use the network namespaces to
access routing informations: Add a network namespace parameter to
ipv6_route_ioctl and propagate the network namespace value to all the
routing code that have not yet been changed.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] route6 - Pass the network namespace parameter to rt6_purge_dflt_routers
Daniel Lezcano [Tue, 4 Mar 2008 21:47:14 +0000 (13:47 -0800)]
[NETNS][IPV6] route6 - Pass the network namespace parameter to rt6_purge_dflt_routers

Add a network namespace parameter to rt6_purge_dflt_routers.  This is
needed to call fib6_get_table with the appropriate network namespace.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] route6 - Pass network namespace to rt6_add_route_info and rt6_get_route...
Daniel Lezcano [Tue, 4 Mar 2008 21:46:48 +0000 (13:46 -0800)]
[NETNS][IPV6] route6 - Pass network namespace to rt6_add_route_info and rt6_get_route_info

Add a network namespace parameter to rt6_add_route_info() and
rt6_get_route_info to enable them to handle multiple network
namespaces.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] route6 - Make proc entry /proc/net/rt6_stats per namespace
Daniel Lezcano [Tue, 4 Mar 2008 21:46:23 +0000 (13:46 -0800)]
[NETNS][IPV6] route6 - Make proc entry /proc/net/rt6_stats per namespace

Make the proc entry /proc/net/rt6_stats work in all network namespace.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] route6 - Pass the network namespace parameter to rt6_lookup
Daniel Lezcano [Tue, 4 Mar 2008 21:45:59 +0000 (13:45 -0800)]
[NETNS][IPV6] route6 - Pass the network namespace parameter to rt6_lookup

Add a network namespace parameter to rt6_lookup().

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] route6 - create route6 proc files for the namespace
Daniel Lezcano [Tue, 4 Mar 2008 21:45:33 +0000 (13:45 -0800)]
[NETNS][IPV6] route6 - create route6 proc files for the namespace

Make /proc/net/ipv6_route and /proc/net/rt6_stats to be per namespace.
These proc files are now created when the network namespace is
initialized.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETPOLL]: Revert two bogus cleanups that broke netconsole.
David S. Miller [Tue, 4 Mar 2008 20:28:49 +0000 (12:28 -0800)]
[NETPOLL]: Revert two bogus cleanups that broke netconsole.

Based upon a report by Andrew Morton and code analysis done
by Jarek Poplawski.

This reverts 33f807ba0d9259e7c75c7a2ce8bd2787e5b540c7 ("[NETPOLL]:
Kill NETPOLL_RX_DROP, set but never tested.")  and
c7b6ea24b43afb5749cb704e143df19d70e23dea ("[NETPOLL]: Don't need
rx_flags.").

The rx_flags did get tested for zero vs. non-zero and therefore we do
need those tests and that code which sets NETPOLL_RX_DROP et al.

Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[AF_KEY]: Dump SA/SP entries non-atomically
Timo Teras [Tue, 4 Mar 2008 07:40:12 +0000 (23:40 -0800)]
[AF_KEY]: Dump SA/SP entries non-atomically

Stop dumping of entries when af_key socket receive queue is getting
full and continue it later when there is more room again.

This fixes dumping of large databases. Currently the entries not
fitting into the receive queue are just dropped (including the
end-of-dump message) which can confuse applications.

Signed-off-by: Timo Teras <timo.teras@iki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[TIPC]: Convert tsock->sem in a mutex
Matthias Kaehlcke [Tue, 4 Mar 2008 07:35:53 +0000 (23:35 -0800)]
[TIPC]: Convert tsock->sem in a mutex

The semaphore tsock->sem is used as mutex, convert it to the mutex API

Signed-off-by: Matthias Kaehlcke <matthias@kaehlcke.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] rt6_stats - make the stats per network namespace
Benjamin Thery [Tue, 4 Mar 2008 07:34:17 +0000 (23:34 -0800)]
[NETNS][IPV6] rt6_stats - make the stats per network namespace

The rt6_stats is now per namespace with this patch. It is allocated
when a network namespace is created and freed when the network
namespace exits and references are relative to the network namespace.

Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] rt6_stats - dynamically allocate the routes statistics
Daniel Lezcano [Tue, 4 Mar 2008 07:33:43 +0000 (23:33 -0800)]
[NETNS][IPV6] rt6_stats - dynamically allocate the routes statistics

This patch allocates the rt6_stats struct dynamically when the fib6 is
initialized. That provides the ability to create several instances of
this structure for the network namespaces.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] fib6_rules - handle several network namespaces
Daniel Lezcano [Tue, 4 Mar 2008 07:33:08 +0000 (23:33 -0800)]
[NETNS][IPV6] fib6_rules - handle several network namespaces

The fib6_rules_ops is moved to the network namespace structure.  All
references are changed to have it relatively to it.

Each time a network namespace is created a new fib6_rules_ops is
allocated, initialized and stored into the network namespace
structure.

The common part of the fib rules is namespace aware, so it is quite
easy to retrieve the network namespace from the rules and use it in
the different callbacks.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] fib6 rule - dynamic allocation of the rules struct ops
Daniel Lezcano [Tue, 4 Mar 2008 07:32:30 +0000 (23:32 -0800)]
[NETNS][IPV6] fib6 rule - dynamic allocation of the rules struct ops

The fib6_rules_ops structure is dynamically allocated, so that allows
to make several instances of it per network namespace.

The global static fib6_rules_ops structure is renamed to
fib6_rules_ops_template in order to quickly memcopy it for the
structure initialization.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] ip6_fib - clean node use namespace
Benjamin Thery [Tue, 4 Mar 2008 07:31:57 +0000 (23:31 -0800)]
[NETNS][IPV6] ip6_fib - clean node use namespace

The fib6_clean_node function should have the network namespace it is
working on. The fib6_cleaner_t structure is extended with the network
namespace field to be passed to the fib6_clean_node function.

The different functions calling the fib6_clean_node function are
extended with the netns parameter when needed to propagate the netns
pointer.

Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] ip6_fib - gc timer per namespace
Daniel Lezcano [Tue, 4 Mar 2008 07:31:11 +0000 (23:31 -0800)]
[NETNS][IPV6] ip6_fib - gc timer per namespace

Move the timer initialization at the network namespace creation and
store the network namespace in the timer argument.

That enables multiple timers (one per network namespace) to do garbage
collecting.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] ip6_fib - dynamically allocate gc-timer
Daniel Lezcano [Tue, 4 Mar 2008 07:29:33 +0000 (23:29 -0800)]
[NETNS][IPV6] ip6_fib - dynamically allocate gc-timer

The ip6_fib_timer gc timer is dynamically allocated and initialized in
the ip6 fib init function. There are no more references to a static
global variable. That will allow to make multiple instance of the
garbage collecting timer and make them per namespace.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] ip6_fib - add net to gc timer parameter
Daniel Lezcano [Tue, 4 Mar 2008 07:28:58 +0000 (23:28 -0800)]
[NETNS][IPV6] ip6_fib - add net to gc timer parameter

The fib tables are now relative to the network namespace. When the
garbage collector timer expires, we must have a network namespace
parameter in order to retrieve the tables. For now this is the
init_net, but we should be able to have a timer per namespace and use
the timer callback parameter to pass the network namespace from the
expired timer.

The timer callback, fib6_run_gc, is actually used to be called
synchronously by some functions and asynchronously when the timer
expires.

When the timer expires, the delay specified for fib6_run_gc parameter
is always zero. So, I changed fib6_run_gc to not be a timer callback
but a function called by the timer callback and I added a timer
callback where its work is just to retrieve from the data arg of the
timer the network namespace and call fib6_run_gc with zero expiring
time and the network namespace parameters. That makes the code cleaner
for the fib6_run_gc callers.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] ip6_fib - fib6_clean_all handle several network namespaces
Daniel Lezcano [Tue, 4 Mar 2008 07:27:06 +0000 (23:27 -0800)]
[NETNS][IPV6] ip6_fib - fib6_clean_all handle several network namespaces

The function fib6_clean_all takes the network namespace as
parameter. That allows to flush the routes related to a specific
network namespace.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] ip6_fib - make it per network namespace
Daniel Lezcano [Tue, 4 Mar 2008 07:25:27 +0000 (23:25 -0800)]
[NETNS][IPV6] ip6_fib - make it per network namespace

The fib table for ipv6 are moved to the network namespace structure.
All references to them are made relatively to the network namespace.

All external calls to the ip6_fib functions taking the network
namespace parameter are made using the init_net variable, so the
ip6_fib engine is ready for the namespaces but the callers not yet.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NETNS][IPV6] ip6_fib - dynamically allocate the fib tables
Daniel Lezcano [Tue, 4 Mar 2008 07:24:31 +0000 (23:24 -0800)]
[NETNS][IPV6] ip6_fib - dynamically allocate the fib tables

This patch changes the fib6 tables to be dynamically allocated.  That
provides the ability to make several instances of them when a new
network namespace is created.

Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[IPV6] MCAST: Use standard path for sending MLD/MLDv2 messages.
YOSHIFUJI Hideaki [Fri, 7 Dec 2007 01:40:56 +0000 (17:40 -0800)]
[IPV6] MCAST: Use standard path for sending MLD/MLDv2 messages.

This is changing the paths for sending MLD/MLDv2 messages
from dev_queue_xmit() to standard dst_output().

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[IPV6]: Make ndisc_dst_alloc() common for later use.
YOSHIFUJI Hideaki [Fri, 7 Dec 2007 00:11:48 +0000 (16:11 -0800)]
[IPV6]: Make ndisc_dst_alloc() common for later use.

For later use, this patch is renaming ndisc_dst_alloc()
(and related function/structures) to icmp6_dst_alloc()
(and so on).  This patch also removing unused function-
pointer argument for it.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[IPV6]: Make ndisc_flow_init() common for later use.
YOSHIFUJI Hideaki [Thu, 6 Dec 2007 23:43:30 +0000 (15:43 -0800)]
[IPV6]: Make ndisc_flow_init() common for later use.

For later use, this patch is renaming ndisc_flow_init() to
icmpv6_flow_init() and putting it in common place.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[IPV6] ADDRCONF: Convert ipv6_get_saddr() to ipv6_dev_get_saddr().
YOSHIFUJI Hideaki [Mon, 3 Mar 2008 12:44:34 +0000 (21:44 +0900)]
[IPV6] ADDRCONF: Convert ipv6_get_saddr() to ipv6_dev_get_saddr().

Since most users of ipv6_get_saddr() pass non-NULL as
dst argument, use ipv6_dev_get_saddr() directly.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[NET] NEIGHBOUR: Remove unpopular neigh_is_connected().
YOSHIFUJI Hideaki [Tue, 4 Mar 2008 05:55:03 +0000 (14:55 +0900)]
[NET] NEIGHBOUR: Remove unpopular neigh_is_connected().

neigh_is_connected() is not popular at all, and the only user
drivers/net/cxgb3/l2t.c:t3_l2t_update() also have raw (expanded) expression.
Let's expand it and remove the inline function.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[IPV6]: Use htonl() instead of __constant_htonl() where appricable.
YOSHIFUJI Hideaki [Thu, 28 Feb 2008 08:03:12 +0000 (17:03 +0900)]
[IPV6]: Use htonl() instead of __constant_htonl() where appricable.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[IPV6] SYSCTL: complete initialization for sysctl table in subsystem code.
YOSHIFUJI Hideaki [Wed, 27 Feb 2008 15:24:28 +0000 (00:24 +0900)]
[IPV6] SYSCTL: complete initialization for sysctl table in subsystem code.

Move initialization bits for subsystem sysctl tables to
appropriate functions.
 - route
 - icmp

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[IPV6]: Move packet_type{} related bits to af_inet6.c.
YOSHIFUJI Hideaki [Wed, 27 Feb 2008 14:14:03 +0000 (23:14 +0900)]
[IPV6]: Move packet_type{} related bits to af_inet6.c.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[IPV6] UDP: Rename IPv6 UDP files.
YOSHIFUJI Hideaki [Thu, 21 Feb 2008 07:13:26 +0000 (16:13 +0900)]
[IPV6] UDP: Rename IPv6 UDP files.

Rename net/ipv6/udp.c to net/ipv6/udp_ipv6.c
Rename net/ipv6/udplite.c to net/ipv6/udplite_ipv6.c.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[IPV4] UDP: Move IPv4-specific bits to other file.
YOSHIFUJI Hideaki [Tue, 4 Mar 2008 05:50:52 +0000 (14:50 +0900)]
[IPV4] UDP: Move IPv4-specific bits to other file.

Move IPv4-specific UDP bits from net/ipv4/udp.c into (new) net/ipv4/udp_ipv4.c.
Rename net/ipv4/udplite.c to net/ipv4/udplite_ipv4.c.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[IPV4]: Fix size description of CONFIG_INET.
YOSHIFUJI Hideaki [Tue, 12 Feb 2008 08:35:16 +0000 (17:35 +0900)]
[IPV4]: Fix size description of CONFIG_INET.

CONFIG_INET now enlarges about 400KB, not 140KB.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[UDP]: Allow users to configure UDP-Lite.
YOSHIFUJI Hideaki [Fri, 29 Feb 2008 16:06:47 +0000 (01:06 +0900)]
[UDP]: Allow users to configure UDP-Lite.

Let's give users an option for disabling UDP-Lite (~4K).

old:
|    text    data     bss     dec     hex filename
|  286498   12432    6072  305002   4a76a net/ipv4/built-in.o
|  193830    8192    3204  205226   321aa net/ipv6/ipv6.o

new (without UDP-Lite):
|    text    data     bss     dec     hex filename
|  284086   12136    5432  301654   49a56 net/ipv4/built-in.o
|  191835    7832    3076  202743   317f7 net/ipv6/ipv6.o

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[TCP]: Add IPv6 support to TCP SYN cookies
Glenn Griffin [Fri, 8 Feb 2008 05:49:26 +0000 (21:49 -0800)]
[TCP]: Add IPv6 support to TCP SYN cookies

Updated to incorporate Eric's suggestion of using a per cpu buffer
rather than allocating on the stack.  Just a two line change, but will
resend in it's entirety.

Signed-off-by: Glenn Griffin <ggriffin.kernel@gmail.com>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[TCP]: lower stack usage in cookie_hash() function
Eric Dumazet [Thu, 7 Feb 2008 09:40:19 +0000 (10:40 +0100)]
[TCP]: lower stack usage in cookie_hash() function

400 bytes allocated on stack might be a litle bit too much. Using a
per_cpu var is more friendly.

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
16 years ago[PPPOL2TP]: Add missing sock_put() in pppol2tp_tunnel_closeall()
Jarek Poplawski [Tue, 4 Mar 2008 04:49:34 +0000 (20:49 -0800)]
[PPPOL2TP]: Add missing sock_put() in pppol2tp_tunnel_closeall()

Every skb removed from session->reorder_q needs sock_put().

Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Acked-by: James Chapman <jchapman@katalix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years agoSubject: [PPPOL2TP] add missing sock_put() in pppol2tp_recv_dequeue()
Jarek Poplawski [Tue, 4 Mar 2008 04:48:53 +0000 (20:48 -0800)]
Subject: [PPPOL2TP] add missing sock_put() in pppol2tp_recv_dequeue()

Every skb removed from session->reorder_q needs sock_put().

Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Acked-by: James Chapman <jchapman@katalix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[ARP]: Introduce the arp_hdr_len helper.
Pavel Emelyanov [Mon, 3 Mar 2008 20:20:57 +0000 (12:20 -0800)]
[ARP]: Introduce the arp_hdr_len helper.

There are some place, that calculate the ARP header length. These
calculations are correct, but
 a) some operate with "magic" constants,
 b) enlarge the code length (sometimes at the cost of coding style),
 c) are not informative from the first glance.

The proposal is to introduce a helper, that includes all the good
sides of these calculations.

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[BLUETOOTH]: l2cap info_timer delete fix in hci_conn_del
Dave Young [Mon, 3 Mar 2008 20:18:55 +0000 (12:18 -0800)]
[BLUETOOTH]: l2cap info_timer delete fix in hci_conn_del

When the l2cap info_timer is active the info_state will be set to
L2CAP_INFO_FEAT_MASK_REQ_SENT, and it will be unset after the timer is
deleted or timeout triggered.

Here in l2cap_conn_del only call del_timer_sync when the info_state is
set to L2CAP_INFO_FEAT_MASK_REQ_SENT.

Signed-off-by: Dave Young <hidave.darkstar@gmail.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[NET]: Fix race in generic address resolution.
Frank Blaschka [Mon, 3 Mar 2008 20:16:04 +0000 (12:16 -0800)]
[NET]: Fix race in generic address resolution.

neigh_update sends skb from neigh->arp_queue while neigh_timer_handler
has increased skbs refcount and calls solicit with the
skb. neigh_timer_handler should not increase skbs refcount but make a
copy of the skb and do solicit with the copy.

Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years agoiucv: fix build error on !SMP
Heiko Carstens [Mon, 3 Mar 2008 20:12:33 +0000 (12:12 -0800)]
iucv: fix build error on !SMP

Since a5fbb6d1064be885d2a6b82f625186753cf74848
"KVM: fix !SMP build error" smp_call_function isn't a define anymore
that folds into nothing but a define that calls up_smp_call_function
with all parameters. Hence we cannot #ifdef out the unused code
anymore...
This seems to be the preferred method, so do this for s390 as well.

net/iucv/iucv.c: In function 'iucv_cleanup_queue':
net/iucv/iucv.c:657: error: '__iucv_cleanup_queue' undeclared

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
Signed-off-by: Ursula Braun <braunu@de.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[TCP]: Must count fack_count also when skipping
Ilpo Järvinen [Mon, 3 Mar 2008 20:10:16 +0000 (12:10 -0800)]
[TCP]: Must count fack_count also when skipping

It makes fackets_out to grow too slowly compared with the
real write queue.

This shouldn't cause those BUG_TRAP(packets <= tp->packets_out)
to trigger but how knows how such inconsistent fackets_out
affects here and there around TCP when everything is nowadays
assuming accurate fackets_out. So lets see if this silences
them all.

Reported by Guillaume Chazarain <guichaz@gmail.com>.

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years agoipv6: fix inet6_init/icmpv6_cleanup sections mismatch
Alexey Dobriyan [Mon, 3 Mar 2008 20:02:54 +0000 (12:02 -0800)]
ipv6: fix inet6_init/icmpv6_cleanup sections mismatch

Signed-off-by: Alexey Dobriyan <adobriyan@sw.ru>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[TCP]: Merge exit paths in tcp_v4_conn_request.
Denis V. Lunev [Mon, 3 Mar 2008 19:59:32 +0000 (11:59 -0800)]
[TCP]: Merge exit paths in tcp_v4_conn_request.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[SCTP]: seq_printf format warning. (fixed)
Denis V. Lunev [Mon, 3 Mar 2008 19:55:54 +0000 (11:55 -0800)]
[SCTP]: seq_printf format warning. (fixed)

sctp_association->hbinterval is unsigned long. Replace %8d with %8lu.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[IPV4]: skb->dst can't be NULL in ip_options_echo.
Denis V. Lunev [Mon, 3 Mar 2008 19:50:10 +0000 (11:50 -0800)]
[IPV4]: skb->dst can't be NULL in ip_options_echo.

ip_options_echo is called on the packet input path after the initial
routing. The dst entry on the packet is cleared only in the several
very specific places and immidiately assigned back (may be new).

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years ago[ICMP]: Section conflict between icmp_sk_init/icmp_sk_exit.
Denis V. Lunev [Fri, 29 Feb 2008 22:15:19 +0000 (14:15 -0800)]
[ICMP]: Section conflict between icmp_sk_init/icmp_sk_exit.

Functions from __exit section should not be called from ones in __init
section. Fix this conflict.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 years agoMerge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/linville/wireles...
David S. Miller [Fri, 29 Feb 2008 21:41:25 +0000 (13:41 -0800)]
Merge branch 'master' of /linux/kernel/git/linville/wireless-2.6.26

16 years agomac80211: fix key replacing, hw accel
Johannes Berg [Wed, 27 Feb 2008 12:39:00 +0000 (13:39 +0100)]
mac80211: fix key replacing, hw accel

Even though I thought about it a lot and had also tested it, some
of my recent changes in the key code broke replacing keys, making
the kernel oops because a key is removed from a list while not on
it.

This patch fixes that using the list as an indication whether or
not the key is on it (an empty list means it's not on any list.)

Also, this patch fixes hw accel enabling, the check for not doing
hw accel when the interface is down was lost and is restored by
this.

Additionally, move adding the key to the list into the function
__ieee80211_key_replace() for more consistency.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>