Dushan Tcholich reports that on his system ksoftirqd can consume
between %6 to %10 of cpu time, and cause ~200 context switches per
second.
He then correlated this with a report by bdupree@techfinesse.com:
http://marc.info/?l=linux-kernel&m=119613299024398&w=2
and the culprit cause seems to be starting the bridge interface.
In particular, when starting the bridge interface, his scripts
are specifying a hello timer interval of "0".
The bridge hello time can't be safely set to values less than 1
second, otherwise it is possible to end up with a runaway timer.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add more ethtool generic operations to dump the bridge offload
settings.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
From: Stephen Hemminger <shemminger@vyatta.com>
Based upon original patch by Herbert Xu, which contained
the following problem description:
--------------------
When the forward delay is set to zero, we still delay the setting
of the forwarding state by one or possibly two timers depending
on whether STP is enabled. This could either turn out to be
instantaneous, or horribly slow depending on the load of the
machine.
As there is nothing preventing us from enabling forwarding straight
away, this patch eliminates this potential delay by executing the
code directly if the forward delay is zero.
The effect of this problem is that immediately after the carrier
comes on a port, the bridge will drop all packets received from
that port until it enters forwarding mode, thus causing unnecessary
packet loss.
Note that this patch doesn't fully remove the delay due to the
link watcher. We should also check the carrier state when we
are about to drop an incoming packet because the port is disabled.
But that's for another patch.
--------------------
This version of the fix takes a different approach, in that
it just does the state change directly.
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch fixes the following warning due to incompatible pointer
assignment:
net/bridge/br_netfilter.c: In function 'br_netfilter_rtable_init':
net/bridge/br_netfilter.c:116: warning: assignment from incompatible
pointer type
This warning is due to commit 4adf0af681
from July 30 (send correct MTU value in PMTU (revised)).
Signed-off-by: Rami Rosen <ramirose@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When bridging interfaces with different MTUs, the bridge correctly chooses
the minimum of the MTUs of the physical devices as the bridges MTU. But
when a frame is passed which fits through the incoming, but not through
the outgoing interface, a "Fragmentation Needed" packet is generated.
However, the propagated MTU is hardcoded to 1500, which is wrong in this
situation. The sender will repeat the packet again with the same frame
size, and the same problem will occur again.
Instead of sending 1500, the (correct) MTU value of the bridge is now sent
via PMTU. To achieve this, the corresponding rtable structure is stored
in its net_bridge structure.
Modified to get rid of fake_net_device as well.
Signed-off-by: Simon Wunderlich <siwu@hrz.tu-chemnitz.de>
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Without CONFIG_NET_NS, namespace is always &init_net.
Compiler will be able to omit namespace comparisons with this patch.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
dev_set_promiscuity/allmulti might overflow.
Commit: "netdevice: Fix promiscuity and allmulti overflow" in net-next makes
dev_set_promiscuity/allmulti return error number if overflow happened.
Here, we check the positive increment for promiscuity to get error return.
Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com>
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The help text should refer to nflog instead of ulog. Noticed by
Krzysztof Halasa <khc@pm.waw.pl>.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use the STP demux layer for receiving STP PDUs instead of directly
registering with LLC.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Unregistering a bridge device may cause virtual devices stacked on the
bridge, like vlan or macvlan devices, to be unregistered as well.
br_cleanup_bridges() uses for_each_netdev_safe() to iterate over all
devices during cleanup. This is not enough however, if one of the
additionally unregistered devices is next in the list to the bridge
device, it will get freed as well and the iteration continues on
the freed element.
Restart iteration after each bridge device removal from the beginning to
fix this, similar to what rtnl_link_unregister() does.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Acked-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add skb_warn_if_lro() to test whether an skb was received with LRO and
warn if so.
Change br_forward(), ip_forward() and ip6_forward() to call it) and
discard the skb if it returns true.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Large Receive Offload (LRO) is only appropriate for packets that are
destined for the host, and should be disabled if received packets may be
forwarded. It can also confuse the GSO on output.
Add dev_disable_lro() function which uses the appropriate ethtool ops to
disable LRO if enabled.
Add calls to dev_disable_lro() in br_add_if() and functions that enable
IPv4 and IPv6 forwarding.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix bridge netfilter code so that it uses CONFIG_IPV6 as needed:
net/built-in.o: In function `ebt_filter_ip6':
ebt_ip6.c:(.text+0x87c37): undefined reference to `ipv6_skip_exthdr'
net/built-in.o: In function `ebt_log_packet':
ebt_log.c:(.text+0x88dee): undefined reference to `ipv6_skip_exthdr'
make[1]: *** [.tmp_vmlinux1] Error 1
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Normally, the bridge just chooses the smallest mac address as the
bridge id and mac address of bridge device. But if the administrator
has explictly set the interface address then don't change it.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Any frame addressed to link-local addresses should be processed by local
receive path. The earlier code would process them only if STP was enabled.
Since there are other frames like LACP for bonding, we should always
process them.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch removes CVS keywords that weren't updated for a long time
from comments.
Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
It implements matching functions for IPv6 address & traffic class
(merged from the patch sent by Jan Engelhardt [jengelh@computergmbh.de]
http://marc.info/?l=netfilter-devel&m=120182168424052&w=2), protocol,
and layer-4 port id. Corresponding watcher logging function is also
added for IPv6.
Signed-off-by: Kuo-lang Tseng <kuo-lang.tseng@intel.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Even though bridges require 6 fields from struct net_device_stats,
the on-device stats are always there, so we may just use them.
The br_dev_get_stats is no longer required after this.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Move rcu-protected lists from list.h into a new header file rculist.h.
This is done because list are a very used primitive structure all over the
kernel and it's currently impossible to include other header files in this
list.h without creating some circular dependencies.
For example, list.h implements rcu-protected list and uses rcu_dereference()
without including rcupdate.h. It actually compiles because users of
rcu_dereference() are macros. Others RCU functions could be used too but
aren't probably because of this.
Therefore this patch creates rculist.h which includes rcupdates without to
many changes/troubles.
Signed-off-by: Franck Bui-Huu <fbuihuu@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Josh Triplett <josh@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This actually had to be merged with the patch #1, but I decided not to
mix two changes in one patch.
There are already two calls to free_netdev() in there, so merge them
into one.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
In case the register_netdevice() call fails the device is leaked,
since the out: label is just rtnl_unlock()+return.
Free the device.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The forwarding table binary interface (my bad choice), only exposes
the port number of the first 8 bits. The bridge code was limited to
256 ports at the time, but now the kernel supports up 1024 ports, so
the upper bits are lost when doing:
brctl showmacs
The fix is to squeeze the extra bits into small hole left in data
structure, to maintain binary compatiablity.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There is a race condition when unloading bridge and netfilter.
The problem happens if __fake_rtable is in use by a skb
coming in, while someone starts to unload bridge.ko.
br_netfilter_fini() is called at the beginning of unload
in br_deinit() while skbs still are being forwarded and
transferred to local ip stack. Thus there is a possibility
of the __fake_rtable pointer not being removed in a skb that
goes up to ip stack. This results in a kernel panic, as
ip_rcv() calls the input-function of __fake_rtable, which
is NULL.
Moving the call of br_netfilter_fini() to the end of
br_deinit() solves the problem.
Signed-off-by: Bodo Stroesser <bstroesser@fujitsu-siemens.com>
Signed-off-by: Stephen Hemminger <stephen.hemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When device is added to bridge its refcnt is incremented (in new_nbp()), but if
error occurs during further br_add_if() operations this counter is not
decremented back. Fix it by adding dev_put() call in the error path.
Signed-off-by: Volodymyr G Lukiianyk <volodymyrgl@gmail.com>
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds the ebtables nflog watcher to the kernel in order to
allow ebtables log through the nfnetlink_log backend.
Signed-off-by: Peter Warasin <peter@endian.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>
The bridge netfilter code attaches a fake dst_entry with a pointer to a
fake net_device structure to skbs it passes up to IPv4 netfilter. This
leads to crashes when the skb is passed to __ip_route_output_key when
dereferencing the namespace pointer.
Since bridging can currently only operate in the init_net namespace,
the easiest fix for now is to initialize the nd_net pointer of the
fake net_device struct to &init_net.
Should fix bugzilla 10323: http://bugzilla.kernel.org/show_bug.cgi?id=10323
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Introduce per-sock inlines: sock_net(), sock_net_set()
and per-inet_timewait_sock inlines: twsk_net(), twsk_net_set().
Without CONFIG_NET_NS, no namespace other than &init_net exists.
Let's explicitly define them to help compiler optimizations.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Introduce per-net_device inlines: dev_net(), dev_net_set().
Without CONFIG_NET_NS, no namespace other than &init_net exists.
Let's explicitly define them to help compiler optimizations.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
In br_fdb_cleanup() next_timer and this_timer are in jiffies, so they
should be compared using the time_after() macro.
Signed-off-by: Fabio Checconi <fabio@gandalf.sssup.it>
Signed-off-by: Stephen Hemminger <stephen.hemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
__FUNCTION__ is gcc-specific, use __func__
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
(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>
The function ebt_do_table doesn't take NF_DROP as a verdict from the targets.
Signed-off-by: Joonwoo Park <joonwpark81@gmail.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
http://bugzilla.kernel.org/show_bug.cgi?id=9920
The function skb_make_writable returns true or false.
Signed-off-by: Joonwoo Park <joonwpark81@gmail.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Update the MODULES_DESCRIPTION() tags for all Ebtables modules.
Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Before the removal of the deferred output hooks, netoutdev was used in
case of VLANs on top of a bridge to store the VLAN device, so the
deferred hooks would see the correct output device. This isn't
necessary anymore since we're calling the output hooks for the correct
device directly in the IP stack.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Needed to propagate it down to the ip_route_output_flow.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Create a specific helper for netlink kernel socket disposal. This just
let the code look better and provides a ground for proper disposal
inside a namespace.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Tested-by: Alexey Dobriyan <adobriyan@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
In net/bridge/netfilter/ebtables.c,
- remove unused include of a header file (linux/tty.h) and remove the
corresponding comment above it.
Signed-off-by: Rami Rosen <ramirose@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove two unused macros, INV_FLAG and SET_BITMASK
from net/bridge/netfilter/ebt_vlan.c.
Signed-off-by: Rami Rosen <ramirose@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stephen Hemminger
Patrick McHardy