On most systems most of the TCP established/time-wait hash buckets are empty.
When walking the hash table for /proc/net/tcp their read locks would
always be aquired just to find out they're empty. This patch changes the code
to check first if the buckets have any entries before taking the lock, which
is much cheaper than taking a lock. Since the hash tables are large
this makes a measurable difference on processing /proc/net/tcp, 
especially on architectures with slow read_lock (e.g. PPC) 

On a 2GB Core2 system time cat /proc/net/tcp > /dev/null (with a mostly
empty hash table) goes from 0.046s to 0.005s.

On systems with slower atomics (like P4 or POWER4) or larger hash tables
(more RAM) the difference is much higher.

This can be noticeable because there are some daemons around who regularly
scan /proc/net/tcp.

Original idea for this patch from Marcus Meissner, but redone by me.
Signed-off-by: NAndi Kleen <ak@suse.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

After integrating ESP into ip_vs_proto_ah, rename it (and the references to
it) to ip_vs_proto_ah_esp.c and delete the old ip_vs_proto_esp.c.
Signed-off-by: NJulius Volz <juliusv@google.com>
Signed-off-by: NSimon Horman <horms@verge.net.au>

Rename all ah_* functions to ah_esp_* (and adjust comments). Move ESP
protocol definition into ip_vs_proto_ah.c and remove all usage of
ip_vs_proto_esp.c.

Make the compilation of ip_vs_proto_ah.c dependent on a new config
variable, IP_VS_PROTO_AH_ESP, which is selected either by
IP_VS_PROTO_ESP or IP_VS_PROTO_AH. Only compile the selected protocols'
structures within this file.
Signed-off-by: NJulius Volz <juliusv@google.com>
Signed-off-by: NSimon Horman <horms@verge.net.au>

Some duplicated code lying around. Located with my suffix tree
tool.
Signed-off-by: NIlpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NIlpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Large block of code duplication removed.

Sadly, the return value thing is a bit tricky here but it
seems the most sensible way to return positive from validator
on success rather than negative.
Signed-off-by: NIlpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We can't access the cache entry outside of our critical read-locked region,
because someone may free that entry. Also getting an entry under read lock,
then locking for write and trying to delete that entry looks fishy, but should
be no problem here, because we're only comparing a pointer. Also there is no
need for our own rwlock, there is already one in the service structure for use
in the schedulers.
Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Signed-off-by: NSimon Horman <horms@verge.net.au>

We can't access the cache entry outside of our critical read-locked region,
because someone may free that entry. And we also need to check in the critical
region wether the destination is still available, i.e. it's not in the trash.
If we drop our reference counter, the destination can be purged from the trash
at any time. Our caller only guarantees that no destination is moved to the
trash, while we are scheduling. Also there is no need for our own rwlock,
there is already one in the service structure for use in the schedulers.
Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Signed-off-by: NSimon Horman <horms@verge.net.au>

Use incoming network tuple as seed for NAT port randomization.
This avoids concerns of leaking net_random() bits, and also gives better
port distribution. Don't have NAT server, compile tested only.
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>

[ added missing EXPORT_SYMBOL_GPL ]
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch fixes matching of inverted destination address type.
Signed-off-by: NAnders Grafström <grfstrm@users.sourceforge.net>
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Let me first state that disabling the route cache hash rebuild
should not be done without extensive analysis on the risk profile
and careful deliberation.

However, there are times when this can be done safely or for
testing.  For example, when you have mechanisms for ensuring
that offending parties do not exist in your network.

This patch lets the user disable the rebuild if the interval is
set to zero.  This also incidentally fixes a divide-by-zero error
with name-spaces.

In addition, this patch makes the effect of an interval change
immediate rather than it taking effect at the next rebuild as
is currently the case.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

For the sake of clarity, rename __ip_vs_wlc_schedule() in lblc.c to
__ip_vs_lblc_schedule() and the version in lblcr.c to __ip_vs_lblc_schedule().

I guess the original name stuck from a copy and paste.

Cc: Sven Wegener <sven.wegener@stealer.net>
Signed-off-by: NSimon Horman <horms@verge.net.au>

Commit 8ab19ea3 ("ipvs: Fix possible deadlock
in estimator code") fixed a deadlock condition, but that condition can only
happen during unload of IPVS, because during normal operation there is at least
our global stats structure in the estimator list. The mod_timer() and
del_timer_sync() calls are actually initialization and cleanup code in
disguise. Let's make it explicit and move them to their own init and cleanup
function.
Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Signed-off-by: NSimon Horman <horms@verge.net.au>

There are schedulers that only schedule based on data available in the service
or destination structures and they don't need any persistent storage or
initialization routine. These schedulers currently provide dummy functions for
the init_service, update_service and/or done_service functions. For the
init_service and done_service cases we already have code that only calls these
functions, if the scheduler provides them. Do the same for the update_service
case and remove the dummy functions from all schedulers.
Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Signed-off-by: NSimon Horman <horms@verge.net.au>

Add the implementation of the new Generic Netlink interface to IPVS and
keep the old set/getsockopt interface for userspace backwards
compatibility.
Signed-off-by: NJulius Volz <juliusv@google.com>
Acked-by: NSven Wegener <sven.wegener@stealer.net>
Signed-off-by: NSimon Horman <horms@verge.net.au>

This patch makes the multicast socket to be per namespace.

When a network namespace is created, other than the init_net and a
multicast packet is received, the kernel goes to a hang or a kernel panic.

How to reproduce ?

 * create a child network namespace
 * create a pair virtual device veth
    * ip link add type veth
 * move one side to the pair network device to the child namespace
    * ip link set netns <childpid> dev veth1
 * ping -I veth0 224.0.0.1

The bug appears because the function ip_mc_init_dev does not initialize
the different multicast fields as it exits because it is not the init_net.

BUG: soft lockup - CPU#0 stuck for 61s! [avahi-daemon:2695]
Modules linked in:
irq event stamp: 50350
hardirqs last  enabled at (50349): [<c03ee949>] _spin_unlock_irqrestore+0x34/0x39
hardirqs last disabled at (50350): [<c03ec639>] schedule+0x9f/0x5ff
softirqs last  enabled at (45712): [<c0374d4b>] ip_setsockopt+0x8e7/0x909
softirqs last disabled at (45710): [<c03ee682>] _spin_lock_bh+0x8/0x27

Pid: 2695, comm: avahi-daemon Not tainted (2.6.27-rc2-00029-g0872073 #3)
EIP: 0060:[<c03ee47c>] EFLAGS: 00000297 CPU: 0
EIP is at __read_lock_failed+0x8/0x10
EAX: c4f38810 EBX: c4f38810 ECX: 00000000 EDX: c04cc22e
ESI: fb0000e0 EDI: 00000011 EBP: 0f02000a ESP: c4e3faa0
 DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
CR0: 8005003b CR2: 44618a40 CR3: 04e37000 CR4: 000006d0
DR0: 00000000 DR1: 00000000 DR2: 00000000 DR3: 00000000
DR6: ffff0ff0 DR7: 00000400
 [<c02311f8>] ? _raw_read_lock+0x23/0x25
 [<c0390666>] ? ip_check_mc+0x1c/0x83
 [<c036d478>] ? ip_route_input+0x229/0xe92
 [<c022e2e4>] ? trace_hardirqs_on_thunk+0xc/0x10
 [<c0104c9c>] ? do_IRQ+0x69/0x7d
 [<c0102e64>] ? restore_nocheck_notrace+0x0/0xe
 [<c036fdba>] ? ip_rcv+0x227/0x505
 [<c0358764>] ? netif_receive_skb+0xfe/0x2b3
 [<c03588d2>] ? netif_receive_skb+0x26c/0x2b3
 [<c035af31>] ? process_backlog+0x73/0xbd
 [<c035a8cd>] ? net_rx_action+0xc1/0x1ae
 [<c01218a8>] ? __do_softirq+0x7b/0xef
 [<c0121953>] ? do_softirq+0x37/0x4d
 [<c035b50d>] ? dev_queue_xmit+0x3d4/0x40b
 [<c0122037>] ? local_bh_enable+0x96/0xab
 [<c035b50d>] ? dev_queue_xmit+0x3d4/0x40b
 [<c012181e>] ? _local_bh_enable+0x79/0x88
 [<c035fcb8>] ? neigh_resolve_output+0x20f/0x239
 [<c0373118>] ? ip_finish_output+0x1df/0x209
 [<c0373364>] ? ip_dev_loopback_xmit+0x62/0x66
 [<c0371db5>] ? ip_local_out+0x15/0x17
 [<c0372013>] ? ip_push_pending_frames+0x25c/0x2bb
 [<c03891b8>] ? udp_push_pending_frames+0x2bb/0x30e
 [<c038a189>] ? udp_sendmsg+0x413/0x51d
 [<c038a1a9>] ? udp_sendmsg+0x433/0x51d
 [<c038f927>] ? inet_sendmsg+0x35/0x3f
 [<c034f092>] ? sock_sendmsg+0xb8/0xd1
 [<c012d554>] ? autoremove_wake_function+0x0/0x2b
 [<c022e6de>] ? copy_from_user+0x32/0x5e
 [<c022e6de>] ? copy_from_user+0x32/0x5e
 [<c034f238>] ? sys_sendmsg+0x18d/0x1f0
 [<c0175e90>] ? pipe_write+0x3cb/0x3d7
 [<c0170347>] ? do_sync_write+0xbe/0x105
 [<c012d554>] ? autoremove_wake_function+0x0/0x2b
 [<c03503b2>] ? sys_socketcall+0x176/0x1b0
 [<c01085ea>] ? syscall_trace_enter+0x6c/0x7b
 [<c0102e1a>] ? syscall_call+0x7/0xb
Signed-off-by: NDaniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In order to align the coding styles of ip_vs_zero_stats() and
its child-function ip_vs_zero_estimator(), clear ip_vs_stats
members explicitlty rather than doing a limited memset().

This was chosen over modifying ip_vs_zero_estimator() to use
memset() as it is more robust against changes in members
in the relevant structures. memset() would be prefered if
all members of the structure were to be cleared.

Cc: Sven Wegener <sven.wegener@stealer.net>
Signed-off-by: NSimon Horman <horms@verge.net.au>
Signed-off-by: NSven Wegener <sven.wegener@stealer.net>

It's a global variable and automatically initialized to zero. And now we can
also initialize the lock at compile time.
Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Acked-by: NSimon Horman <horms@verge.net.au>

There's no reason for dynamically allocating an estimator object for every
stats object. Directly embed an estimator object into every stats object and
switch to using the kernel-provided list implementation. This makes the code
much simpler and faster, as we do not need to traverse the list of all
estimators to find the one belonging to a stats object. There's no need to use
an rwlock, as we only have one reader. Also reorder the members of the
estimator structure slightly to avoid padding overhead. This can't be done
with the stats object as the members are currently copied to our user space
object via memcpy() and changing it would break ABI.
Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Acked-by: NSimon Horman <horms@verge.net.au>

Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Acked-by: NSimon Horman <horms@verge.net.au>

Being able to discard these functions saves a couple of bytes at runtime. The
cleanup functions can't be annotated with __exit as they are also called from
init functions.
Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Acked-by: NSimon Horman <horms@verge.net.au>

No need to do it at runtime and this saves a couple of bytes in the text
section.
Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Acked-by: NSimon Horman <horms@verge.net.au>

Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Acked-by: NSimon Horman <horms@verge.net.au>

There is a slight chance for a deadlock in the estimator code. We can't call
del_timer_sync() while holding our lock, as the timer might be active and
spinning for the lock on another cpu. Work around this issue by using
try_to_del_timer_sync() and releasing the lock. We could actually delete the
timer outside of our lock, as the add and kill functions are only every called
from userspace via [gs]etsockopt() and are serialized by a mutex, but better
make this explicit.
Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Cc: stable <stable@kernel.org>
Acked-by: NSimon Horman <horms@verge.net.au>

Commit 998e7a76 ("ipvs: Use kthread_run()
instead of doing a double-fork via kernel_thread()") introduced a possible
deadlock in the sync code. We need to use the _bh versions for the lock, as the
lock is also accessed from a bottom half.
Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Acked-by: NSimon Horman <horms@verge.net.au>

The socket lock is there to protect the normal UDP receive path.
Encapsulation UDP sockets don't need that protection.  In fact
the locking is deadly for them as they may contain another UDP
packet within, possibly with the same addresses.

Also the nested bit was copied from TCP.  TCP needs it because
of accept(2) spawning sockets.  This simply doesn't apply to UDP
so I've removed it.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The indentation in part of tcp_minisocks makes it look like one of the if
statements is much more important than it actually is.
Signed-off-by: NAdam Langley <agl@imperialviolet.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If the following packet flow happen, kernel will panic.
MathineA			MathineB
		SYN
	---------------------->    
        	SYN+ACK
	<----------------------
		ACK(bad seq)
	---------------------->
When a bad seq ACK is received, tcp_v4_md5_do_lookup(skb->sk, ip_hdr(skb)->daddr))
is finally called by tcp_v4_reqsk_send_ack(), but the first parameter(skb->sk) is 
NULL at that moment, so kernel panic happens.
This patch fixes this bug.

OOPS output is as following:
[  302.812793] IP: [<c05cfaa6>] tcp_v4_md5_do_lookup+0x12/0x42
[  302.817075] Oops: 0000 [#1] SMP 
[  302.819815] Modules linked in: ipv6 loop dm_multipath rtc_cmos rtc_core rtc_lib pcspkr pcnet32 mii i2c_piix4 parport_pc i2c_core parport ac button ata_piix libata dm_mod mptspi mptscsih mptbase scsi_transport_spi sd_mod scsi_mod crc_t10dif ext3 jbd mbcache uhci_hcd ohci_hcd ehci_hcd [last unloaded: scsi_wait_scan]
[  302.849946] 
[  302.851198] Pid: 0, comm: swapper Not tainted (2.6.27-rc1-guijf #5)
[  302.855184] EIP: 0060:[<c05cfaa6>] EFLAGS: 00010296 CPU: 0
[  302.858296] EIP is at tcp_v4_md5_do_lookup+0x12/0x42
[  302.861027] EAX: 0000001e EBX: 00000000 ECX: 00000046 EDX: 00000046
[  302.864867] ESI: ceb69e00 EDI: 1467a8c0 EBP: cf75f180 ESP: c0792e54
[  302.868333]  DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068
[  302.871287] Process swapper (pid: 0, ti=c0792000 task=c0712340 task.ti=c0746000)
[  302.875592] Stack: c06f413a 00000000 cf75f180 ceb69e00 00000000 c05d0d86 000016d0 ceac5400 
[  302.883275]        c05d28f8 000016d0 ceb69e00 ceb69e20 681bf6e3 00001000 00000000 0a67a8c0 
[  302.890971]        ceac5400 c04250a3 c06f413a c0792eb0 c0792edc cf59a620 cf59a620 cf59a634 
[  302.900140] Call Trace:
[  302.902392]  [<c05d0d86>] tcp_v4_reqsk_send_ack+0x17/0x35
[  302.907060]  [<c05d28f8>] tcp_check_req+0x156/0x372
[  302.910082]  [<c04250a3>] printk+0x14/0x18
[  302.912868]  [<c05d0aa1>] tcp_v4_do_rcv+0x1d3/0x2bf
[  302.917423]  [<c05d26be>] tcp_v4_rcv+0x563/0x5b9
[  302.920453]  [<c05bb20f>] ip_local_deliver_finish+0xe8/0x183
[  302.923865]  [<c05bb10a>] ip_rcv_finish+0x286/0x2a3
[  302.928569]  [<c059e438>] dev_alloc_skb+0x11/0x25
[  302.931563]  [<c05a211f>] netif_receive_skb+0x2d6/0x33a
[  302.934914]  [<d0917941>] pcnet32_poll+0x333/0x680 [pcnet32]
[  302.938735]  [<c05a3b48>] net_rx_action+0x5c/0xfe
[  302.941792]  [<c042856b>] __do_softirq+0x5d/0xc1
[  302.944788]  [<c042850e>] __do_softirq+0x0/0xc1
[  302.948999]  [<c040564b>] do_softirq+0x55/0x88
[  302.951870]  [<c04501b1>] handle_fasteoi_irq+0x0/0xa4
[  302.954986]  [<c04284da>] irq_exit+0x35/0x69
[  302.959081]  [<c0405717>] do_IRQ+0x99/0xae
[  302.961896]  [<c040422b>] common_interrupt+0x23/0x28
[  302.966279]  [<c040819d>] default_idle+0x2a/0x3d
[  302.969212]  [<c0402552>] cpu_idle+0xb2/0xd2
[  302.972169]  =======================
[  302.974274] Code: fc ff 84 d2 0f 84 df fd ff ff e9 34 fe ff ff 83 c4 0c 5b 5e 5f 5d c3 90 90 57 89 d7 56 53 89 c3 50 68 3a 41 6f c0 e8 e9 55 e5 ff <8b> 93 9c 04 00 00 58 85 d2 59 74 1e 8b 72 10 31 db 31 c9 85 f6 
[  303.011610] EIP: [<c05cfaa6>] tcp_v4_md5_do_lookup+0x12/0x42 SS:ESP 0068:c0792e54
[  303.018360] Kernel panic - not syncing: Fatal exception in interrupt
Signed-off-by: NGui Jianfeng <guijianfeng@cn.fujitsu.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Noticed by Paulius Zaleckas.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Here's a revised version, based on Herbert's comments, of a fix for
the ipv6-inner, ipv4-outer interfamily ipsec beet mode. It fixes the
network header adjustment in interfamily, and doesn't reserve space
for the pseudo header anymore when we have ipv6 as the inner family.
Signed-off-by: NJoakim Koskela <jookos@gmail.com>
Acked-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch replaces dst_metric() with dst_mtu() in net/ipv4/route.c.
Signed-off-by: NRami Rosen <ramirose@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 76e6ebfb ("netns: add namespace
parameter to rt_cache_flush") acceses the extra2 parameter of the
ip_default_ttl ctl_table, but it is never set to a meaningful
value. When e84f84f2 ("netns: place
rt_genid into struct net") is applied, we'll oops in
rt_cache_invalidate(). Set extra2 to init_net, to avoid that.
Reported-by: NMarcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: NSven Wegener <sven.wegener@stealer.net>
Tested-by: NMarcin Slusarz <marcin.slusarz@gmail.com>
Acked-by: NDenis V. Lunev <den@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

fix:

net/ipv4/route.c: In function 'ip_static_sysctl_init':
net/ipv4/route.c:3225: error: 'ipv4_route_path' undeclared (first use in this function)
net/ipv4/route.c:3225: error: (Each undeclared identifier is reported only once
net/ipv4/route.c:3225: error: for each function it appears in.)
net/ipv4/route.c:3225: error: 'ipv4_route_table' undeclared (first use in this function)
Signed-off-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

I noticed, looking at tcpdumps, that timewait ACKs were getting sent
with an incorrect MD5 signature when signatures were enabled.

I broke this in 49a72dfb ("tcp: Fix
MD5 signatures for non-linear skbs"). I didn't take into account that
the skb passed to tcp_*_send_ack was the inbound packet, thus the
source and dest addresses need to be swapped when calculating the MD5
pseudoheader.
Signed-off-by: NAdam Langley <agl@imperialviolet.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The thing is that recent_mt_destroy first flushes the entries
from table with the recent_table_flush and only *after* this
removes the proc file, corresponding to that table.

Thus, if we manage to write to this file the '+XXX' command we
will leak some entries. If we manage to write there a 'clean'
command we'll race in two recent_table_flush flows, since the
recent_mt_destroy calls this outside the recent_lock.

The proper solution as I see it is to remove the proc file first
and then go on with flushing the table. This flushing becomes
safe w/o the lock, since the table is already inaccessible from
the outside.
Signed-off-by: NPavel Emelyanov <xemul@openvz.org>
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NHarvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

From a report by Matti Aarnio, and preliminary patch by Adam Langley.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Piss-poor sysctl registration API strikes again, film at 11...

What we really need is _pathname_ required to be present in already
registered table, so that kernel could warn about bad order.  That's the
next target for sysctl stuff (and generally saner and more explicit
order of initialization of ipv[46] internals wouldn't hurt either).

For the time being, here are full fixups required by ..._rotable()
stuff; we make per-net sysctl sets descendents of "ro" one and make sure
that sufficient skeleton is there before we start registering per-net
sysctls.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Piss-poor sysctl registration API strikes again, film at 11...
What we really need is _pathname_ required to be present in
already registered table, so that kernel could warn about bad
order.  That's the next target for sysctl stuff (and generally
saner and more explicit order of initialization of ipv[46]
internals wouldn't hurt either).

For the time being, here are full fixups required by ..._rotable()
stuff; we make per-net sysctl sets descendents of "ro" one and
make sure that sufficient skeleton is there before we start registering
per-net sysctls.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>