David noticed :

------------------
Eric, I was profiling the non-routing-cache case and something that
stuck out is the case of calling inet_getpeer() with create==0.

If an entry is not found, we have to redo the lookup under a spinlock
to make certain that a concurrent writer rebalancing the tree does
not "hide" an existing entry from us.

This makes the case of a create==0 lookup for a not-present entry
really expensive.  It is on the order of 600 cpu cycles on my
Niagara2.

I added a hack to not do the relookup under the lock when create==0
and it now costs less than 300 cycles.

This is now a pretty common operation with the way we handle COW'd
metrics, so I think it's definitely worth optimizing.
-----------------

One solution is to use a seqlock instead of a spinlock to protect struct
inet_peer_base.

After a failed avl tree lookup, we can easily detect if a writer did
some changes during our lookup. Taking the lock and redo the lookup is
only necessary in this case.

Note: Add one private rcu_deref_locked() macro to place in one spot the
access to spinlock included in seqlock.
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Validity of the cached PMTU information is indicated by it's
expiration value being non-zero, just as per dst->expires.

The scheme we will use is that we will remember the pre-ICMP value
held in the metrics or route entry, and then at expiration time
we will restore that value.

In this way PMTU expiration does not kill off the cached route as is
done currently.

Redirect information is permanent, or at least until another redirect
is received.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Future changes will add caching information, and some of
these new elements will be addresses.

Since the family is implicit via the ->daddr.family member,
replicating the family in ever address we store is entirely
redundant.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Like metrics, the ICMP rate limiting bits are cached state about
a destination.  So move it into the inet_peer entries.

If an inet_peer cannot be bound (the reason is memory allocation
failure or similar), the policy is to allow.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Set the RTAX_LOCKED metric to INETPEER_METRICS_NEW (basically,
all ones) on fresh inetpeer entries.

This way code can determine if default metrics have been loaded
in from a routing table entry already.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Family was hard-coded to AF_INET but should be daddr->family.

This fixes crashes when unlinking ipv6 peer entries, since the
unlink code was looking up the base pointer properly.
Reported-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

They are verboten these days.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

They are only allowed on cached ipv6 routes.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add the ipv6 peer tree instance, and adapt remaining
direct references to 'v4_peers' as needed.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Now v4 and v6 addresses will both work properly.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

And make an inet_getpeer_v4() helper, update callers.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently only the v4 aspect is used, but this will change.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Instead of directly accessing "peer", change to code to
operate using a "struct inet_peer_base *" pointer.

This will facilitate the addition of a seperate tree for
ipv6 peer entries.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Adds __rcu annotations to inetpeer
	(struct inet_peer)->avl_left
	(struct inet_peer)->avl_right

This is a tedious cleanup, but removes one smp_wmb() from link_to_pool()
since we now use more self documenting rcu_assign_pointer().

Note the use of RCU_INIT_POINTER() instead of rcu_assign_pointer() in
all cases we dont need a memory barrier.
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Addition of rcu_head to struct inet_peer added 16bytes on 64bit arches.

Thats a bit unfortunate, since old size was exactly 64 bytes.

This can be solved, using an union between this rcu_head an four fields,
that are normally used only when a refcount is taken on inet_peer.
rcu_head is used only when refcnt=-1, right before structure freeing.

Add a inet_peer_refcheck() function to check this assertion for a while.

We can bring back SLAB_HWCACHE_ALIGN qualifier in kmem cache creation.
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Followup of commit aa1039e7 (inetpeer: RCU conversion)

Unused inet_peer entries have a null refcnt.

Using atomic_inc_not_zero() in rcu lookups is not going to work for
them, and slow path is taken.

Fix this using -1 marker instead of 0 for deleted entries.
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

inetpeer currently uses an AVL tree protected by an rwlock.

It's possible to make most lookups use RCU

1) Add a struct rcu_head to struct inet_peer

2) add a lookup_rcu_bh() helper to perform lockless and opportunistic
lookup. This is a normal function, not a macro like lookup().

3) Add a limit to number of links followed by lookup_rcu_bh(). This is
needed in case we fall in a loop.

4) add an smp_wmb() in link_to_pool() right before node insert.

5) make unlink_from_pool() use atomic_cmpxchg() to make sure it can take
last reference to an inet_peer, since lockless readers could increase
refcount, even while we hold peers.lock.

6) Delay struct inet_peer freeing after rcu grace period so that
lookup_rcu_bh() cannot crash.

7) inet_getpeer() first attempts lockless lookup.
   Note this lookup can fail even if target is in AVL tree, but a
concurrent writer can let tree in a non correct form.
   If this attemps fails, lock is taken a regular lookup is performed
again.

8) convert peers.lock from rwlock to a spinlock

9) Remove SLAB_HWCACHE_ALIGN when peer_cachep is created, because
rcu_head adds 16 bytes on 64bit arches, doubling effective size (64 ->
128 bytes)
In a future patch, this is probably possible to revert this part, if rcu
field is put in an union to share space with rid, ip_id_count, tcp_ts &
tcp_ts_stamp. These fields being manipulated only with refcnt > 0.
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Try to reduce cache line contentions in peer management, to reduce IP
defragmentation overhead.

- peer_fake_node is marked 'const' to make sure its not modified.
  (tested with CONFIG_DEBUG_RODATA=y)

- Group variables in two structures to reduce number of dirtied cache
lines. One named "peers" for avl tree root, its number of entries, and
associated lock. (candidate for RCU conversion)

- A second one named "unused_peers" for unused list and its lock

- Add a !list_empty() test in unlink_from_unused() to avoid taking lock
when entry is not unused.

- Use atomic_dec_and_lock() in inet_putpeer() to avoid taking lock in
some cases.
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

While investigating for network latencies, I found inet_getid() was a
contention point for some workloads, as inet_peer_idlock is shared
by all inet_getid() users regardless of peers.

One way to fix this is to make ip_id_count an atomic_t instead
of __u16, and use atomic_add_return().

In order to keep sizeof(struct inet_peer) = 64 on 64bit arches
tcp_ts_stamp is also converted to __u32 instead of "unsigned long".
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NJianjun Kong <jianjun@zeuux.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch removes CVS keywords that weren't updated for a long time
from comments.
Signed-off-by: NAdrian Bunk <bunk@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The inetpeer.c tracks the LRU list of inet_perr-s, but makes
it by hands. Use the list_head-s for this.
Signed-off-by: NPavel Emelyanov <xemul@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

  CC      net/ipv4/inetpeer.o
net/ipv4/inetpeer.c: In function 'unlink_from_pool':
net/ipv4/inetpeer.c:297: warning: the address of 'stack' will always evaluate as 'true'
net/ipv4/inetpeer.c:297: warning: the address of 'stack' will always evaluate as 'true'
net/ipv4/inetpeer.c: In function 'inet_getpeer':
net/ipv4/inetpeer.c:409: warning: the address of 'stack' will always evaluate as 'true'
net/ipv4/inetpeer.c:409: warning: the address of 'stack' will always evaluate as 'true'

"Fix" by checking for != NULL.
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Slab destructors were no longer supported after Christoph's
c59def9f change. They've been
BUGs for both slab and slub, and slob never supported them
either.

This rips out support for the dtor pointer from kmem_cache_create()
completely and fixes up every single callsite in the kernel (there were
about 224, not including the slab allocator definitions themselves,
or the documentation references).
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

1) Some sysctl vars are declared __read_mostly

2) We can avoid updating stack[] when doing an AVL lookup only.

    lookup() macro is extended to receive a second parameter, that may be NULL
in case of a pure lookup (no need to save the AVL path). This removes
unnecessary instructions, because compiler knows if this _stack parameter is
NULL or not.

    text size of net/ipv4/inetpeer.o is 2063 bytes instead of 2107 on x86_64
Signed-off-by: NEric Dumazet <dada1@cosmosbay.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

After Al Viro (finally) succeeded in removing the sched.h #include in module.h
recently, it makes sense again to remove other superfluous sched.h includes.
There are quite a lot of files which include it but don't actually need
anything defined in there.  Presumably these includes were once needed for
macros that used to live in sched.h, but moved to other header files in the
course of cleaning it up.

To ease the pain, this time I did not fiddle with any header files and only
removed #includes from .c-files, which tend to cause less trouble.

Compile tested against 2.6.20-rc2 and 2.6.20-rc2-mm2 (with offsets) on alpha,
arm, i386, ia64, mips, powerpc, and x86_64 with allnoconfig, defconfig,
allmodconfig, and allyesconfig as well as a few randconfigs on x86_64 and all
configs in arch/arm/configs on arm.  I also checked that no new warnings were
introduced by the patch (actually, some warnings are removed that were emitted
by unnecessarily included header files).
Signed-off-by: NTim Schmielau <tim@physik3.uni-rostock.de>
Acked-by: NRussell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Replace all uses of kmem_cache_t with struct kmem_cache.

The patch was generated using the following script:

	#!/bin/sh
	#
	# Replace one string by another in all the kernel sources.
	#

	set -e

	for file in `find * -name "*.c" -o -name "*.h"|xargs grep -l $1`; do
		quilt add $file
		sed -e "1,\$s/$1/$2/g" $file >/tmp/$$
		mv /tmp/$$ $file
		quilt refresh
	done

The script was run like this

	sh replace kmem_cache_t "struct kmem_cache"
Signed-off-by: NChristoph Lameter <clameter@sgi.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

1) shrink struct inet_peer on 64 bits platforms.

This one is interesting - we use net-endian value as search key, but
order the tree by *host-endian* comparisons of keys.  OK since we only
care about lookups.  Annotated inet_getpeer() and friends.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The variable peer_total is protected by a lock.  The volatile marker
makes no sense.  This shaves off 20 bytes on i386.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This changes some simple "if (x) BUG();" statements to "BUG_ON(x);"
Signed-off-by: NKris Katterjohn <kjak@users.sourceforge.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Another spin of Herbert Xu's "safer ip reassembly" patch
for 2.6.16.

(The original patch is here:
http://marc.theaimsgroup.com/?l=linux-netdev&m=112281936522415&w=2
and my only contribution is to have tested it.)

This patch (optionally) does additional checks before accepting IP
fragments, which can greatly reduce the possibility of reassembling
fragments which originated from different IP datagrams.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NArthur Kepner <akepner@sgi.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Clean up timer initialization by introducing DEFINE_TIMER a'la
DEFINE_SPINLOCK.  Build and boot-tested on x86.  A similar patch has been
been in the -RT tree for some time.
Signed-off-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

This patch puts mostly read only data in the right section
(read_mostly), to help sharing of these data between CPUS without
memory ping pongs.

On one of my production machine, tcp_statistics was sitting in a
heavily modified cache line, so *every* SNMP update had to force a
reload.
Signed-off-by: NEric Dumazet <dada1@cosmosbay.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Of this type, mostly:

CHECK   net/ipv6/netfilter.c
net/ipv6/netfilter.c:96:12: warning: symbol 'ipv6_netfilter_init' was not declared. Should it be static?
net/ipv6/netfilter.c:101:6: warning: symbol 'ipv6_netfilter_fini' was not declared. Should it be static?
Signed-off-by: NArnaldo Carvalho de Melo <acme@mandriva.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch contains the following possible cleanups:
- make needlessly global code static
- #if 0 the following unused global function:
  - xfrm4_state.c: xfrm4_state_fini
- remove the following unneeded EXPORT_SYMBOL's:
  - ip_output.c: ip_finish_output
  - ip_output.c: sysctl_ip_default_ttl
  - fib_frontend.c: ip_dev_find
  - inetpeer.c: inet_peer_idlock
  - ip_options.c: ip_options_compile
  - ip_options.c: ip_options_undo
  - net/core/request_sock.c: sysctl_max_syn_backlog
Signed-off-by: NAdrian Bunk <bunk@stusta.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

From: Dave Johnson <djohnson+linux-kernel@sw.starentnetworks.com>

Found this bug while doing some scaling testing that created 500K inet
peers.

peer_check_expire() in net/ipv4/inetpeer.c isn't using inet_peer_gc_mintime
correctly and will end up creating an expire timer with less than the
minimum duration, and even zero/negative if enough active peers are
present.

If >65K peers, the timer will be less than inet_peer_gc_mintime, and with
>70K peers, the timer duration will reach zero and go negative.

The timer handler will continue to schedule another zero/negative timer in
a loop until peers can be aged.  This can continue for at least a few
minutes or even longer if the peers remain active due to arriving packets
while the loop is occurring.

Bug is present in both 2.4 and 2.6.  Same patch will apply to both just
fine.
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.

Let it rip!