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: NDenis V. Lunev <den@openvz.org>
Tested-by: NAlexey Dobriyan <adobriyan@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Network namespace allocates 2 kernel netlink sockets, fibnl &
rtnl. These sockets should be disposed properly, i.e. by
sock_release. Plain sock_put is not enough.
Signed-off-by: NDenis V. Lunev <den@openvz.org>
Tested-by: NAlexey Dobriyan <adobriyan@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The garbage collection function receive the dst_ops structure as
parameter. This is useful for the next incoming patchset because it
will need the dst_ops (there will be several instances) and the
network namespace pointer (contained in the dst_ops).

The protocols which do not take care of the namespaces will not be
impacted by this change (expect for the function signature), they do
just ignore the parameter.
Signed-off-by: NDaniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, sizeof(struct fib_alias) is 24 or 48 bytes on 32/64 bits
arches.

Because of SLAB_HWCACHE_ALIGN requirement, these are rounded to 32 and
64 bytes respectively.

This patch moves rcu to the end of fib_alias, and conditionally
defines it only for CONFIG_IP_FIB_TRIE.

We also remove SLAB_HWCACHE_ALIGN requirement for fib_alias and
fib_node objects because it is not necessary.

(BTW SLUB currently denies it for objects smaller than
cache_line_size() / 2, but not SLAB)

Finally, sizeof(fib_alias) go back to 16 and 32 bytes.

Then, we can embed one fib_alias on each fib_node, to favor locality.
Most of the time access to the fib_alias will be free because one
cache line contains both the list head (fn_alias) and (one of) the
list element.
Signed-off-by: NEric Dumazet <dada1@cosmosbay.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

node_parent() and tnode_get_child() currently use rcu_dereference().

These functions are called from both
- readers only paths (where rcu_dereference() is needed), and
- writer path (where rcu_dereference() is not needed)

To make explicit where rcu_dereference() is really needed, I
introduced new node_parent_rcu() and tnode_get_child_rcu() functions
which use rcu_dereference(), while node_parent() and tnode_get_child()
dont use it.

Then I changed calling sites where rcu_dereference() was really needed
to call the _rcu() variants.

This should have no impact but for alpha architecture, and may help
future sparse checks.
Signed-off-by: NEric Dumazet <dada1@cosmosbay.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Allows to remove five empty implementations.
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Its unused and unlikely to ever be used.
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since there now is generic support for shared sysctl paths, the only
remains are the net/netfilter and net/ipv4/netfilter paths. Move them
to net/netfilter/core.c and net/ipv4/netfilter.c and kill nf_sysctl.c.
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The current TCP RST construction reuses the old packet and can't
deal with IP options as a consequence of that. Construct the
RST from scratch instead.
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch removes inlines except those which are used
by packet matching code and thus are performance-critical.

Before:

$ size */*/*/ip*tables*.o
   text    data     bss     dec     hex filename
   6402     500      16    6918    1b06 net/ipv4/netfilter/ip_tables.o
   7130     500      16    7646    1dde net/ipv6/netfilter/ip6_tables.o

After:

$ size */*/*/ip*tables*.o
   text    data     bss     dec     hex filename
   6307     500      16    6823    1aa7 net/ipv4/netfilter/ip_tables.o
   7010     500      16    7526    1d66 net/ipv6/netfilter/ip6_tables.o
Signed-off-by: NDenys Vlasenko <vda.linux@googlemail.com>
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch moves ipt_iprange to xt_iprange, in preparation for adding
IPv6 support to xt_iprange.
Signed-off-by: NJan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Updates the MODULE_DESCRIPTION() tags for all Netfilter modules,
actually describing what the module does and not just
"netfilter XYZ target".
Signed-off-by: NJan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 88c85d81f74f92371745158aebc5cbf490412002 forgot to remove the
old ipt_TOS file (whose code has been merged into xt_DSCP). Remove
it now.
Signed-off-by: NJan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Most of the netfilter modules are not considered experimental anymore,
the only ones I want to keep marked as EXPERIMENTAL are:

- TCPOPTSTRIP target, which is brand new.

- SANE helper, which is quite new.

- CLUSTERIP target, which I believe hasn't had much testing despite
  being in the kernel for quite a long time.

- SCTP match and conntrack protocol, which are a mess and need to
  be reviewed and cleaned up before I would trust them.
Signed-off-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Initialization of the slab cache's should be done when IP is
initialized to make sure of available memory, and that code can be
marked __init.
Signed-off-by: NStephen Hemminger <stephen.hemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Show number of entries in trie, the size field was being set but never used,
but it only counted leaves, not all entries. Refactor the two cases in
fib_triestat_seq_show into a single routine.

Note: the stat structure was being malloc'd but the stack usage isn't so
high (288 bytes) that it is worth the additional complexity.
Signed-off-by: NStephen Hemminger <stephen.hemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

fib_trie_seq_show() uses two helper functions, rtn_scope() and
rtn_type() that can write to static storage without locking.

Just pass to them a temporary buffer to avoid potential corruption
(probably not triggerable but still...)
Signed-off-by: NEric Dumazet <dada1@cosmosbay.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

inet_confirm_addr can be called with NULL in_dev from arp_ignore iff
scope is RT_SCOPE_LINK.

Lets always pass the device and check for RT_SCOPE_LINK scope inside
inet_confirm_addr. This let us take network namespace from in_device a
need for an additional argument.
Signed-off-by: NDenis V. Lunev <den@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

arp_ignore has two arguments: dev & in_dev. dev is used for
inet_confirm_addr calling only.

inet_confirm_addr, in turn, either gets in_dev from the device passed
or iterates over all network devices if the device passed is NULL. It
seems logical to directly pass in_dev into inet_confirm_addr.
Signed-off-by: NDenis V. Lunev <den@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Some calls in the arp.c have network namespace as an argument. Getting
init_net inside these functions is simply inconsistent. Fix this.
Signed-off-by: NDenis V. Lunev <den@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The neighbour entry will be destroyed in the case of error, so it is
pointless to perform constly routing table lookup in this case.
Signed-off-by: NDenis V. Lunev <den@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

To do so, just register the proper subsystem and create files in
->init callbacks.

No other special per-namespace handling for raw sockets is required.
Signed-off-by: NPavel Emelyanov <xemul@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Happily, in all the rest places (->bind callbacks only), that require the
struct net, we have a socket, so get the net from it.
Signed-off-by: NPavel Emelyanov <xemul@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Pull the struct net pointer up to the showing functions
to filter the sockets depending on their namespaces.
Signed-off-by: NPavel Emelyanov <xemul@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This requires just to pass the appropriate struct net pointer
into __raw_v[46]_lookup and skip sockets that do not belong
to a needed namespace.

The proper net is get from skb->dev in all the cases.
Signed-off-by: NPavel Emelyanov <xemul@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If declared as unsigned short, these fields can overflow, and whole
trie logic is broken. I could not make the machine crash, but some
tnode can never be freed.

Note for 64 bit arches : By reordering t_key and parent in [node,
leaf, tnode] structures, we can use 32 bits hole after t_key so that
sizeof(struct tnode) doesnt change after this patch.
Signed-off-by: NEric Dumazet <dada1@cosmosbay.com>
Signed-off-by: NRobert Olsson <robert.olsson@its.uu.se>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

tnode_alloc() already clears allocated memory, using kcalloc() or
alloc_pages(GFP_KERNEL|__GFP_ZERO, ...)
Signed-off-by: NEric Dumazet <dada1@cosmosbay.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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

In struct tnode, we use two fields of 5 bits for 'pos' and 'bits'.
Switching to plain 'unsigned char' (8 bits) take the same space
because of compiler alignments, and reduce text size by 435 bytes
on i386.

On i386 :
$ size net/ipv4/fib_trie.o.before_patch net/ipv4/fib_trie.o
    text    data     bss     dec     hex filename
   13714       4      64   13782    35d6 net/ipv4/fib_trie.o.before
   13279       4      64   13347    3423 net/ipv4/fib_trie.o
Signed-off-by: NEric Dumazet <dada1@cosmosbay.com>
Acked-by: NStephen Hemminger <stephen.hemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Make FIB TRIE go through sparse checker without warnings.
Signed-off-by: NStephen Hemminger <stephen.hemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The FIB TRIE code has a bunch of statistics, but the code is hidden
behind an ifdef that was never implemented. Since it was dead code, it
was broken as well.

This patch fixes that by making it a config option.
Signed-off-by: NStephen Hemminger <stephen.hemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

printk related cleanups:
 * Get rid of unused printk wrappers.
 * Make bug checks into KERN_WARNING because KERN_DEBUG gets ignored
 * Turn one cryptic old message into something real
 * Make sure all messages have KERN_XXX
Signed-off-by: NStephen Hemminger <stephen.hemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The only error from fib_insert_node is if memory allocation fails, so
instead of passing by reference, just use the convention of returning
NULL.
Signed-off-by: NStephen Hemminger <stephen.hemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use %u instead of %d when printing unsigned values.
Signed-off-by: NStephen Hemminger <stephen.hemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The revision element must of been part of an earlier design, because
currently it is set but never used.
Signed-off-by: NStephen Hemminger <stephen.hemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

trie_init is worthless it is just zeroing stuff that is already zero!
Move the memset() down to make it obvious.
Signed-off-by: NStephen Hemminger <stephen.hemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

net/ipv4/tcp_cong.c:
  tcp_reno_cong_avoid |  -65
 1 function changed, 65 bytes removed, diff: -65

net/ipv4/arp.c:
  arp_ignore |   -5
 1 function changed, 5 bytes removed, diff: -5

net/ipv4/tcp_bic.c:
  bictcp_cong_avoid |  -57
 1 function changed, 57 bytes removed, diff: -57

net/ipv4/tcp_cubic.c:
  bictcp_cong_avoid |  -61
 1 function changed, 61 bytes removed, diff: -61

net/ipv4/tcp_highspeed.c:
  hstcp_cong_avoid |  -63
 1 function changed, 63 bytes removed, diff: -63

net/ipv4/tcp_hybla.c:
  hybla_cong_avoid |  -85
 1 function changed, 85 bytes removed, diff: -85

net/ipv4/tcp_htcp.c:
  htcp_cong_avoid |  -57
 1 function changed, 57 bytes removed, diff: -57

net/ipv4/tcp_veno.c:
  tcp_veno_cong_avoid |  -52
 1 function changed, 52 bytes removed, diff: -52

net/ipv4/tcp_scalable.c:
  tcp_scalable_cong_avoid |  -61
 1 function changed, 61 bytes removed, diff: -61

net/ipv4/tcp_yeah.c:
  tcp_yeah_cong_avoid |  -75
 1 function changed, 75 bytes removed, diff: -75

net/ipv4/tcp_illinois.c:
  tcp_illinois_cong_avoid |  -54
 1 function changed, 54 bytes removed, diff: -54

net/dccp/ccids/ccid3.c:
  ccid3_update_send_interval |   -7
  ccid3_hc_tx_packet_recv    |   +7
 2 functions changed, 7 bytes added, 7 bytes removed, diff: +0

net/ipv4/tcp_cong.c:
  tcp_is_cwnd_limited |  +88
 1 function changed, 88 bytes added, diff: +88

built-in.o:
 14 functions changed, 95 bytes added, 642 bytes removed, diff: -547

...Again some gcc artifacts visible as well.
Signed-off-by: NIlpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

net/ipv4/tcp.c:
  tcp_close_state | -226
  tcp_done        | -145
  tcp_close       | -564
  tcp_disconnect  | -141
 4 functions changed, 1076 bytes removed, diff: -1076

net/ipv4/tcp_input.c:
  tcp_fin               |  -86
  tcp_rcv_state_process | -164
 2 functions changed, 250 bytes removed, diff: -250

net/ipv4/tcp_ipv4.c:
  tcp_v4_connect | -209
 1 function changed, 209 bytes removed, diff: -209

net/ipv4/arp.c:
  arp_ignore |   +5
 1 function changed, 5 bytes added, diff: +5

net/ipv6/tcp_ipv6.c:
  tcp_v6_connect | -158
 1 function changed, 158 bytes removed, diff: -158

net/sunrpc/xprtsock.c:
  xs_sendpages |   -2
 1 function changed, 2 bytes removed, diff: -2

net/dccp/ccids/ccid3.c:
  ccid3_update_send_interval |   +7
 1 function changed, 7 bytes added, diff: +7

net/ipv4/tcp.c:
  tcp_set_state | +238
 1 function changed, 238 bytes added, diff: +238

built-in.o:
 12 functions changed, 250 bytes added, 1695 bytes removed, diff: -1445

I've no explanation why some unrelated changes seem to occur
consistently as well (arp_ignore, ccid3_update_send_interval;
I checked the arp_ignore asm and it seems to be due to some
reordered of operation order causing some extra opcodes to be
generated). Still, the benefits are pretty obvious from the
codiff's results.
Signed-off-by: NIlpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Otherwise we beat heavily on the global tcp_memory atomics
when all of the sockets in the system are slowly sending
perioding packet clumps.

Noticed and suggested by Eric Dumazet.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

I.e. remove the net != &init_net checks from the places, that now can
handle other-than-init net namespace.
Acked-by: NBenjamin Thery <benjamin.thery@bull.net>
Acked-by: NDaniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: NDenis V. Lunev <den@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>