linux.git/net/netfilter, branch v4.2-rc2

netfilter: nf_qeueue: Drop queue entries on nf_unregister_hook

2015-06-23T13:23:23+00:00

Add code to nf_unregister_hook to flush the nf_queue when a hook is
unregistered.  This guarantees that the pointer that the nf_queue code
retains into the nf_hook list will remain valid while a packet is
queued.

I tested what would happen if we do not flush queued packets and was
trivially able to obtain the oops below.  All that was required was
to stop the nf_queue listening process, to delete all of the nf_tables,
and to awaken the nf_queue listening process.

> BUG: unable to handle kernel paging request at 0000000100000001
> IP: [<0000000100000001>] 0x100000001
> PGD b9c35067 PUD 0
> Oops: 0010 [#1] SMP
> Modules linked in:
> CPU: 0 PID: 519 Comm: lt-nfqnl_test Not tainted
> task: ffff8800b9c8c050 ti: ffff8800ba9d8000 task.ti: ffff8800ba9d8000
> RIP: 0010:[<0000000100000001>]  [<0000000100000001>] 0x100000001
> RSP: 0018:ffff8800ba9dba40  EFLAGS: 00010a16
> RAX: ffff8800bab48a00 RBX: ffff8800ba9dba90 RCX: ffff8800ba9dba90
> RDX: ffff8800b9c10128 RSI: ffff8800ba940900 RDI: ffff8800bab48a00
> RBP: ffff8800b9c10128 R08: ffffffff82976660 R09: ffff8800ba9dbb28
> R10: dead000000100100 R11: dead000000200200 R12: ffff8800ba940900
> R13: ffffffff8313fd50 R14: ffff8800b9c95200 R15: 0000000000000000
> FS:  00007fb91fc34700(0000) GS:ffff8800bfa00000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 0000000100000001 CR3: 00000000babfb000 CR4: 00000000000007f0
> Stack:
>  ffffffff8206ab0f ffffffff82982240 ffff8800bab48a00 ffff8800b9c100a8
>  ffff8800b9c10100 0000000000000001 ffff8800ba940900 ffff8800b9c10128
>  ffffffff8206bd65 ffff8800bfb0d5e0 ffff8800bab48a00 0000000000014dc0
> Call Trace:
>  [] ? nf_iterate+0x4f/0xa0
>  [] ? nf_reinject+0x125/0x190
>  [] ? nfqnl_recv_verdict+0x255/0x360
>  [] ? nla_parse+0x80/0xf0
>  [] ? nfnetlink_rcv_msg+0x13c/0x240
>  [] ? __memcg_kmem_get_cache+0x4c/0x150
>  [] ? nfnl_lock+0x20/0x20
>  [] ? netlink_rcv_skb+0xa9/0xc0
>  [] ? netlink_unicast+0x12f/0x1c0
>  [] ? netlink_sendmsg+0x28e/0x650
>  [] ? sock_sendmsg+0x44/0x50
>  [] ? ___sys_sendmsg+0x2ab/0x2c0
>  [] ? __wake_up+0x43/0x70
>  [] ? tty_write+0x1c4/0x2a0
>  [] ? __sys_sendmsg+0x44/0x80
>  [] ? system_call_fastpath+0x12/0x6a
> Code:  Bad RIP value.
> RIP  [<0000000100000001>] 0x100000001
>  RSP 
> CR2: 0000000100000001
> ---[ end trace 08eb65d42362793f ]---

Cc: stable@vger.kernel.org
Signed-off-by: "Eric W. Biederman" 
Signed-off-by: David S. Miller

netfilter: nftables: Do not run chains in the wrong network namespace

2015-06-23T13:23:22+00:00

Currenlty nf_tables chains added in one network namespace are being
run in all network namespace.  The issues are myriad with the simplest
being an unprivileged user can cause any network packets to be dropped.

Address this by simply not running nf_tables chains in the wrong
network namespace.

Cc: stable@vger.kernel.org
Signed-off-by: "Eric W. Biederman" 
Acked-by: Pablo Neira Ayuso 
Signed-off-by: David S. Miller

netfilter: use forward declaration instead of including linux/proc_fs.h

2015-06-18T19:14:30+00:00

We don't need to pull the full definitions in that file, a simple forward
declaration is enough.

Moreover, include linux/procfs.h from nf_synproxy_core, otherwise this hits a
compilation error due to missing declarations, ie.

net/netfilter/nf_synproxy_core.c: In function ‘synproxy_proc_init’:
net/netfilter/nf_synproxy_core.c:326:2: error: implicit declaration of function ‘proc_create’ [-Werror=implicit-function-declaration]
  if (!proc_create("synproxy", S_IRUGO, net->proc_net_stat,
  ^

Signed-off-by: Pablo Neira Ayuso 
Signed-off-by: Eric W. Biederman

netfilter: Kill unused copies of RCV_SKB_FAIL

2015-06-18T19:14:27+00:00

This appears to have been a dead macro in both nfnetlink_log.c and
nfnetlink_queue_core.c since these pieces of code were added in 2005.

Signed-off-by: "Eric W. Biederman" 
Signed-off-by: Pablo Neira Ayuso

netfilter: xt_socket: add XT_SOCKET_RESTORESKMARK flag

2015-06-18T11:05:09+00:00

xt_socket is useful for matching sockets with IP_TRANSPARENT and
taking some action on the matching packets. However, it lacks the
ability to match only a small subset of transparent sockets.

Suppose there are 2 applications, each with its own set of transparent
sockets. The first application wants all matching packets dropped,
while the second application wants them forwarded somewhere else.

Add the ability to retore the skb->mark from the sk_mark. The mark
is only restored if a matching socket is found and the transparent /
nowildcard conditions are satisfied.

Now the 2 hypothetical applications can differentiate their sockets
based on a mark value set with SO_MARK.

iptables -t mangle -I PREROUTING -m socket --transparent \
                                           --restore-skmark -j action
iptables -t mangle -A action -m mark --mark 10 -j action2
iptables -t mangle -A action -m mark --mark 11 -j action3

Signed-off-by: Harout Hedeshian 
Signed-off-by: Pablo Neira Ayuso

netfilter: nfnetlink_queue: add security context information

2015-06-18T11:02:24+00:00

This patch adds an additional attribute when sending
packet information via netlink in netfilter_queue module.
It will send additional security context data, so that
userspace applications can verify this context against
their own security databases.

Signed-off-by: Roman Kubiak 
Acked-by: Florian Westphal 
Signed-off-by: Pablo Neira Ayuso

netfilter: nf_tables_netdev: unregister hooks on net_device removal

2015-06-15T21:02:35+00:00

In case the net_device is gone, we have to unregister the hooks and put back
the reference on the net_device object. Once it comes back, register them
again. This also covers the device rename case.

This patch also adds a new flag to indicate that the basechain is disabled, so
their hooks are not registered. This flag is used by the netdev family to
handle the case where the net_device object is gone. Currently this flag is not
exposed to userspace.

Signed-off-by: Pablo Neira Ayuso

netfilter: nf_tables: add nft_register_basechain() and nft_unregister_basechain()

2015-06-15T21:02:33+00:00

This wrapper functions take care of hook registration for basechains.

Signed-off-by: Pablo Neira Ayuso

netfilter: nf_tables: attach net_device to basechain

2015-06-15T21:02:31+00:00

The device is part of the hook configuration, so instead of a global
configuration per table, set it to each of the basechain that we create.

This patch reworks ebddf1a8d78a ("netfilter: nf_tables: allow to bind table to
net_device").

Note that this adds a dev_name field in the nft_base_chain structure which is
required the netdev notification subscription that follows up in a patch to
handle gone net_devices.

Suggested-by: Patrick McHardy 
Signed-off-by: Pablo Neira Ayuso

netfilter: x_tables: remove XT_TABLE_INFO_SZ and a dereference.

2015-06-15T18:19:20+00:00

After Florian patches, there is no need for XT_TABLE_INFO_SZ anymore :
Only one copy of table is kept, instead of one copy per cpu.

We also can avoid a dereference if we put table data right after
xt_table_info. It reduces register pressure and helps compiler.

Then, we attempt a kmalloc() if total size is under order-3 allocation,
to reduce TLB pressure, as in many cases, rules fit in 32 KB.

Signed-off-by: Eric Dumazet 
Cc: Florian Westphal 
Signed-off-by: Pablo Neira Ayuso