Instead of the actual interface index or name, set destination register
to just 1 or 0 depending on whether the lookup succeeded or not if
NFTA_FIB_F_PRESENT was set in userspace.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
this allows to assign connection tracking helpers to
connections via nft objref infrastructure.
The idea is to first specifiy a helper object:
table ip filter {
ct helper some-name {
type "ftp"
protocol tcp
l3proto ip
}
}
and then assign it via
nft add ... ct helper set "some-name"
helper assignment works for new conntracks only as we cannot expand the
conntrack extension area once it has been committed to the main conntrack
table.
ipv4 and ipv6 protocols are tracked stored separately so
we can also handle families that observe both ipv4 and ipv6 traffic.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
this is needed by the upcoming ct helper object type --
we'd like to be able use the table family (ip, ip6, inet) to figure
out which helper has to be requested.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Element comments may come without any prior set flag, so we have to keep
a list of dummy struct nft_set_ext to keep this information around. This
is only useful for set dumps to userspace. From the packet path, this
set type relies on the bitmap representation. This patch simplifies the
logic since we don't need to allocate the dummy nft_set_ext structure
anymore on the fly at the cost of increasing memory consumption because
of the list of dummy struct nft_set_ext.
Fixes: 665153ff57 ("netfilter: nf_tables: add bitmap set type")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Since the nfct and nfctinfo have been combined, the nf_conn structure
must be at least 8 bytes aligned, as the 3 LSB bits are used for the
nfctinfo. But there's a fake nf_conn structure to denote untracked
connections, which is created by a PER_CPU construct. This does not
guarantee that it will be 8 bytes aligned and can break the logic in
determining the correct nfctinfo.
I triggered this on a 32bit machine with the following error:
BUG: unable to handle kernel NULL pointer dereference at 00000af4
IP: nf_ct_deliver_cached_events+0x1b/0xfb
*pdpt = 0000000031962001 *pde = 0000000000000000
Oops: 0000 [#1] SMP
[Modules linked in: ip6t_REJECT nf_reject_ipv6 nf_conntrack_ipv6 nf_defrag_ipv6 ip6table_filter ip6_tables ipv6 crc_ccitt ppdev r8169 parport_pc parport
OK ]
CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.10.0-test+ #75
Hardware name: MSI MS-7823/CSM-H87M-G43 (MS-7823), BIOS V1.6 02/22/2014
task: c126ec00 task.stack: c1258000
EIP: nf_ct_deliver_cached_events+0x1b/0xfb
EFLAGS: 00010202 CPU: 0
EAX: 0021cd01 EBX: 00000000 ECX: 27b0c767 EDX: 32bcb17a
ESI: f34135c0 EDI: f34135c0 EBP: f2debd60 ESP: f2debd3c
DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068
CR0: 80050033 CR2: 00000af4 CR3: 309a0440 CR4: 001406f0
Call Trace:
<SOFTIRQ>
? ipv6_skip_exthdr+0xac/0xcb
ipv6_confirm+0x10c/0x119 [nf_conntrack_ipv6]
nf_hook_slow+0x22/0xc7
nf_hook+0x9a/0xad [ipv6]
? ip6t_do_table+0x356/0x379 [ip6_tables]
? ip6_fragment+0x9e9/0x9e9 [ipv6]
ip6_output+0xee/0x107 [ipv6]
? ip6_fragment+0x9e9/0x9e9 [ipv6]
dst_output+0x36/0x4d [ipv6]
NF_HOOK.constprop.37+0xb2/0xba [ipv6]
? icmp6_dst_alloc+0x2c/0xfd [ipv6]
? local_bh_enable+0x14/0x14 [ipv6]
mld_sendpack+0x1c5/0x281 [ipv6]
? mark_held_locks+0x40/0x5c
mld_ifc_timer_expire+0x1f6/0x21e [ipv6]
call_timer_fn+0x135/0x283
? detach_if_pending+0x55/0x55
? mld_dad_timer_expire+0x3e/0x3e [ipv6]
__run_timers+0x111/0x14b
? mld_dad_timer_expire+0x3e/0x3e [ipv6]
run_timer_softirq+0x1c/0x36
__do_softirq+0x185/0x37c
? test_ti_thread_flag.constprop.19+0xd/0xd
do_softirq_own_stack+0x22/0x28
</SOFTIRQ>
irq_exit+0x5a/0xa4
smp_apic_timer_interrupt+0x2a/0x34
apic_timer_interrupt+0x37/0x3c
By using DEFINE/DECLARE_PER_CPU_ALIGNED we can enforce at least 8 byte
alignment as all cache line sizes are at least 8 bytes or more.
Fixes: a9e419dc7b ("netfilter: merge ctinfo into nfct pointer storage area")
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Currently, there are two different methods to store an u16 integer to
the u32 data register. For example:
u32 *dest = ®s->data[priv->dreg];
1. *dest = 0; *(u16 *) dest = val_u16;
2. *dest = val_u16;
For method 1, the u16 value will be stored like this, either in
big-endian or little-endian system:
0 15 31
+-+-+-+-+-+-+-+-+-+-+-+-+
| Value | 0 |
+-+-+-+-+-+-+-+-+-+-+-+-+
For method 2, in little-endian system, the u16 value will be the same
as listed above. But in big-endian system, the u16 value will be stored
like this:
0 15 31
+-+-+-+-+-+-+-+-+-+-+-+-+
| 0 | Value |
+-+-+-+-+-+-+-+-+-+-+-+-+
So later we use "memcmp(®s->data[priv->sreg], data, 2);" to do
compare in nft_cmp, nft_lookup expr ..., method 2 will get the wrong
result in big-endian system, as 0~15 bits will always be zero.
For the similar reason, when loading an u16 value from the u32 data
register, we should use "*(u16 *) sreg;" instead of "(u16)*sreg;",
the 2nd method will get the wrong value in the big-endian system.
So introduce some wrapper functions to store/load an u8 or u16
integer to/from the u32 data register, and use them in the right
place.
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Currently we just assume the element key as a u32 integer, regardless of
the set key length.
This is incorrect, for example, the tcp port number is only 16 bits.
So when we use the nft_payload expr to get the tcp dport and store
it to dreg, the dport will be stored at 0~15 bits, and 16~31 bits
will be padded with zero.
So the reg->data[dreg] will be looked like as below:
0 15 31
+-+-+-+-+-+-+-+-+-+-+-+-+
| tcp dport | 0 |
+-+-+-+-+-+-+-+-+-+-+-+-+
But for these big-endian systems, if we treate this register as a u32
integer, the element key will be larger than 65535, so the following
lookup in bitmap set will cause out of bound access.
Another issue is that if we add element with comments in bitmap
set(although the comments will be ignored eventually), the element will
vanish strangely. Because we treate the element key as a u32 integer, so
the comments will become the part of the element key, then the element
key will also be larger than 65535 and out of bound access will happen:
# nft add element t s { 1 comment test }
Since set->klen is 1 or 2, it's fine to treate the element key as a u8 or
u16 integer.
Fixes: 665153ff57 ("netfilter: nf_tables: add bitmap set type")
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Regarding RFC 792, the first 64 bits of the original SCTP datagram's
data could be contained in ICMP packet, such as:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type | Code | Checksum |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| unused |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Internet Header + 64 bits of Original Data Datagram |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
However, according to RFC 4960, SCTP datagram header is as below:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Source Port Number | Destination Port Number |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Verification Tag |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Checksum |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
It means only the first three fields of SCTP header can be carried in
ICMP packet except for Checksum field.
At present in sctp_manip_pkt(), no matter whether the packet is ICMP or
not, it always calculates SCTP packet checksum. However, not only the
calculation of checksum is unnecessary for ICMP, but also it causes
another fatal issue that ICMP packet is dropped. The header size of
SCTP is used to identify whether the writeable length of skb is bigger
than skb->len through skb_make_writable() in sctp_manip_pkt(). But
when it deals with ICMP packet, skb_make_writable() directly returns
false as the writeable length of skb is bigger than skb->len.
Subsequently ICMP is dropped.
Now we correct this misbahavior. When sctp_manip_pkt() handles ICMP
packet, 8 bytes rather than the whole SCTP header size is used to check
if writeable length of skb is overflowed. Meanwhile, as it's meaningless
to calculate checksum when packet is ICMP, the computation of checksum
is ignored as well.
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This new function consolidates set lookup via either name or ID by
introducing a new nft_set_lookup() function. Replace existing spots
where we can use this too.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
When we want to validate the expr's dependency or hooks, we must do two
things to accomplish it. First, write a X_validate callback function
and point ->validate to it. Second, call X_validate in init routine.
This is very common, such as fib, nat, reject expr and so on ...
It is a little ugly, since we will call X_validate in the expr's init
routine, it's better to do it in nf_tables_newexpr. So we can avoid to
do this again and again. After doing this, the second step listed above
is not useful anymore, remove them now.
Patch was tested by nftables/tests/py/nft-test.py and
nftables/tests/shell/run-tests.sh.
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch provides symmetric hash support according to source
ip address and port, and destination ip address and port.
For this purpose, the __skb_get_hash_symmetric() is used to
identify the flow as it uses FLOW_DISSECTOR_F_STOP_AT_FLOW_LABEL
flag by default.
The new attribute NFTA_HASH_TYPE has been included to support
different types of hashing functions. Currently supported
NFT_HASH_JENKINS through jhash and NFT_HASH_SYM through symhash.
The main difference between both types are:
- jhash requires an expression with sreg, symhash doesn't.
- symhash supports modulus and offset, but not seed.
Examples:
nft add rule ip nat prerouting ct mark set jhash ip saddr mod 2
nft add rule ip nat prerouting ct mark set symhash mod 2
By default, jenkins hash will be used if no hash type is
provided for compatibility reasons.
Signed-off-by: Laura Garcia Liebana <laura.garcia@zevenet.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch renames the local nft_hash structure and functions
to nft_jhash in order to prepare the nft_hash module code to
add new hash functions.
Signed-off-by: Laura Garcia Liebana <laura.garcia@zevenet.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Honor NFT_EXTHDR_F_PRESENT flag so we check if the TCP option is
present.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Pull networking fixes from David Miller:
1) Fix double-free in batman-adv, from Sven Eckelmann.
2) Fix packet stats for fast-RX path, from Joannes Berg.
3) Netfilter's ip_route_me_harder() doesn't handle request sockets
properly, fix from Florian Westphal.
4) Fix sendmsg deadlock in rxrpc, from David Howells.
5) Add missing RCU locking to transport hashtable scan, from Xin Long.
6) Fix potential packet loss in mlxsw driver, from Ido Schimmel.
7) Fix race in NAPI handling between poll handlers and busy polling,
from Eric Dumazet.
8) TX path in vxlan and geneve need proper RCU locking, from Jakub
Kicinski.
9) SYN processing in DCCP and TCP need to disable BH, from Eric
Dumazet.
10) Properly handle net_enable_timestamp() being invoked from IRQ
context, also from Eric Dumazet.
11) Fix crash on device-tree systems in xgene driver, from Alban Bedel.
12) Do not call sk_free() on a locked socket, from Arnaldo Carvalho de
Melo.
13) Fix use-after-free in netvsc driver, from Dexuan Cui.
14) Fix max MTU setting in bonding driver, from WANG Cong.
15) xen-netback hash table can be allocated from softirq context, so use
GFP_ATOMIC. From Anoob Soman.
16) Fix MAC address change bug in bgmac driver, from Hari Vyas.
17) strparser needs to destroy strp_wq on module exit, from WANG Cong.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (69 commits)
strparser: destroy workqueue on module exit
sfc: fix IPID endianness in TSOv2
sfc: avoid max() in array size
rds: remove unnecessary returned value check
rxrpc: Fix potential NULL-pointer exception
nfp: correct DMA direction in XDP DMA sync
nfp: don't tell FW about the reserved buffer space
net: ethernet: bgmac: mac address change bug
net: ethernet: bgmac: init sequence bug
xen-netback: don't vfree() queues under spinlock
xen-netback: keep a local pointer for vif in backend_disconnect()
netfilter: nf_tables: don't call nfnetlink_set_err() if nfnetlink_send() fails
netfilter: nft_set_rbtree: incorrect assumption on lower interval lookups
netfilter: nf_conntrack_sip: fix wrong memory initialisation
can: flexcan: fix typo in comment
can: usb_8dev: Fix memory leak of priv->cmd_msg_buffer
can: gs_usb: fix coding style
can: gs_usb: Don't use stack memory for USB transfers
ixgbe: Limit use of 2K buffers on architectures with 256B or larger cache lines
ixgbe: update the rss key on h/w, when ethtool ask for it
...
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
The following patchset contains Netfilter fixes for your net tree,
they are:
1) Missing check for full sock in ip_route_me_harder(), from
Florian Westphal.
2) Incorrect sip helper structure initilization that breaks it when
several ports are used, from Christophe Leroy.
3) Fix incorrect assumption when looking up for matching with adjacent
intervals in the nft_set_rbtree.
4) Fix broken netlink event error reporting in nf_tables that results
in misleading ESRCH errors propagated to userspace listeners.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The underlying nlmsg_multicast() already sets sk->sk_err for us to
notify socket overruns, so we should not do anything with this return
value. So we just call nfnetlink_set_err() if:
1) We fail to allocate the netlink message.
or
2) We don't have enough space in the netlink message to place attributes,
which means that we likely need to allocate a larger message.
Before this patch, the internal ESRCH netlink error code was propagated
to userspace, which is quite misleading. Netlink semantics mandate that
listeners just hit ENOBUFS if the socket buffer overruns.
Reported-by: Alexander Alemayhu <alexander@alemayhu.com>
Tested-by: Alexander Alemayhu <alexander@alemayhu.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
In case of adjacent ranges, we may indeed see either the high part of
the range in first place or the low part of it. Remove this incorrect
assumption, let's make sure we annotate the low part of the interval in
case of we have adjacent interva intervals so we hit a matching in
lookups.
Reported-by: Simon Hanisch <hanisch@wh2.tu-dresden.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
In commit 82de0be686 ("netfilter: Add helper array
register/unregister functions"),
struct nf_conntrack_helper sip[MAX_PORTS][4] was changed to
sip[MAX_PORTS * 4], so the memory init should have been changed to
memset(&sip[4 * i], 0, 4 * sizeof(sip[i]));
But as the sip[] table is allocated in the BSS, it is already set to 0
Fixes: 82de0be686 ("netfilter: Add helper array register/unregister functions")
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Add #include <linux/cred.h> dependencies to all .c files rely on sched.h
doing that for them.
Note that even if the count where we need to add extra headers seems high,
it's still a net win, because <linux/sched.h> is included in over
2,200 files ...
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull networking fixes from David Miller:
1) Don't save TIPC header values before the header has been validated,
from Jon Paul Maloy.
2) Fix memory leak in RDS, from Zhu Yanjun.
3) We miss to initialize the UID in the flow key in some paths, from
Julian Anastasov.
4) Fix latent TOS masking bug in the routing cache removal from years
ago, also from Julian.
5) We forget to set the sockaddr port in sctp_copy_local_addr_list(),
fix from Xin Long.
6) Missing module ref count drop in packet scheduler actions, from
Roman Mashak.
7) Fix RCU annotations in rht_bucket_nested, from Herbert Xu.
8) Fix use after free which happens because L2TP's ipv4 support returns
non-zero values from it's backlog_rcv function which ipv4 interprets
as protocol values. Fix from Paul Hüber.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (35 commits)
qed: Don't use attention PTT for configuring BW
qed: Fix race with multiple VFs
l2tp: avoid use-after-free caused by l2tp_ip_backlog_recv
xfrm: provide correct dst in xfrm_neigh_lookup
rhashtable: Fix RCU dereference annotation in rht_bucket_nested
rhashtable: Fix use before NULL check in bucket_table_free
net sched actions: do not overwrite status of action creation.
rxrpc: Kernel calls get stuck in recvmsg
net sched actions: decrement module reference count after table flush.
lib: Allow compile-testing of parman
ipv6: check sk sk_type and protocol early in ip_mroute_set/getsockopt
sctp: set sin_port for addr param when checking duplicate address
net/mlx4_en: fix overflow in mlx4_en_init_timestamp()
netfilter: nft_set_bitmap: incorrect bitmap size
net: s2io: fix typo argumnet argument
net: vxge: fix typo argumnet argument
netfilter: nf_ct_expect: Change __nf_ct_expect_check() return value.
ipv4: mask tos for input route
ipv4: add missing initialization for flowi4_uid
lib: fix spelling mistake: "actualy" -> "actually"
...
Now that %z is standartised in C99 there is no reason to support %Z.
Unlike %L it doesn't even make format strings smaller.
Use BUILD_BUG_ON in a couple ATM drivers.
In case anyone didn't notice lib/vsprintf.o is about half of SLUB which
is in my opinion is quite an achievement. Hopefully this patch inspires
someone else to trim vsprintf.c more.
Link: http://lkml.kernel.org/r/20170103230126.GA30170@avx2
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Fix typos and add the following to the scripts/spelling.txt:
aligment||alignment
I did not touch the "N_BYTE_ALIGMENT" macro in
drivers/net/wireless/realtek/rtlwifi/wifi.h to avoid unpredictable
impact.
I fixed "_aligment_handler" in arch/openrisc/kernel/entry.S because
it is surrounded by #if 0 ... #endif. It is surely safe and I
confirmed "_alignment_handler" is correct.
I also fixed the "controler" I found in the same hunk in
arch/openrisc/kernel/head.S.
Link: http://lkml.kernel.org/r/1481573103-11329-8-git-send-email-yamada.masahiro@socionext.com
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Fix typos and add the following to the scripts/spelling.txt:
an user||a user
an userspace||a userspace
I also added "userspace" to the list since it is a common word in Linux.
I found some instances for "an userfaultfd", but I did not add it to the
list. I felt it is endless to find words that start with "user" such as
"userland" etc., so must draw a line somewhere.
Link: http://lkml.kernel.org/r/1481573103-11329-4-git-send-email-yamada.masahiro@socionext.com
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
priv->bitmap_size stores the real bitmap size, instead of the full
struct nft_bitmap object.
Fixes: 665153ff57 ("netfilter: nf_tables: add bitmap set type")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Commit 4dee62b1b9 ("netfilter: nf_ct_expect: nf_ct_expect_insert()
returns void") inadvertently changed the successful return value of
nf_ct_expect_related_report() from 0 to 1 due to
__nf_ct_expect_check() returning 1 on success. Prevent this
regression in the future by changing the return value of
__nf_ct_expect_check() to 0 on success.
Signed-off-by: Jarno Rajahalme <jarno@ovn.org>
Acked-by: Joe Stringer <joe@ovn.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Commit 4dee62b1b9 ("netfilter: nf_ct_expect: nf_ct_expect_insert()
returns void") inadvertently changed the successful return value of
nf_ct_expect_related_report() from 0 to 1, which caused openvswitch
conntrack integration fail in FTP test cases.
Fix this by always returning zero on the success code path.
Fixes: 4dee62b1b9 ("netfilter: nf_ct_expect: nf_ct_expect_insert() returns void")
Signed-off-by: Jarno Rajahalme <jarno@ovn.org>
Acked-by: Joe Stringer <joe@ovn.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Dan reports:
net/netfilter/nft_ct.c:549 nft_ct_set_init()
error: uninitialized symbol 'len'.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Fixes: edee4f1e92 ("netfilter: nft_ct: add zone id set support")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
The following patchset contains Netfilter fixes for your net tree,
they are:
1) Revisit warning logic when not applying default helper assignment.
Jiri Kosina considers we are breaking existing setups and not warning
our users accordinly now that automatic helper assignment has been
turned off by default. So let's make him happy by spotting the warning
by when we find a helper but we cannot attach, instead of warning on the
former deprecated behaviour. Patch from Jiri Kosina.
2) Two patches to fix regression in ctnetlink interfaces with
nfnetlink_queue. Specifically, perform more relaxed in CTA_STATUS
and do not bail out if CTA_HELP indicates the same helper that we
already have. Patches from Kevin Cernekee.
3) A couple of bugfixes for ipset via Jozsef Kadlecsik. Due to wrong
index logic in hash set types and null pointer exception in the
list:set type.
4) hashlimit bails out with correct userspace parameters due to wrong
arithmetics in the code that avoids "divide by zero" when
transforming the userspace timing in milliseconds to token credits.
Patch from Alban Browaeys.
5) Fix incorrect NFQA_VLAN_MAX definition, patch from
Ken-ichirou MATSUZAWA.
6) Don't not declare nfnetlink batch error list as static, since this
may be used by several subsystems at the same time. Patch from
Liping Zhang.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Jozsef Kadlecsik says:
====================
ipset patches for nf
Please apply the next patches for ipset in your nf branch.
Both patches should go into the stable kernel branches as well,
because these are important bugfixes:
* Sometimes valid entries in hash:* types of sets were evicted
due to a typo in an index. The wrong evictions happen when
entries are deleted from the set and the bucket is shrinked.
Bug was reported by Eric Ewanco and the patch fixes
netfilter bugzilla id #1119.
* Fixing of a null pointer exception when someone wants to add an
entry to an empty list type of set and specifies an add before/after
option. The fix is from Vishwanath Pai.
====================
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Otherwise, different subsys will race to access the err_list, with holding
the different nfnl_lock(subsys_id).
But this will not happen now, since ->call_batch is only implemented by
nftables, so the err_list is protected by nfnl_lock(NFNL_SUBSYS_NFTABLES).
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Diving the divider by the multiplier before applying to the input.
When this would "divide by zero", divide the multiplier by the divider
first then multiply the input by this value.
Currently user2creds outputs zero when input value is bigger than the
number of slices and lower than scale.
This as then user input is applied an integer divide operation to
a number greater than itself (scale).
That rounds up to zero, then we multiply zero by the credits slice size.
iptables -t filter -I INPUT --protocol tcp --match hashlimit
--hashlimit 40/second --hashlimit-burst 20 --hashlimit-mode srcip
--hashlimit-name syn-flood --jump RETURN
thus trigger the overflow detection code:
xt_hashlimit: overflow, try lower: 25000/20
(25000 as hashlimit avg and 20 the burst)
Here:
134217 slices of (HZ * CREDITS_PER_JIFFY) size.
500000 is user input value
1000000 is XT_HASHLIMIT_SCALE_v2
gives: 0 as user2creds output
Setting burst to "1" typically solve the issue ...
but setting it to "40" does too !
This is on 32bit arch calling into revision 2 of hashlimit.
Signed-off-by: Alban Browaeys <alban.browaeys@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
If we use before/after to add an element to an empty list it will cause
a kernel panic.
$> cat crash.restore
create a hash:ip
create b hash:ip
create test list:set timeout 5 size 4
add test b before a
$> ipset -R < crash.restore
Executing the above will crash the kernel.
Signed-off-by: Vishwanath Pai <vpai@akamai.com>
Reviewed-by: Josh Hunt <johunt@akamai.com>
Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Wrong index was used and therefore when shrinking a hash bucket at
deleting an entry, valid entries could be evicted as well.
Thanks to Eric Ewanco for the thorough bugreport.
Fixes netfilter bugzilla #1119
Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
This new attribute allows us to uniquely identify a rule in transaction.
Robots may trigger an insertion followed by deletion in a batch, in that
scenario we still don't have a public rule handle that we can use to
delete the rule. This is similar to the NFTA_SET_ID attribute that
allows us to refer to an anonymous set from a batch.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch implements the check generation id as provided by nfnetlink.
This allows us to reject ruleset updates against stale baseline, so
userspace can retry update with a fresh ruleset cache.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch allows userspace to specify the generation ID that has been
used to build an incremental batch update.
If userspace specifies the generation ID in the batch message as
attribute, then nfnetlink compares it to the current generation ID so
you make sure that you work against the right baseline. Otherwise, bail
out with ERESTART so userspace knows that its changeset is stale and
needs to respin. Userspace can do this transparently at the cost of
taking slightly more time to refresh caches and rework the changeset.
This check is optional, if there is no NFNL_BATCH_GENID attribute in the
batch begin message, then no check is performed.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Because nf_ct_expect_insert() always succeeds now, its return value can
be just void instead of int. And remove code that checks for its return
value.
Signed-off-by: Gao Feng <fgao@ikuai8.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
timer_del() followed by timer_add() can be replaced by
mod_timer_pending().
Signed-off-by: Gao Feng <fgao@ikuai8.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch implements the kernel side of the TCP option patch.
Signed-off-by: Manuel Messner <mm@skelett.io>
Reviewed-by: Florian Westphal <fw@strlen.de>
Acked-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
zones allow tracking multiple connections sharing identical tuples,
this is needed e.g. when tracking distinct vlans with overlapping ip
addresses (conntrack is l2 agnostic).
Thus the zone has to be set before the packet is picked up by the
connection tracker. This is done by means of 'conntrack templates' which
are conntrack structures used solely to pass this info from one netfilter
hook to the next.
The iptables CT target instantiates these connection tracking templates
once per rule, i.e. the template is fixed/tied to particular zone, can
be read-only and therefore be re-used by as many skbs simultaneously as
needed.
We can't follow this model because we want to take the zone id from
an sreg at rule eval time so we could e.g. fill in the zone id from
the packets vlan id or a e.g. nftables key : value maps.
To avoid cost of per packet alloc/free of the template, use a percpu
template 'scratch' object and use the refcount to detect the (unlikely)
case where the template is still attached to another skb (i.e., previous
skb was nfqueued ...).
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Next patch will add ZONE_ID set support which will need similar
error unwind (put operation) as conntrack labels.
Prepare for this: remove the 'label_got' boolean in favor
of a switch statement that can be extended in next patch.
As we already have that in the set_destroy function place that in
a separate function and call it from the set init function.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Just like with counters the direction attribute is optional.
We set priv->dir to MAX unconditionally to avoid duplicating the assignment
for all keys with optional direction.
For keys where direction is mandatory, existing code already returns
an error.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch adds a new bitmap set type. This bitmap uses two bits to
represent one element. These two bits determine the element state in the
current and the future generation that fits into the nf_tables commit
protocol. When dumping elements back to userspace, the two bits are
expanded into a struct nft_set_ext object.
If no NFTA_SET_DESC_SIZE is specified, the existing automatic set
backend selection prefers bitmap over hash in case of keys whose size is
<= 16 bit. If the set size is know, the bitmap set type is selected if
with 16 bit kets and more than 390 elements in the set, otherwise the
hash table set implementation is used.
For 8 bit keys, the bitmap consumes 66 bytes. For 16 bit keys, the
bitmap takes 16388 bytes.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
The space notation allows us to classify the set backend implementation
based on the amount of required memory. This provides an order of the
set representation scalability in terms of memory. The size field is
still left in place so use this if the userspace provides no explicit
number of elements, so we cannot calculate the real memory that this set
needs. This also helps us break ties in the set backend selection
routine, eg. two backend implementations provide the same performance.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Use lookup as field name instead, to prepare the introduction of the
memory class in a follow up patch.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This provides context to walk callback iterator, thus, we know if the
walk happens from the set flush path. This is required by the new bitmap
set type coming in a follow up patch which has no real struct
nft_set_ext, so it has to allocate it based on the two bit compact
element representation.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Although semantics are similar to deactivate() with no implicit element
lookup, this is only called from the set flush path, so better rename
this to flush().
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
If NFT_EXTHDR_F_PRESENT is set, exthdr will not copy any header field
data into *dest, but instead set it to 1 if the header is found and 0
otherwise.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Prior to Linux 4.4, it was usually harmless to send a CTA_HELP attribute
containing the name of the current helper. That is no longer the case:
as of Linux 4.4, if ctnetlink_change_helper() returns an error from
the ct->master check, processing of the request will fail, skipping the
NFQA_EXP attribute (if present).
This patch changes the behavior to improve compatibility with user
programs that expect the kernel interface to work the way it did prior
to Linux 4.4. If a user program specifies CTA_HELP but the argument
matches the current conntrack helper name, ignore it instead of generating
an error.
Signed-off-by: Kevin Cernekee <cernekee@chromium.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
The libnetfilter_conntrack userland library always sets IPS_CONFIRMED
when building a CTA_STATUS attribute. If this toggles the bit from
0->1, the parser will return an error. On Linux 4.4+ this will cause any
NFQA_EXP attribute in the packet to be ignored. This breaks conntrackd's
userland helpers because they operate on unconfirmed connections.
Instead of returning -EBUSY if the user program asks to modify an
unchangeable bit, simply ignore the change.
Also, fix the logic so that user programs are allowed to clear
the bits that they are allowed to change.
Signed-off-by: Kevin Cernekee <cernekee@chromium.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Commit 3bb398d925 ("netfilter: nf_ct_helper: disable automatic helper
assignment") is causing behavior regressions in firewalls, as traffic
handled by conntrack helpers is now by default not passed through even
though it was before due to missing CT targets (which were not necessary
before this commit).
The default had to be switched off due to security reasons [1] [2] and
therefore should stay the way it is, but let's be friendly to firewall
admins and issue a warning the first time we're in situation where packet
would be likely passed through with the old default but we're likely going
to drop it on the floor now.
Rewrite the code a little bit as suggested by Linus, so that we avoid
spaghettiing the code even more -- namely the whole decision making
process regarding helper selection (either automatic or not) is being
separated, so that the whole logic can be simplified and code (condition)
duplication reduced.
[1] https://cansecwest.com/csw12/conntrack-attack.pdf
[2] https://home.regit.org/netfilter-en/secure-use-of-helpers/
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Pablo Neira Ayuso says:
====================
Netfilter updates for net-next
The following patchset contains Netfilter updates for your net-next
tree, they are:
1) Stash ctinfo 3-bit field into pointer to nf_conntrack object from
sk_buff so we only access one single cacheline in the conntrack
hotpath. Patchset from Florian Westphal.
2) Don't leak pointer to internal structures when exporting x_tables
ruleset back to userspace, from Willem DeBruijn. This includes new
helper functions to copy data to userspace such as xt_data_to_user()
as well as conversions of our ip_tables, ip6_tables and arp_tables
clients to use it. Not surprinsingly, ebtables requires an ad-hoc
update. There is also a new field in x_tables extensions to indicate
the amount of bytes that we copy to userspace.
3) Add nf_log_all_netns sysctl: This new knob allows you to enable
logging via nf_log infrastructure for all existing netnamespaces.
Given the effort to provide pernet syslog has been discontinued,
let's provide a way to restore logging using netfilter kernel logging
facilities in trusted environments. Patch from Michal Kubecek.
4) Validate SCTP checksum from conntrack helper, from Davide Caratti.
5) Merge UDPlite conntrack and NAT helpers into UDP, this was mostly
a copy&paste from the original helper, from Florian Westphal.
6) Reset netfilter state when duplicating packets, also from Florian.
7) Remove unnecessary check for broadcast in IPv6 in pkttype match and
nft_meta, from Liping Zhang.
8) Add missing code to deal with loopback packets from nft_meta when
used by the netdev family, also from Liping.
9) Several cleanups on nf_tables, one to remove unnecessary check from
the netlink control plane path to add table, set and stateful objects
and code consolidation when unregister chain hooks, from Gao Feng.
10) Fix harmless reference counter underflow in IPVS that, however,
results in problems with the introduction of the new refcount_t
type, from David Windsor.
11) Enable LIBCRC32C from nf_ct_sctp instead of nf_nat_sctp,
from Davide Caratti.
12) Missing documentation on nf_tables uapi header, from Liping Zhang.
13) Use rb_entry() helper in xt_connlimit, from Geliang Tang.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 69b34fb996 ("netfilter: xt_LOG: add net namespace support for
xt_LOG") disabled logging packets using the LOG target from non-init
namespaces. The motivation was to prevent containers from flooding
kernel log of the host. The plan was to keep it that way until syslog
namespace implementation allows containers to log in a safe way.
However, the work on syslog namespace seems to have hit a dead end
somewhere in 2013 and there are users who want to use xt_LOG in all
network namespaces. This patch allows to do so by setting
/proc/sys/net/netfilter/nf_log_all_netns
to a nonzero value. This sysctl is only accessible from init_net so that
one cannot switch the behaviour from inside a container.
Signed-off-by: Michal Kubecek <mkubecek@suse.cz>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Currently, the ip_vs_dest cache frees ip_vs_dest objects when their
reference count becomes < 0. Aside from not being semantically sound,
this is problematic for the new type refcount_t, which will be introduced
shortly in a separate patch. refcount_t is the new kernel type for
holding reference counts, and provides overflow protection and a
constrained interface relative to atomic_t (the type currently being
used for kernel reference counts).
Per Julian Anastasov: "The problem is that dest_trash currently holds
deleted dests (unlinked from RCU lists) with refcnt=0." Changing
dest_trash to hold dest with refcnt=1 will allow us to free ip_vs_dest
structs when their refcnt=0, in ip_vs_dest_put_and_free().
Signed-off-by: David Windsor <dwindsor@gmail.com>
Signed-off-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
After this change conntrack operations (lookup, creation, matching from
ruleset) only access one instead of two sk_buff cache lines.
This works for normal conntracks because those are allocated from a slab
that guarantees hw cacheline or 8byte alignment (whatever is larger)
so the 3 bits needed for ctinfo won't overlap with nf_conn addresses.
Template allocation now does manual address alignment (see previous change)
on arches that don't have sufficent kmalloc min alignment.
Some spots intentionally use skb->_nfct instead of skb_nfct() helpers,
this is to avoid undoing the skb_nfct() use when we remove untracked
conntrack object in the future.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
The next change will merge skb->nfct pointer and skb->nfctinfo
status bits into single skb->_nfct (unsigned long) area.
For this to work nf_conn addresses must always be aligned at least on
an 8 byte boundary since we will need the lower 3bits to store nfctinfo.
Conntrack templates are allocated via kmalloc.
kbuild test robot reported
BUILD_BUG_ON failed: NFCT_INFOMASK >= ARCH_KMALLOC_MINALIGN
on v1 of this patchset, so not all platforms meet this requirement.
Do manual alignment if needed, the alignment offset is stored in the
nf_conn entry protocol area. This works because templates are not
handed off to L4 protocol trackers.
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Add a helper to assign a nf_conn entry and the ctinfo bits to an sk_buff.
This avoids changing code in followup patch that merges skb->nfct and
skb->nfctinfo into skb->_nfct.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Followup patch renames skb->nfct and changes its type so add a helper to
avoid intrusive rename change later.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Next patch makes direct skb->nfct access illegal, reduce noise
in next patch by using accessors we already have.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
It is never accessed for reading and the only places that write to it
are the icmp(6) handlers, which also set skb->nfct (and skb->nfctinfo).
The conntrack core specifically checks for attached skb->nfct after
->error() invocation and returns early in this case.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
If something fails in nf_tables_table_enable(), it unregisters the
chains. But the rollback code is the same as nf_tables_table_disable()
almostly, except there is one counter check. Now create one wrapper
function to eliminate the duplicated codes.
Signed-off-by: Feng <fgao@ikuai8.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Add missing set->ndeact update on each deactivated element from the set
flush path. Otherwise, sets with fixed size break after flush since
accounting breaks.
# nft add set x y { type ipv4_addr\; size 2\; }
# nft add element x y { 1.1.1.1 }
# nft add element x y { 1.1.1.2 }
# nft flush set x y
# nft add element x y { 1.1.1.1 }
<cmdline>:1:1-28: Error: Could not process rule: Too many open files in system
Fixes: 8411b6442e ("netfilter: nf_tables: support for set flushing")
Reported-by: Elise Lennion <elise.lennion@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
If the element exists and no NLM_F_EXCL is specified, do not bump
set->nelems, otherwise we leak one set element slot. This problem
amplifies if the set is full since the abort path always decrements the
counter for the -ENFILE case too, giving one spare extra slot.
Fix this by moving set->nelems update to nft_add_set_elem() after
successful element insertion. Moreover, remove the element if the set is
full so there is no need to rely on the abort path to undo things
anymore.
Fixes: c016c7e45d ("netfilter: nf_tables: honor NLM_F_EXCL flag in set element insertion")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
First, log prefix will be truncated to NF_LOG_PREFIXLEN-1, i.e. 127,
at nf_log_packet(), so the extra part is useless.
Second, after adding a log rule with a very very long prefix, we will
fail to dump the nft rules after this _special_ one, but acctually,
they do exist. For example:
# name_65000=$(printf "%0.sQ" {1..65000})
# nft add rule filter output log prefix "$name_65000"
# nft add rule filter output counter
# nft add rule filter output counter
# nft list chain filter output
table ip filter {
chain output {
type filter hook output priority 0; policy accept;
}
}
So now, restrict the log prefix length to NF_LOG_PREFIXLEN-1.
Fixes: 96518518cc ("netfilter: add nftables")
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Add net.ipv4.ip_unprivileged_port_start, which is a per namespace sysctl
that denotes the first unprivileged inet port in the namespace. To
disable all privileged ports set this to zero. It also checks for
overlap with the local port range. The privileged and local range may
not overlap.
The use case for this change is to allow containerized processes to bind
to priviliged ports, but prevent them from ever being allowed to modify
their container's network configuration. The latter is accomplished by
ensuring that the network namespace is not a child of the user
namespace. This modification was needed to allow the container manager
to disable a namespace's priviliged port restrictions without exposing
control of the network namespace to processes in the user namespace.
Signed-off-by: Krister Johansen <kjlx@templeofstupid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, if the user add a stateful object with the name size exceed
NFT_OBJ_MAXNAMELEN - 1 (i.e. 31), we truncate it down to 31 silently.
This is not friendly, furthermore, this will cause duplicated stateful
objects when the first 31 characters of the name is same. So limit the
stateful object's name size to NFT_OBJ_MAXNAMELEN - 1.
After apply this patch, error message will be printed out like this:
# name_32=$(printf "%0.sQ" {1..32})
# nft add counter filter $name_32
<cmdline>:1:1-52: Error: Could not process rule: Numerical result out
of range
add counter filter QQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQ
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Also this patch cleans up the codes which missing the name size limit
validation in nftables.
Fixes: e50092404c ("netfilter: nf_tables: add stateful objects")
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This further refines the changes made to conntrack gc_worker in
commit e0df8cae6c ("netfilter: conntrack: refine gc worker heuristics").
The main idea of that change was to reduce the scan interval when evictions
take place.
However, on the reporters' setup, there are 1-2 million conntrack entries
in total and roughly 8k new (and closing) connections per second.
In this case we'll always evict at least one entry per gc cycle and scan
interval is always at 1 jiffy because of this test:
} else if (expired_count) {
gc_work->next_gc_run /= 2U;
next_run = msecs_to_jiffies(1);
being true almost all the time.
Given we scan ~10k entries per run its clearly wrong to reduce interval
based on nonzero eviction count, it will only waste cpu cycles since a vast
majorities of conntracks are not timed out.
Thus only look at the ratio (scanned entries vs. evicted entries) to make
a decision on whether to reduce or not.
Because evictor is supposed to only kick in when system turns idle after
a busy period, pick a high ratio -- this makes it 50%. We thus keep
the idea of increasing scan rate when its likely that table contains many
expired entries.
In order to not let timed-out entries hang around for too long
(important when using event logging, in which case we want to timely
destroy events), we now scan the full table within at most
GC_MAX_SCAN_JIFFIES (16 seconds) even in worst-case scenario where all
timed-out entries sit in same slot.
I tested this with a vm under synflood (with
sysctl net.netfilter.nf_conntrack_tcp_timeout_syn_recv=3).
While flood is ongoing, interval now stays at its max rate
(GC_MAX_SCAN_JIFFIES / GC_MAX_BUCKETS_DIV -> 125ms).
With feedback from Nicolas Dichtel.
Reported-by: Denys Fedoryshchenko <nuclearcat@nuclearcat.com>
Cc: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Fixes: b87a2f9199 ("netfilter: conntrack: add gc worker to remove timed-out entries")
Signed-off-by: Florian Westphal <fw@strlen.de>
Tested-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Tested-by: Denys Fedoryshchenko <nuclearcat@nuclearcat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Instead of breaking loop and instant resched, don't bother checking
this in first place (the loop calls cond_resched for every bucket anyway).
Suggested-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
The return value of nf_tables_table_lookup() is valid pointer or one
pointer error. There are two cases:
1) IS_ERR(table) is true, it would return the error or reset the
table as NULL, it is unnecessary to perform the latter check
"table != NULL".
2) IS_ERR(obj) is false, the table is one valid pointer. It is also
unnecessary to perform that check.
The nf_tables_newset() and nf_tables_newobj() have same logic codes.
In summary, we could move the block of condition check "table != NULL"
in the else block to eliminate the original condition checks.
Signed-off-by: Gao Feng <fgao@ikuai8.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
After adding the following nft rule, then ping 224.0.0.1:
# nft add rule netdev t c pkttype host counter
The warning complain message will be printed out again and again:
WARNING: CPU: 0 PID: 10182 at net/netfilter/nft_meta.c:163 \
nft_meta_get_eval+0x3fe/0x460 [nft_meta]
[...]
Call Trace:
<IRQ>
dump_stack+0x85/0xc2
__warn+0xcb/0xf0
warn_slowpath_null+0x1d/0x20
nft_meta_get_eval+0x3fe/0x460 [nft_meta]
nft_do_chain+0xff/0x5e0 [nf_tables]
So we should deal with PACKET_LOOPBACK in netdev family too. For ipv4,
convert it to PACKET_BROADCAST/MULTICAST according to the destination
address's type; For ipv6, convert it to PACKET_MULTICAST directly.
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Since there's no broadcast address in IPV6, so in ipv6 family, the
PACKET_LOOPBACK must be multicast packets, there's no need to check
it again.
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
The NF_CONNTRACK Kconfig option description makes an incorrect reference
to the "meta" expression where the "ct" expression would be correct.This
patch fixes the respective typographical error.
Fixes: d497c63527 ("netfilter: add help information to new nf_tables Kconfig options")
Signed-off-by: William Breathitt Gray <vilhelm.gray@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
When dumping nft stateful objects, if NFTA_OBJ_TABLE and NFTA_OBJ_TYPE
attributes are not specified either, filter will become NULL, so oops
will happen(actually nft utility will always set NFTA_OBJ_TABLE attr,
so I write a test program to make this happen):
BUG: unable to handle kernel NULL pointer dereference at (null)
IP: nf_tables_dump_obj+0x17c/0x330 [nf_tables]
[...]
Call Trace:
? nf_tables_dump_obj+0x5/0x330 [nf_tables]
? __kmalloc_reserve.isra.35+0x31/0x90
? __alloc_skb+0x5b/0x1e0
netlink_dump+0x124/0x2a0
__netlink_dump_start+0x161/0x190
nf_tables_getobj+0xe8/0x280 [nf_tables]
Fixes: a9fea2a3c3 ("netfilter: nf_tables: allow to filter stateful object dumps by type")
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
In matches and targets that define a kernel-only tail to their
xt_match and xt_target data structs, add a field .usersize that
specifies up to where data is to be shared with userspace.
Performed a search for comment "Used internally by the kernel" to find
relevant matches and targets. Manually inspected the structs to derive
a valid offsetof.
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Convert compat to copying entries, matches and targets one by one,
using the xt_match_to_user and xt_target_to_user helper functions.
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
xt_entry_target, xt_entry_match and their private data may contain
kernel data.
Introduce helper functions xt_match_to_user, xt_target_to_user and
xt_data_to_user that copy only the expected fields. These replace
existing logic that calls copy_to_user on entire structs, then
overwrites select fields.
Private data is defined in xt_match and xt_target. All matches and
targets that maintain kernel data store this at the tail of their
private structure. Extend xt_match and xt_target with .usersize to
limit how many bytes of data are copied. The remainder is cleared.
If compatsize is specified, usersize can only safely be used if all
fields up to usersize use platform-independent types. Otherwise, the
compat_to_user callback must be defined.
This patch does not yet enable the support logic.
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
The following patchset contains accumulated Netfilter fixes for your
net tree:
1) Ensure quota dump and reset happens iff we can deliver numbers to
userspace.
2) Silence splat on incorrect use of smp_processor_id() from nft_queue.
3) Fix an out-of-bound access reported by KASAN in
nf_tables_rule_destroy(), patch from Florian Westphal.
4) Fix layer 4 checksum mangling in the nf_tables payload expression
with IPv6.
5) Fix a race in the CLUSTERIP target from control plane path when two
threads run to add a new configuration object. Serialize invocations
of clusterip_config_init() using spin_lock. From Xin Long.
6) Call br_nf_pre_routing_finish_bridge_finish() once we are done with
the br_nf_pre_routing_finish() hook. From Artur Molchanov.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
To make the code clearer, use rb_entry() instead of container_of() to
deal with rbtree.
Signed-off-by: Geliang Tang <geliangtang@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
implement sctp_error to let nf_conntrack_in validate crc32c on the packet
transport header. Assign skb->ip_summed to CHECKSUM_UNNECESSARY and return
NF_ACCEPT in case of successful validation; otherwise, return -NF_ACCEPT to
let netfilter skip connection tracking, like other protocols do.
Besides preventing corrupted packets from matching conntrack entries, this
fixes functionality of REJECT target: it was not generating any ICMP upon
reception of SCTP packets, because it was computing RFC 1624 checksum on
the packet and systematically mismatching crc32c in the SCTP header.
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
nf_conntrack needs to compute crc32c when dealing with SCTP packets.
Moreover, NF_NAT_PROTO_SCTP (currently selecting LIBCRC32C) can be enabled
only if conntrack support for SCTP is enabled. Therefore, move enabling of
kernel support for crc32c so that it is selected when NF_CT_PROTO_SCTP=y.
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Reviewed-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Similar to xt_connbytes, user can match how many average bytes per packet
a connection has transferred so far.
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
udplite nat was copied from udp nat, they are virtually 100% identical.
Not really surprising given udplite is just udp with partial csum coverage.
old:
text data bss dec hex filename
11606 1457 210 13273 33d9 nf_nat.ko
330 0 2 332 14c nf_nat_proto_udp.o
276 0 2 278 116 nf_nat_proto_udplite.o
new:
text data bss dec hex filename
11598 1457 210 13265 33d1 nf_nat.ko
640 0 4 644 284 nf_nat_proto_udp.o
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
udplite was copied from udp, they are virtually 100% identical.
This adds udplite tracker to udp instead, removes udplite module,
and then makes the udplite tracker builtin.
udplite will then simply re-use udp timeout settings.
It makes little sense to add separate sysctls, nowadays we have
fine-grained timeout policy support via the CT target.
old:
text data bss dec hex filename
1633 672 0 2305 901 nf_conntrack_proto_udp.o
1756 672 0 2428 97c nf_conntrack_proto_udplite.o
69526 17937 268 87731 156b3 nf_conntrack.ko
new:
text data bss dec hex filename
2442 1184 0 3626 e2a nf_conntrack_proto_udp.o
68565 17721 268 86554 1521a nf_conntrack.ko
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
ktime is a union because the initial implementation stored the time in
scalar nanoseconds on 64 bit machine and in a endianess optimized timespec
variant for 32bit machines. The Y2038 cleanup removed the timespec variant
and switched everything to scalar nanoseconds. The union remained, but
become completely pointless.
Get rid of the union and just keep ktime_t as simple typedef of type s64.
The conversion was done with coccinelle and some manual mopping up.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
This was entirely automated, using the script by Al:
PATT='^[[:blank:]]*#[[:blank:]]*include[[:blank:]]*<asm/uaccess.h>'
sed -i -e "s!$PATT!#include <linux/uaccess.h>!" \
$(git grep -l "$PATT"|grep -v ^include/linux/uaccess.h)
to do the replacement at the end of the merge window.
Requested-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
If the NFT_PAYLOAD_L4CSUM_PSEUDOHDR flag is set, then mangle layer 4
checksum. This should not depend on csum_type NFT_PAYLOAD_CSUM_INET
since IPv6 header has no checksum field, but still an update of any of
the pseudoheader fields may trigger a layer 4 checksum update.
Fixes: 1814096980 ("netfilter: nft_payload: layer 4 checksum adjustment for pseudoheader fields")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
BUG: KASAN: slab-out-of-bounds in nf_tables_rule_destroy+0xf1/0x130 at addr ffff88006a4c35c8
Read of size 8 by task nft/1607
When we've destroyed last valid expr, nft_expr_next() returns an invalid expr.
We must not dereference it unless it passes != nft_expr_last() check.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Using smp_processor_id() causes splats with PREEMPT_RCU:
[19379.552780] BUG: using smp_processor_id() in preemptible [00000000] code: ping/32389
[19379.552793] caller is debug_smp_processor_id+0x17/0x19
[...]
[19379.552823] Call Trace:
[19379.552832] [<ffffffff81274e9e>] dump_stack+0x67/0x90
[19379.552837] [<ffffffff8129a4d4>] check_preemption_disabled+0xe5/0xf5
[19379.552842] [<ffffffff8129a4fb>] debug_smp_processor_id+0x17/0x19
[19379.552849] [<ffffffffa07c42dd>] nft_queue_eval+0x35/0x20c [nft_queue]
No need to disable preemption since we only fetch the numeric value, so
let's use raw_smp_processor_id() instead.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Dumping of netlink attributes may fail due to insufficient room in the
skbuff, so let's reset consumed quota if we succeed to put netlink
attributes into the skbuff.
Fixes: 43da04a593 ("netfilter: nf_tables: atomic dump and reset for stateful objects")
Reported-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Dump and reset doesn't work unless cmpxchg64() is used both from packet
and control plane paths. This approach is going to be slow though.
Instead, use a percpu seqcount to fetch counters consistently, then
subtract bytes and packets in case a reset was requested.
The cpu that running over the reset code is guaranteed to own this stats
exclusively, we have to turn counters into signed 64bit though so stats
update on reset don't get wrong on underflow.
This patch is based on original sketch from Eric Dumazet.
Fixes: 43da04a593 ("netfilter: nf_tables: atomic dump and reset for stateful objects")
Suggested-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pablo Neira Ayuso says:
====================
Netfilter/IPVS updates for net-next
The following patchset contains a large Netfilter update for net-next,
to summarise:
1) Add support for stateful objects. This series provides a nf_tables
native alternative to the extended accounting infrastructure for
nf_tables. Two initial stateful objects are supported: counters and
quotas. Objects are identified by a user-defined name, you can fetch
and reset them anytime. You can also use a maps to allow fast lookups
using any arbitrary key combination. More info at:
http://marc.info/?l=netfilter-devel&m=148029128323837&w=2
2) On-demand registration of nf_conntrack and defrag hooks per netns.
Register nf_conntrack hooks if we have a stateful ruleset, ie.
state-based filtering or NAT. The new nf_conntrack_default_on sysctl
enables this from newly created netnamespaces. Default behaviour is not
modified. Patches from Florian Westphal.
3) Allocate 4k chunks and then use these for x_tables counter allocation
requests, this improves ruleset load time and also datapath ruleset
evaluation, patches from Florian Westphal.
4) Add support for ebpf to the existing x_tables bpf extension.
From Willem de Bruijn.
5) Update layer 4 checksum if any of the pseudoheader fields is updated.
This provides a limited form of 1:1 stateless NAT that make sense in
specific scenario, eg. load balancing.
6) Add support to flush sets in nf_tables. This series comes with a new
set->ops->deactivate_one() indirection given that we have to walk
over the list of set elements, then deactivate them one by one.
The existing set->ops->deactivate() performs an element lookup that
we don't need.
7) Two patches to avoid cloning packets, thus speed up packet forwarding
via nft_fwd from ingress. From Florian Westphal.
8) Two IPVS patches via Simon Horman: Decrement ttl in all modes to
prevent infinite loops, patch from Dwip Banerjee. And one minor
refactoring from Gao feng.
9) Revisit recent log support for nf_tables netdev families: One patch
to ensure that we correctly handle non-ethernet packets. Another
patch to add missing logger definition for netdev. Patches from
Liping Zhang.
10) Three patches for nft_fib, one to address insufficient register
initialization and another to solve incorrect (although harmless)
byteswap operation. Moreover update xt_rpfilter and nft_fib to match
lbcast packets with zeronet as source, eg. DHCP Discover packets
(0.0.0.0 -> 255.255.255.255). Also from Liping Zhang.
11) Built-in DCCP, SCTP and UDPlite conntrack and NAT support, from
Davide Caratti. While DCCP is rather hopeless lately, and UDPlite has
been broken in many-cast mode for some little time, let's give them a
chance by placing them at the same level as other existing protocols.
Thus, users don't explicitly have to modprobe support for this and
NAT rules work for them. Some people point to the lack of support in
SOHO Linux-based routers that make deployment of new protocols harder.
I guess other middleboxes outthere on the Internet are also to blame.
Anyway, let's see if this has any impact in the midrun.
12) Skip software SCTP software checksum calculation if the NIC comes
with SCTP checksum offload support. From Davide Caratti.
13) Initial core factoring to prepare conversion to hook array. Three
patches from Aaron Conole.
14) Gao Feng made a wrong conversion to switch in the xt_multiport
extension in a patch coming in the previous batch. Fix it in this
batch.
15) Get vmalloc call in sync with kmalloc flags to avoid a warning
and likely OOM killer intervention from x_tables. From Marcelo
Ricardo Leitner.
16) Update Arturo Borrero's email address in all source code headers.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for attaching an eBPF object by file descriptor.
The iptables binary can be called with a path to an elf object or a
pinned bpf object. Also pass the mode and path to the kernel to be
able to return it later for iptables dump and save.
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Andrey Konovalov reported that this vmalloc call is based on an
userspace request and that it's spewing traces, which may flood the logs
and cause DoS if abused.
Florian Westphal also mentioned that this call should not trigger OOM
killer.
This patch brings the vmalloc call in sync to kmalloc and disables the
warn trace on allocation failure and also disable OOM killer invocation.
Note, however, that under such stress situation, other places may
trigger OOM killer invocation.
Reported-by: Andrey Konovalov <andreyknvl@google.com>
Cc: Florian Westphal <fw@strlen.de>
Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch adds support for set flushing, that consists of walking over
the set elements if the NFTA_SET_ELEM_LIST_ELEMENTS attribute is set.
This patch requires the following changes:
1) Add set->ops->deactivate_one() operation: This allows us to
deactivate an element from the set element walk path, given we can
skip the lookup that happens in ->deactivate().
2) Add a new nft_trans_alloc_gfp() function since we need to allocate
transactions using GFP_ATOMIC given the set walk path happens with
held rcu_read_lock.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This new function allows us to deactivate one single element, this is
required by the set flush command that comes in a follow up patch.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
SCTP GSO and hardware can do CRC32c computation after netfilter processing,
so we can avoid calling sctp_compute_checksum() on skb if skb->ip_summed
is equal to CHECKSUM_PARTIAL. Moreover, set skb->ip_summed to CHECKSUM_NONE
when the NAT code computes the CRC, to prevent offloaders from computing
it again (on ixgbe this resulted in a transmission with wrong L4 checksum).
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch adds the netlink code to filter out dump of stateful objects,
through the NFTA_OBJ_TYPE netlink attribute.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch allows us to refer to stateful object dictionaries, the
source register indicates the key data to be used to look up for the
corresponding state object. We can refer to these maps through names or,
alternatively, the map transaction id. This allows us to refer to both
anonymous and named maps.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch allows you to refer to stateful objects from set elements.
This provides the infrastructure to create maps where the right hand
side of the mapping is a stateful object.
This allows us to build dictionaries of stateful objects, that you can
use to perform fast lookups using any arbitrary key combination.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Notify on depleted quota objects. The NFT_QUOTA_F_DEPLETED flag
indicates we have reached overquota.
Add pointer to table from nft_object, so we can use it when sending the
depletion notification to userspace.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Introduce nf_tables_obj_notify() to notify internal state changes in
stateful objects. This is used by the quota object to report depletion
in a follow up patch.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch adds a new NFT_MSG_GETOBJ_RESET command perform an atomic
dump-and-reset of the stateful object. This also comes with add support
for atomic dump and reset for counter and quota objects.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Add a new attribute NFTA_QUOTA_CONSUMED that displays the amount of
quota that has been already consumed. This allows us to restore the
internal state of the quota object between reboots as well as to monitor
how wasted it is.
This patch changes the logic to account for the consumed bytes, instead
of the bytes that remain to be consumed.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch augments nf_tables to support stateful objects. This new
infrastructure allows you to create, dump and delete stateful objects,
that are identified by a user-defined name.
This patch adds the generic infrastructure, follow up patches add
support for two stateful objects: counters and quotas.
This patch provides a native infrastructure for nf_tables to replace
nfacct, the extended accounting infrastructure for iptables.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
... so we can use current skb instead of working with a clone.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
I lost one test case in the last commit for xt_multiport.
For example, the rule is "-m multiport --dports 22,80,443".
When first port is unmatched and the second is matched, the curent codes
could not return the right result.
It would return false directly when the first port is unmatched.
Fixes: dd2602d00f ("netfilter: xt_multiport: Use switch case instead
of multiple condition checks")
Signed-off-by: Gao Feng <fgao@ikuai8.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch adds a new flag that signals the kernel to update layer 4
checksum if the packet field belongs to the layer 4 pseudoheader. This
implicitly provides stateless NAT 1:1 that is useful under very specific
usecases.
Since rules mangling layer 3 fields that are part of the pseudoheader
may potentially convey any layer 4 packet, we have to deal with the
layer 4 checksum adjustment using protocol specific code.
This patch adds support for TCP, UDP and ICMPv6, since they include the
pseudoheader in the layer 4 checksum calculation. ICMP doesn't, so we
can skip it.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Acctually ntohl and htonl are identical, so this doesn't affect
anything, but it is conceptually wrong.
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
instead of allocating each xt_counter individually, allocate 4k chunks
and then use these for counter allocation requests.
This should speed up rule evaluation by increasing data locality,
also speeds up ruleset loading because we reduce calls to the percpu
allocator.
As Eric points out we can't use PAGE_SIZE, page_allocator would fail on
arches with 64k page size.
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Keeps some noise away from a followup patch.
Signed-off-by: Florian Westphal <fw@strlen.de>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
On SMP we overload the packet counter (unsigned long) to contain
percpu offset. Hide this from callers and pass xt_counters address
instead.
Preparation patch to allocate the percpu counters in page-sized batch
chunks.
Signed-off-by: Florian Westphal <fw@strlen.de>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This is to facilitate converting from a singly-linked list to an array
of elements.
Signed-off-by: Aaron Conole <aconole@bytheb.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
nf_defrag modules for ipv4 and ipv6 export an empty stub function.
Any module that needs the defragmentation hooks registered simply 'calls'
this empty function to create a phony module dependency -- modprobe will
then load the defrag module too.
This extends netfilter ipv4/ipv6 defragmentation modules to delay the hook
registration until the functionality is requested within a network namespace
instead of module load time for all namespaces.
Hooks are only un-registered on module unload or when a namespace that used
such defrag functionality exits.
We have to use struct net for this as the register hooks can be called
before netns initialization here from the ipv4/ipv6 conntrack module
init path.
There is no unregister functionality support, defrag will always be
active once it was requested inside a net namespace.
The reason is that defrag has impact on nft and iptables rulesets
(without defrag we might see framents).
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
1) Old code was hard to maintain, due to complex lock chains.
(We probably will be able to remove some kfree_rcu() in callers)
2) Using a single timer to update all estimators does not scale.
3) Code was buggy on 32bit kernel (WRITE_ONCE() on 64bit quantity
is not supposed to work well)
In this rewrite :
- I removed the RB tree that had to be scanned in
gen_estimator_active(). qdisc dumps should be much faster.
- Each estimator has its own timer.
- Estimations are maintained in net_rate_estimator structure,
instead of dirtying the qdisc. Minor, but part of the simplification.
- Reading the estimator uses RCU and a seqcount to provide proper
support for 32bit kernels.
- We reduce memory need when estimators are not used, since
we store a pointer, instead of the bytes/packets counters.
- xt_rateest_mt() no longer has to grab a spinlock.
(In the future, xt_rateest_tg() could be switched to per cpu counters)
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This switch (default on) can be used to disable automatic registration
of connection tracking functionality in newly created network
namespaces.
This means that when net namespace goes down (or the tracker protocol
module is unloaded) we *might* have to unregister the hooks.
We can either add another per-netns variable that tells if
the hooks got registered by default, or, alternatively, just call
the protocol _put() function and have the callee deal with a possible
'extra' put() operation that doesn't pair with a get() one.
This uses the latter approach, i.e. a put() without a get has no effect.
Conntrack is still enabled automatically regardless of the new sysctl
setting if the new net namespace requires connection tracking, e.g. when
NAT rules are created.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This makes use of nf_ct_netns_get/put added in previous patch.
We add get/put functions to nf_conntrack_l3proto structure, ipv4 and ipv6
then implement use-count to track how many users (nft or xtables modules)
have a dependency on ipv4 and/or ipv6 connection tracking functionality.
When count reaches zero, the hooks are unregistered.
This delays activation of connection tracking inside a namespace until
stateful firewall rule or nat rule gets added.
This patch breaks backwards compatibility in the sense that connection
tracking won't be active anymore when the protocol tracker module is
loaded. This breaks e.g. setups that ctnetlink for flow accounting and
the like, without any '-m conntrack' packet filter rules.
Followup patch restores old behavour and makes new delayed scheme
optional via sysctl.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
so that conntrack core will add the needed hooks in this namespace.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
MASQUERADE, S/DNAT and REDIRECT already call functions that depend on the
conntrack module.
However, since the conntrack hooks are now registered in a lazy fashion
(i.e., only when needed) a symbol reference is not enough.
Thus, when something is added to a nat table, make sure that it will see
packets by calling nf_ct_netns_get() which will register the conntrack
hooks in the current netns.
An alternative would be to add these dependencies to the NAT table.
However, that has problems when using non-modular builds -- we might
register e.g. ipv6 conntrack before its initcall has run, leading to NULL
deref crashes since its per-netns storage has not yet been allocated.
Adding the dependency in the modules instead has the advantage that nat
table also does not register its hooks until rules are added.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
currently aliased to try_module_get/_put.
Will be changed in next patch when we add functions to make use of ->net
argument to store usercount per l3proto tracker.
This is needed to avoid registering the conntrack hooks in all netns and
later only enable connection tracking in those that need conntrack.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
since adf0516845 ("netfilter: remove ip_conntrack* sysctl compat code")
the only user (ipv4 tracker) sets this to an empty stub function.
After this change nf_ct_l3proto_pernet_register() is also empty,
but this will change in a followup patch to add conditional register
of the hooks.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
CONFIG_NF_CT_PROTO_UDPLITE is no more a tristate. When set to y,
connection tracking support for UDPlite protocol is built-in into
nf_conntrack.ko.
footprint test:
$ ls -l net/netfilter/nf_conntrack{_proto_udplite,}.ko \
net/ipv4/netfilter/nf_conntrack_ipv4.ko \
net/ipv6/netfilter/nf_conntrack_ipv6.ko
(builtin)|| udplite| ipv4 | ipv6 |nf_conntrack
---------++--------+--------+--------+--------------
none || 432538 | 828755 | 828676 | 6141434
UDPlite || - | 829649 | 829362 | 6498204
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
CONFIG_NF_CT_PROTO_SCTP is no more a tristate. When set to y, connection
tracking support for SCTP protocol is built-in into nf_conntrack.ko.
footprint test:
$ ls -l net/netfilter/nf_conntrack{_proto_sctp,}.ko \
net/ipv4/netfilter/nf_conntrack_ipv4.ko \
net/ipv6/netfilter/nf_conntrack_ipv6.ko
(builtin)|| sctp | ipv4 | ipv6 | nf_conntrack
---------++--------+--------+--------+--------------
none || 498243 | 828755 | 828676 | 6141434
SCTP || - | 829254 | 829175 | 6547872
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
CONFIG_NF_CT_PROTO_DCCP is no more a tristate. When set to y, connection
tracking support for DCCP protocol is built-in into nf_conntrack.ko.
footprint test:
$ ls -l net/netfilter/nf_conntrack{_proto_dccp,}.ko \
net/ipv4/netfilter/nf_conntrack_ipv4.ko \
net/ipv6/netfilter/nf_conntrack_ipv6.ko
(builtin)|| dccp | ipv4 | ipv6 | nf_conntrack
---------++--------+--------+--------+--------------
none || 469140 | 828755 | 828676 | 6141434
DCCP || - | 830566 | 829935 | 6533526
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Simon Horman says:
====================
IPVS Updates for v4.10
please consider these enhancements to the IPVS for v4.10.
* Decrement the IP ttl in all the modes in order to prevent infinite
route loops. Thanks to Dwip Banerjee.
* Use IS_ERR_OR_NULL macro. Clean-up from Gao Feng.
====================
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
So we can autoload nfnetlink_log.ko when the user adding nft log
group X rule in netdev family.
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
In netdev family, we will handle non ethernet packets, so using
eth_hdr(skb)->h_proto is incorrect.
Meanwhile, we can use socket(AF_PACKET...) to sending packets, so
skb->protocol is not always set in bridge family.
Add an extra parameter into nf_log_l2packet to solve this issue.
Fixes: 1fddf4bad0 ("netfilter: nf_log: add packet logging for netdev family")
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
CONFIG_NF_NAT_PROTO_UDPLITE is no more a tristate. When set to y, NAT
support for UDPlite protocol is built-in into nf_nat.ko.
footprint test:
(nf_nat_proto_) |udplite || nf_nat
--------------------------+--------++--------
no builtin | 408048 || 2241312
UDPLITE builtin | - || 2577256
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
CONFIG_NF_NAT_PROTO_SCTP is no more a tristate. When set to y, NAT
support for SCTP protocol is built-in into nf_nat.ko.
footprint test:
(nf_nat_proto_) | sctp || nf_nat
--------------------------+--------++--------
no builtin | 428344 || 2241312
SCTP builtin | - || 2597032
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
CONFIG_NF_NAT_PROTO_DCCP is no more a tristate. When set to y, NAT
support for DCCP protocol is built-in into nf_nat.ko.
footprint test:
(nf_nat_proto_) | dccp || nf_nat
--------------------------+--------++--------
no builtin | 409800 || 2241312
DCCP builtin | - || 2578968
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
The email address has changed, let's update the copyright statements.
Signed-off-by: Arturo Borrero Gonzalez <arturo@debian.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Couple conflicts resolved here:
1) In the MACB driver, a bug fix to properly initialize the
RX tail pointer properly overlapped with some changes
to support variable sized rings.
2) In XGBE we had a "CONFIG_PM" --> "CONFIG_PM_SLEEP" fix
overlapping with a reorganization of the driver to support
ACPI, OF, as well as PCI variants of the chip.
3) In 'net' we had several probe error path bug fixes to the
stmmac driver, meanwhile a lot of this code was cleaned up
and reorganized in 'net-next'.
4) The cls_flower classifier obtained a helper function in
'net-next' called __fl_delete() and this overlapped with
Daniel Borkamann's bug fix to use RCU for object destruction
in 'net'. It also overlapped with Jiri's change to guard
the rhashtable_remove_fast() call with a check against
tc_skip_sw().
5) In mlx4, a revert bug fix in 'net' overlapped with some
unrelated changes in 'net-next'.
6) In geneve, a stale header pointer after pskb_expand_head()
bug fix in 'net' overlapped with a large reorganization of
the same code in 'net-next'. Since the 'net-next' code no
longer had the bug in question, there was nothing to do
other than to simply take the 'net-next' hunks.
Signed-off-by: David S. Miller <davem@davemloft.net>
Otherwise, kernel panic will happen if the user does not specify
the related attributes.
Fixes: 0f3cd9b369 ("netfilter: nf_tables: add range expression")
Signed-off-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
As Liping Zhang reports, after commit a8b1e36d0d ("netfilter: nft_dynset:
fix element timeout for HZ != 1000"), priv->timeout was stored in jiffies,
while set->timeout was stored in milliseconds. This is inconsistent and
incorrect.
Firstly, we already call msecs_to_jiffies in nft_set_elem_init, so
priv->timeout will be converted to jiffies twice.
Secondly, if the user did not specify the NFTA_DYNSET_TIMEOUT attr,
set->timeout will be used, but we forget to call msecs_to_jiffies
when do update elements.
Fix this by using jiffies internally for traditional sets and doing the
conversions to/from msec when interacting with userspace - as dynset
already does.
This is preferable to doing the conversions, when elements are inserted or
updated, because this can happen very frequently on busy dynsets.
Fixes: a8b1e36d0d ("netfilter: nft_dynset: fix element timeout for HZ != 1000")
Reported-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Anders K. Pedersen <akp@cohaesio.com>
Acked-by: Liping Zhang <zlpnobody@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
I got offlist bug report about failing connections and high cpu usage.
This happens because we hit 'elasticity' checks in rhashtable that
refuses bucket list exceeding 16 entries.
The nat bysrc hash unfortunately needs to insert distinct objects that
share same key and are identical (have same source tuple), this cannot
be avoided.
Switch to the rhlist interface which is designed for this.
The nulls_base is removed here, I don't think its needed:
A (unlikely) false positive results in unneeded port clash resolution,
a false negative results in packet drop during conntrack confirmation,
when we try to insert the duplicate into main conntrack hash table.
Tested by adding multiple ip addresses to host, then adding
iptables -t nat -A POSTROUTING -o eth0 -j MASQUERADE
... and then creating multiple connections, from same source port but
different addresses:
for i in $(seq 2000 2032);do nc -p 1234 192.168.7.1 $i > /dev/null & done
(all of these then get hashed to same bysource slot)
Then, to test that nat conflict resultion is working:
nc -s 10.0.0.1 -p 1234 192.168.7.1 2000
nc -s 10.0.0.2 -p 1234 192.168.7.1 2000
tcp .. src=10.0.0.1 dst=192.168.7.1 sport=1234 dport=2000 src=192.168.7.1 dst=192.168.7.10 sport=2000 dport=1024 [ASSURED]
tcp .. src=10.0.0.2 dst=192.168.7.1 sport=1234 dport=2000 src=192.168.7.1 dst=192.168.7.10 sport=2000 dport=1025 [ASSURED]
tcp .. src=192.168.7.10 dst=192.168.7.1 sport=1234 dport=2000 src=192.168.7.1 dst=192.168.7.10 sport=2000 dport=1234 [ASSURED]
tcp .. src=192.168.7.10 dst=192.168.7.1 sport=1234 dport=2001 src=192.168.7.1 dst=192.168.7.10 sport=2001 dport=1234 [ASSURED]
[..]
-> nat altered source ports to 1024 and 1025, respectively.
This can also be confirmed on destination host which shows
ESTAB 0 0 192.168.7.1:2000 192.168.7.10:1024
ESTAB 0 0 192.168.7.1:2000 192.168.7.10:1025
ESTAB 0 0 192.168.7.1:2000 192.168.7.10:1234
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Fixes: 870190a9ec ("netfilter: nat: convert nat bysrc hash to rhashtable")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
The comparator works like memcmp, i.e. 0 means objects are equal.
In other words, when objects are distinct they are treated as identical,
when they are distinct they are allegedly the same.
The first case is rare (distinct objects are unlikely to get hashed to
same bucket).
The second case results in unneeded port conflict resolutions attempts.
Fixes: 870190a9ec ("netfilter: nat: convert nat bysrc hash to rhashtable")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Use the function nft_parse_u32_check() to fetch the value and validate
the u32 attribute into the hash len u8 field.
This patch revisits 4da449ae1d ("netfilter: nft_exthdr: Add size check
on u8 nft_exthdr attributes").
Fixes: cb1b69b0b1 ("netfilter: nf_tables: add hash expression")
Signed-off-by: Laura Garcia Liebana <nevola@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Make struct pernet_operations::id unsigned.
There are 2 reasons to do so:
1)
This field is really an index into an zero based array and
thus is unsigned entity. Using negative value is out-of-bound
access by definition.
2)
On x86_64 unsigned 32-bit data which are mixed with pointers
via array indexing or offsets added or subtracted to pointers
are preffered to signed 32-bit data.
"int" being used as an array index needs to be sign-extended
to 64-bit before being used.
void f(long *p, int i)
{
g(p[i]);
}
roughly translates to
movsx rsi, esi
mov rdi, [rsi+...]
call g
MOVSX is 3 byte instruction which isn't necessary if the variable is
unsigned because x86_64 is zero extending by default.
Now, there is net_generic() function which, you guessed it right, uses
"int" as an array index:
static inline void *net_generic(const struct net *net, int id)
{
...
ptr = ng->ptr[id - 1];
...
}
And this function is used a lot, so those sign extensions add up.
Patch snipes ~1730 bytes on allyesconfig kernel (without all junk
messing with code generation):
add/remove: 0/0 grow/shrink: 70/598 up/down: 396/-2126 (-1730)
Unfortunately some functions actually grow bigger.
This is a semmingly random artefact of code generation with register
allocator being used differently. gcc decides that some variable
needs to live in new r8+ registers and every access now requires REX
prefix. Or it is shifted into r12, so [r12+0] addressing mode has to be
used which is longer than [r8]
However, overall balance is in negative direction:
add/remove: 0/0 grow/shrink: 70/598 up/down: 396/-2126 (-1730)
function old new delta
nfsd4_lock 3886 3959 +73
tipc_link_build_proto_msg 1096 1140 +44
mac80211_hwsim_new_radio 2776 2808 +32
tipc_mon_rcv 1032 1058 +26
svcauth_gss_legacy_init 1413 1429 +16
tipc_bcbase_select_primary 379 392 +13
nfsd4_exchange_id 1247 1260 +13
nfsd4_setclientid_confirm 782 793 +11
...
put_client_renew_locked 494 480 -14
ip_set_sockfn_get 730 716 -14
geneve_sock_add 829 813 -16
nfsd4_sequence_done 721 703 -18
nlmclnt_lookup_host 708 686 -22
nfsd4_lockt 1085 1063 -22
nfs_get_client 1077 1050 -27
tcf_bpf_init 1106 1076 -30
nfsd4_encode_fattr 5997 5930 -67
Total: Before=154856051, After=154854321, chg -0.00%
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>