When an SKB cannot be chained to a session, the current code attempts
to "restore" its ip_summed field from lro_mgr->ip_summed. However,
lro_mgr->ip_summed does not hold the original value; in fact, we'd
better not touch skb->ip_summed since it is not modified by the code
in the path leading to a failure to chain it. Also use a cleaer
comment to the describe the ip_summed field of struct net_lro_mgr.
Issue raised by Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>
The problem is that while we work w/o the inet_frags.lock even
read-locked the secret rebuild timer may occur (on another CPU, since
BHs are still disabled in the inet_frag_find) and change the rnd seed
for ipv4/6 fragments.
It was caused by my patch fd9e63544c
([INET]: Omit double hash calculations in xxx_frag_intern) late
in the 2.6.24 kernel, so this should probably be queued to -stable.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
I found another case where we are sending information to userspace
in the wrong HZ scale. This should have been fixed back in 2.5 :-(
This means an ABI change but as it stands there is no way for an application
like ss to get the right value.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The tcp_mem array which contains limits on the total amount of memory
used by TCP sockets is calculated based on nr_all_pages. On a 32 bits
x86 system, we should base this on the number of lowmem pages.
Signed-off-by: Miquel van Smoorenburg <miquels@cistron.nl>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add skb_warn_if_lro() to test whether an skb was received with LRO and
warn if so.
Change br_forward(), ip_forward() and ip6_forward() to call it) and
discard the skb if it returns true.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Large Receive Offload (LRO) is only appropriate for packets that are
destined for the host, and should be disabled if received packets may be
forwarded. It can also confuse the GSO on output.
Add dev_disable_lro() function which uses the appropriate ethtool ops to
disable LRO if enabled.
Add calls to dev_disable_lro() in br_add_if() and functions that enable
IPv4 and IPv6 forwarding.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In commits 33c732c361 ([IPV4]: Add raw
drops counter) and a92aa318b4 ([IPV6]:
Add raw drops counter), Wang Chen added raw drops counter for
/proc/net/raw & /proc/net/raw6
This patch adds this capability to UDP sockets too (/proc/net/udp &
/proc/net/udp6).
This means that 'RcvbufErrors' errors found in /proc/net/snmp can be also
be examined for each udp socket.
# grep Udp: /proc/net/snmp
Udp: InDatagrams NoPorts InErrors OutDatagrams RcvbufErrors SndbufErrors
Udp: 23971006 75 899420 16390693 146348 0
# cat /proc/net/udp
sl local_address rem_address st tx_queue rx_queue tr tm->when retrnsmt ---
uid timeout inode ref pointer drops
75: 00000000:02CB 00000000:0000 07 00000000:00000000 00:00000000 00000000 ---
0 0 2358 2 ffff81082a538c80 0
111: 00000000:006F 00000000:0000 07 00000000:00000000 00:00000000 00000000 ---
0 0 2286 2 ffff81042dd35c80 146348
In this example, only port 111 (0x006F) was flooded by messages that
user program could not read fast enough. 146348 messages were lost.
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When generating the ip header for the transformed packet we just copy
the frag_off field of the ip header from the original packet to the ip
header of the new generated packet. If we receive a packet as a chain
of fragments, all but the last of the new generated packets have the
IP_MF flag set. We have to mask the frag_off field to only keep the
IP_DF flag from the original packet. This got lost with git commit
36cf9acf93 ("[IPSEC]: Separate
inner/outer mode processing on output")
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix three ct_extend/NAT extension related races:
- When cleaning up the extension area and removing it from the bysource hash,
the nat->ct pointer must not be set to NULL since it may still be used in
a RCU read side
- When replacing a NAT extension area in the bysource hash, the nat->ct
pointer must be assigned before performing the replacement
- When reallocating extension storage in ct_extend, the old memory must
not be freed immediately since it may still be used by a RCU read side
Possibly fixes https://bugzilla.redhat.com/show_bug.cgi?id=449315
and/or http://bugzilla.kernel.org/show_bug.cgi?id=10875
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Although this hash takes addresses into account, the ehash chains
can also be too long when, for instance, communications via lo occur.
So, prepare the inet_hashfn to take struct net into account.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Listening-on-one-port sockets in many namespaces produce long
chains in the listening_hash-es, so prepare the inet_lhashfn to
take struct net into account.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Binding to some port in many namespaces may create too long
chains in bhash-es, so prepare the hashfn to take struct net
into account.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Every caller already has this one. The new argument is currently
unused, but this will be fixed shortly.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
They both calculate the hash chain, but currently do not have
a struct net pointer, so pass one there via additional argument,
all the more so their callers already have such.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently the chain to store a UDP socket is calculated with
simple (x & (UDP_HTABLE_SIZE - 1)). But taking net into account
would make this calculation a bit more complex, so moving it into
a function would help.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
1) Remove ICMP_MIN_LENGTH, as it is unused.
2) Remove unneeded tcp_v4_send_check() declaration.
Signed-off-by: Rami Rosen <ramirose@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
I just noticed "cat /proc/net/raw" was buggy, missing '\n' separators.
I believe this was introduced by commit 8cd850efa4
([RAW]: Cleanup IPv4 raw_seq_show.)
This trivial patch restores correct behavior, and applies to current
Linus tree (should also be applied to stable tree as well.)
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Ingo's system is still seeing strange behavior, and he
reports that is goes away if the rest of the deferred
accept changes are reverted too.
Therefore this reverts e4c7884028
("[TCP]: TCP_DEFER_ACCEPT updates - dont retxmt synack") and
539fae89be ("[TCP]: TCP_DEFER_ACCEPT
updates - defer timeout conflicts with max_thresh").
Just like the other revert, these ideas can be revisited for
2.6.27
Signed-off-by: David S. Miller <davem@davemloft.net>
Change struct proto destroy function pointer to return void. Noticed
by Al Viro.
Signed-off-by: Brian Haley <brian.haley@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6:
tcp: Revert 'process defer accept as established' changes.
ipv6: Fix duplicate initialization of rawv6_prot.destroy
bnx2x: Updating the Maintainer
net: Eliminate flush_scheduled_work() calls while RTNL is held.
drivers/net/r6040.c: correct bad use of round_jiffies()
fec_mpc52xx: MPC52xx_MESSAGES_DEFAULT: 2nd NETIF_MSG_IFDOWN => IFUP
ipg: fix receivemode IPG_RM_RECEIVEMULTICAST{,HASH} in ipg_nic_set_multicast_list()
netfilter: nf_conntrack: fix ctnetlink related crash in nf_nat_setup_info()
netfilter: Make nflog quiet when no one listen in userspace.
ipv6: Fail with appropriate error code when setting not-applicable sockopt.
ipv6: Check IPV6_MULTICAST_LOOP option value.
ipv6: Check the hop limit setting in ancillary data.
ipv6 route: Fix route lifetime in netlink message.
ipv6 mcast: Check address family of gf_group in getsockopt(MS_FILTER).
dccp: Bug in initial acknowledgment number assignment
dccp ccid-3: X truncated due to type conversion
dccp ccid-3: TFRC reverse-lookup Bug-Fix
dccp ccid-2: Bug-Fix - Ack Vectors need to be ignored on request sockets
dccp: Fix sparse warnings
dccp ccid-3: Bug-Fix - Zero RTT is possible
This reverts two changesets, ec3c0982a2
("[TCP]: TCP_DEFER_ACCEPT updates - process as established") and
the follow-on bug fix 9ae27e0adb
("tcp: Fix slab corruption with ipv6 and tcp6fuzz").
This change causes several problems, first reported by Ingo Molnar
as a distcc-over-loopback regression where connections were getting
stuck.
Ilpo Järvinen first spotted the locking problems. The new function
added by this code, tcp_defer_accept_check(), only has the
child socket locked, yet it is modifying state of the parent
listening socket.
Fixing that is non-trivial at best, because we can't simply just grab
the parent listening socket lock at this point, because it would
create an ABBA deadlock. The normal ordering is parent listening
socket --> child socket, but this code path would require the
reverse lock ordering.
Next is a problem noticed by Vitaliy Gusev, he noted:
----------------------------------------
>--- a/net/ipv4/tcp_timer.c
>+++ b/net/ipv4/tcp_timer.c
>@@ -481,6 +481,11 @@ static void tcp_keepalive_timer (unsigned long data)
> goto death;
> }
>
>+ if (tp->defer_tcp_accept.request && sk->sk_state == TCP_ESTABLISHED) {
>+ tcp_send_active_reset(sk, GFP_ATOMIC);
>+ goto death;
Here socket sk is not attached to listening socket's request queue. tcp_done()
will not call inet_csk_destroy_sock() (and tcp_v4_destroy_sock() which should
release this sk) as socket is not DEAD. Therefore socket sk will be lost for
freeing.
----------------------------------------
Finally, Alexey Kuznetsov argues that there might not even be any
real value or advantage to these new semantics even if we fix all
of the bugs:
----------------------------------------
Hiding from accept() sockets with only out-of-order data only
is the only thing which is impossible with old approach. Is this really
so valuable? My opinion: no, this is nothing but a new loophole
to consume memory without control.
----------------------------------------
So revert this thing for now.
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch removes CVS keywords that weren't updated for a long time
from comments.
Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
As we do for other socket/timewait-socket specific parameters,
let the callers pass appropriate arguments to
tcp_v{4,6}_do_calc_md5_hash().
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
We can share most part of the hash calculation code because
the only difference between IPv4 and IPv6 is their pseudo headers.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (42 commits)
net: Fix routing tables with id > 255 for legacy software
sky2: Hold RTNL while calling dev_close()
s2io iomem annotations
atl1: fix suspend regression
qeth: start dev queue after tx drop error
qeth: Prepare-function to call s390dbf was wrong
qeth: reduce number of kernel messages
qeth: Use ccw_device_get_id().
qeth: layer 3 Oops in ip event handler
virtio: use callback on empty in virtio_net
virtio: virtio_net free transmit skbs in a timer
virtio: Fix typo in virtio_net_hdr comments
virtio_net: Fix skb->csum_start computation
ehea: set mac address fix
sfc: Recover from RX queue flush failure
add missing lance_* exports
ixgbe: fix typo
forcedeth: msi interrupts
ipsec: pfkey should ignore events when no listeners
pppoe: Unshare skb before anything else
...
Most legacy software do not like tables > 255 as rtm_table is u8
so tb_id is sent &0xff and it is possible to mismatch for example
table 510 with table 254 (main).
This patch introduces RT_TABLE_COMPAT=252 so the code uses it if
tb_id > 255. It makes such old applications happy, new
ones are still able to use RTA_TABLE to get a proper table id.
Signed-off-by: Krzysztof Piotr Oledzki <ole@ans.pl>
Acked-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Makes people happy who try to keep a list of addresses up to date by
listening to notifications.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Wei Yongjun noticed that we may call reqsk_free on request sock objects where
the opt fields may not be initialized, fix it by introducing inet_reqsk_alloc
where we initialize ->opt to NULL and set ->pktopts to NULL in
inet6_reqsk_alloc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
- No need to perform data_len = 0 in the switch command, since data_len
is initialized to 0 in the beginning of the ipq_build_packet_message()
method.
- {ip,ip6}_queue: We can reach nlmsg_failure only from one place; skb is
sure to be NULL when getting there; since skb is NULL, there is no need
to check this fact and call kfree_skb().
Signed-off-by: Rami Rosen <ramirose@gmail.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently the last packet of a connection isn't accounted when its causing
abnormal termination.
Introduces nf_ct_kill_acct() which increments the accounting counters on
conntrack kill. The new function was necessary, because there are calls
to nf_ct_kill() which don't need accounting:
nf_conntrack_proto_tcp.c line ~847:
Kills ct and returns NF_REPEAT. We don't want to count twice.
nf_conntrack_proto_tcp.c line ~880:
Kills ct and returns NF_DROP. I think we don't want to count dropped
packets.
nf_conntrack_netlink.c line ~824:
As far as I can see ctnetlink_del_conntrack() is used to destroy a
conntrack on behalf of the user. There is an sk_buff, but I don't think
this is an actual packet. Incrementing counters here is therefore not
desired.
Signed-off-by: Fabian Hugelshofer <hugelshofer2006@gmx.ch>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Encapsulate the common
if (del_timer(&ct->timeout))
ct->timeout.function((unsigned long)ct)
sequence in a new function.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
The following patch implements a new "security" table for iptables, so
that MAC (SELinux etc.) networking rules can be managed separately to
standard DAC rules.
This is to help with distro integration of the new secmark-based
network controls, per various previous discussions.
The need for a separate table arises from the fact that existing tools
and usage of iptables will likely clash with centralized MAC policy
management.
The SECMARK and CONNSECMARK targets will still be valid in the mangle
table to prevent breakage of existing users.
Signed-off-by: James Morris <jmorris@namei.org>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
- Don't trust a length which is greater than the working buffer.
An invalid length could cause overflow when calculating buffer size
for decoding oid.
- An oid length of zero is invalid and allows for an off-by-one error when
decoding oid because the first subid actually encodes first 2 subids.
- A primitive encoding may not have an indefinite length.
Thanks to Wei Wang from McAfee for report.
Cc: Steven French <sfrench@us.ibm.com>
Cc: stable@kernel.org
Acked-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
skb_splice_bits temporary drops the socket lock while iterating over
the socket queue in order to break a reverse locking condition which
happens with sendfile. This, however, opens a window of opportunity
for tcp_collapse() to aggregate skbs and thus potentially free the
current skb used in skb_splice_bits and tcp_read_sock.
This patch fixes the problem by (re-)getting the same "logical skb"
after the lock has been temporary dropped.
Based on idea and initial patch from Evgeniy Polyakov.
Signed-off-by: Octavian Purdila <opurdila@ixiacom.com>
Acked-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: David S. Miller <davem@davemloft.net>
TCP "resets sent" counter is not incremented when a TCP Reset is
sent via tcp_send_active_reset().
Signed-off-by: Sridhar Samudrala <sri@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The program below just leaks the raw kernel socket
int main() {
int fd = socket(PF_INET, SOCK_RAW, IPPROTO_UDP);
struct sockaddr_in addr;
memset(&addr, 0, sizeof(addr));
inet_aton("127.0.0.1", &addr.sin_addr);
addr.sin_family = AF_INET;
addr.sin_port = htons(2048);
sendto(fd, "a", 1, MSG_MORE, &addr, sizeof(addr));
return 0;
}
Corked packet is allocated via sock_wmalloc which holds the owner socket,
so one should uncork it and flush all pending data on close. Do this in the
same way as in UDP.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Acked-by: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Signed-off-by: David S. Miller <davem@davemloft.net>
This bug is able to corrupt fackets_out in very rare cases.
In order for this to cause corruption:
1) DSACK in the middle of previous SACK block must be generated.
2) In order to take that particular branch, part or all of the
DSACKed segment must already be SACKed so that we have that
in cache in the first place.
3) The new info must be top enough so that fackets_out will be
updated on this iteration.
...then fack_count is updated while skb wasn't, then we walk again
that particular segment thus updating fack_count twice for
a single skb and finally that value is assigned to fackets_out
by tcp_sacktag_one.
It is safe to call tcp_sacktag_one just once for a segment (at
DSACK), no need to call again for plain SACK.
Potential problem of the miscount are limited to premature entry
to recovery and to inflated reordering metric (which could even
cancel each other out in the most the luckiest scenarios :-)).
Both are quite insignificant in worst case too and there exists
also code to reset them (fackets_out once sacked_out becomes zero
and reordering metric on RTO).
This has been reported by a number of people, because it occurred
quite rarely, it has been very evasive. Andy Furniss was able to
get it to occur couple of times so that a bit more info was
collected about the problem using a debug patch, though it still
required lot of checking around. Thanks also to others who have
tried to help here.
This is listed as Bugzilla #10346. The bug was introduced by
me in commit 68f8353b48 ([TCP]: Rewrite SACK block processing &
sack_recv_cache use), I probably thought back then that there's
need to scan that entry twice or didn't dare to make it go
through it just once there. Going through twice would have
required restoring fack_count after the walk but as noted above,
I chose to drop the additional walk step altogether here.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
IPv6 UDP sockets wth IPv4 mapped address use udp_sendmsg to send the data
actually. In this case ip_flush_pending_frames should be called instead
of ip6_flush_pending_frames.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
It is possible that this skip path causes TCP to end up into an
invalid state where ca_state was left to CA_Open while some
segments already came into sacked_out. If next valid ACK doesn't
contain new SACK information TCP fails to enter into
tcp_fastretrans_alert(). Thus at least high_seq is set
incorrectly to a too high seqno because some new data segments
could be sent in between (and also, limited transmit is not
being correctly invoked there). Reordering in both directions
can easily cause this situation to occur.
I guess we would want to use tcp_moderate_cwnd(tp) there as well
as it may be possible to use this to trigger oversized burst to
network by sending an old ACK with huge amount of SACK info, but
I'm a bit unsure about its effects (mainly to FlightSize), so to
be on the safe side I just currently fixed it minimally to keep
TCP's state consistent (obviously, such nasty ACKs have been
possible this far). Though it seems that FlightSize is already
underestimated by some amount, so probably on the long term we
might want to trigger recovery there too, if appropriate, to make
FlightSize calculation to resemble reality at the time when the
losses where discovered (but such change scares me too much now
and requires some more thinking anyway how to do that as it
likely involves some code shuffling).
This bug was found by Brian Vowell while running my TCP debug
patch to find cause of another TCP issue (fackets_out
miscount).
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
The field was supposed to allow the creation of an anycast route by
assigning an anycast address to an address prefix. It was never
implemented so this field is unused and serves no purpose. Remove it.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
