commit bcf49342 ("netfilter: ebtables: Fix extension lookup with
identical name") added a second lookup in case the extension that was
found during the first lookup matched another extension with the same
name, but didn't release the reference on the incorrect module.

Fixes: bcf49342 ("netfilter: ebtables: Fix extension lookup with identical name")
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Acked-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

"meta pkttype set" is only supported on prerouting chain with bridge
family and ingress chain with netdev family.

But the validate check is incomplete, and the user can add the nft
rules on input chain with bridge family, for example:
  # nft add table bridge filter
  # nft add chain bridge filter input {type filter hook input \
    priority 0 \;}
  # nft add chain bridge filter test
  # nft add rule bridge filter test meta pkttype set unicast
  # nft add rule bridge filter input jump test

This patch fixes the problem.
Signed-off-by: Liping Zhang <liping.zhang@spreadtrum.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

KASAN reported this bug:
  BUG: KASAN: use-after-free in icmp_packet+0x25/0x50 [nf_conntrack_ipv4] at
  addr ffff880002db08c8
  Read of size 4 by task lt-nf-queue/19041
  Call Trace:
  <IRQ>  [<ffffffff815eeebb>] dump_stack+0x63/0x88
  [<ffffffff813386f8>] kasan_report_error+0x528/0x560
  [<ffffffff81338cc8>] kasan_report+0x58/0x60
  [<ffffffffa07393f5>] ? icmp_packet+0x25/0x50 [nf_conntrack_ipv4]
  [<ffffffff81337551>] __asan_load4+0x61/0x80
  [<ffffffffa07393f5>] icmp_packet+0x25/0x50 [nf_conntrack_ipv4]
  [<ffffffffa06ecaa0>] nf_conntrack_in+0x550/0x980 [nf_conntrack]
  [<ffffffffa06ec550>] ? __nf_conntrack_confirm+0xb10/0xb10 [nf_conntrack]
  [ ... ]

The main reason is that we missed to unlink the timeout objects in the
unconfirmed ct lists, so we will access the timeout objects that have
already been freed.
Signed-off-by: Liping Zhang <liping.zhang@spreadtrum.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

We forget to call nf_ct_l4proto_put when replacing the existing
timeout policy. Acctually, there's no need to get ct l4proto
before doing replace, so we can move it to a later position.
Signed-off-by: Liping Zhang <liping.zhang@spreadtrum.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

cttimeout and acct objects are deleted from the list while traversing
it, so use list_for_each_entry is unsafe here.
Signed-off-by: Liping Zhang <liping.zhang@spreadtrum.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

After I add the nft rule "nft add rule filter prerouting reject
with tcp reset", kernel panic happened on my system:
  NULL pointer dereference at ...
  IP: [<ffffffff81b9db2f>] nf_send_reset+0xaf/0x400
  Call Trace:
  [<ffffffff81b9da80>] ? nf_reject_ip_tcphdr_get+0x160/0x160
  [<ffffffffa0928061>] nft_reject_ipv4_eval+0x61/0xb0 [nft_reject_ipv4]
  [<ffffffffa08e836a>] nft_do_chain+0x1fa/0x890 [nf_tables]
  [<ffffffffa08e8170>] ? __nft_trace_packet+0x170/0x170 [nf_tables]
  [<ffffffffa06e0900>] ? nf_ct_invert_tuple+0xb0/0xc0 [nf_conntrack]
  [<ffffffffa07224d4>] ? nf_nat_setup_info+0x5d4/0x650 [nf_nat]
  [...]

Because in the PREROUTING chain, routing information is not exist,
then we will dereference the NULL pointer and oops happen.

So we restrict reject expression to INPUT, FORWARD and OUTPUT chain.
This is consistent with iptables REJECT target.
Signed-off-by: Liping Zhang <liping.zhang@spreadtrum.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Make the function mlxsw_router_neigh_construct search the rif according
to the neighbour dev other than the dev that was passed to the ndo, thus
allowing creating neigbhours upon stacked devices.

Fixes: 6cf3c971 ("mlxsw: spectrum_router: Add private neigh table")
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In case we have a layer 3 interface on top of a bridge (VLAN / FID RIF),
then we should flood the following packet types to the router:

* Broadcast: If DIP is the broadcast address of the interface, then we
need to be able to get it to CPU by trapping it following route lookup.

* Reserved IP multicast (224.0.0.X): Some control packets (e.g. OSPF)
use this range and are trapped in the router block.

Fixes: 99f44bb3 ("mlxsw: spectrum: Enable L3 interfaces on top of bridge devices")
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

inet_diag_find_one_icsk takes a reference to a socket that is not
released if sock_diag_destroy returns an error. Fix by changing
tcp_diag_destroy to manage the refcnt for all cases and remove
the sock_put calls from tcp_abort.

Fixes: c1e64e29 ("net: diag: Support destroying TCP sockets")
Reported-by: Lorenzo Colitti <lorenzo@google.com>
Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Instead of using sock_tx_timestamp, use skb_tx_timestamp to record
software transmit timestamp of a packet.

sock_tx_timestamp resets and overrides the tx_flags of the skb.
The function is intended to be called from within the protocol
layer when creating the skb, not from a device driver. This is
inconsistent with other drivers and will cause issues for TCP.

In TCP, we intend to sample the timestamps for the last byte
for each sendmsg/sendpage. For that reason, tcp_sendmsg calls
tcp_tx_timestamp only with the last skb that it generates.
For example, if a 128KB message is split into two 64KB packets
we want to sample the SND timestamp of the last packet. The current
code in the tun driver, however, will result in sampling the SND
timestamp for both packets.

Also, when the last packet is split into smaller packets for
retranmission (see tcp_fragment), the tun driver will record
timestamps for all of the retransmitted packets and not only the
last packet.

Fixes: eda29772 (tun: Support software transmit time stamping.)
Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: Francis Yan <francisyyan@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

After commit ca065d0c ("udp: no longer use SLAB_DESTROY_BY_RCU")
we do not need this special allocation mode anymore, even if it is
harmless.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The function sctp_diag_dump_one() currently performs a memcpy()
of 64 bytes from a 16 byte field into another 16 byte field. Fix
by using correct size, use sizeof to obtain correct size instead
of using a hard-coded constant.

Fixes: 8f840e47 ("sctp: add the sctp_diag.c file")
Signed-off-by: Lance Richardson <lrichard@redhat.com>
Reviewed-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

We currently enable interrupts before we enable NAPI. If an RX interrupt
hits before we enabled NAPI then the NAPI callback is never called and
we leave the hardware with RX interrupts disabled, which of course leads
us to never handling received packets.  Fix this by moving the interrupt
enable to after we've enable NAPI and the reclaim tasklet.

Fixes: cd5e4123 ("dwc_eth_qos: do phy_start before resetting hardware")
Signed-off-by: Rabin Vincent <rabinv@axis.com>
Signed-off-by: Lars Persson <larper@axis.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

clk_prepare_enable() may fail, so we should better check its return
value and propagate it in the case of failure

While at it, replace __lpc_eth_clock_enable() with a plain
clk_prepare_enable/clk_disable_unprepare() call in order to
simplify the code.
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Acked-by: Vladimir Zapolskiy <vz@mleia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The PORT_RATE_CONTROL register works differently on 88e6095/6095f/6131
in comparison to 6123/61/65, and 0x0 disables. The distinction was lost
Linux 4.1 --> 4.2
Signed-off-by: Jamie Lentin <jm@lentin.co.uk>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Like the ksz8081, the ksz9031 has the behavior where it will clear the
interrupt enable bits when leaving power down. This takes advantage of the
solution provided by f5aba91d.
Signed-off-by: Xander Huff <xander.huff@ni.com>
Signed-off-by: Nathan Sullivan <nathan.sullivan@ni.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

When sending an ack in SYN_RECV state, we must scale the offered
window if wscale option was negotiated and accepted.

Tested:
 Following packetdrill test demonstrates the issue :

0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0

+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0

// Establish a connection.
+0 < S 0:0(0) win 20000 <mss 1000,sackOK,wscale 7, nop, TS val 100 ecr 0>
+0 > S. 0:0(0) ack 1 win 28960 <mss 1460,sackOK, TS val 100 ecr 100, nop, wscale 7>

+0 < . 1:11(10) ack 1 win 156 <nop,nop,TS val 99 ecr 100>
// check that window is properly scaled !
+0 > . 1:1(0) ack 1 win 226 <nop,nop,TS val 200 ecr 100>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The current scatter-gather logic in gianfar is flawed, since
it does not consider the eTSEC's RxBD 'Data Length' field is
context depening: for the last fragment it contains the full
frame size, while fragments contain the fragment size, which
equals the value written to register MRBLR.

This causes data corruption as soon as the hardware starts
to fragment receiving frames. As a result, the size of
fragmented frames is increased by
(nr_frags - 1) * MRBLR

We first noticed this issue working with DSA, where an ICMP
request sized 1472 bytes causes the scatter-gather logic to
kick in. The full Ethernet frame (1518) gets increased by
DSA (4), GMAC_FCB_LEN (8), and FSL_GIANFAR_DEV_HAS_TIMER
(priv->padding=8) to a total of 1538 octets, which is
fragmented by the hardware and reconstructed by the driver
to a 3074 octet frame.

This patch fixes the problem by adjusting the size of
the last fragment.

It was tested by setting MRBLR to different multiples of
64, proving correct scatter-gather operation on frames
with up to 9000 octets in size.
Signed-off-by: Zefir Kurtisi <zefir.kurtisi@neratec.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The eTSEC register MRBLR defines the maximum space in
the RX buffers and is set to 1536 by gianfar. This
reasonably covers the common use case where the MTU
is kept at default 1500. In that case, the largest
Ethernet frame size of 1518 plus an optional
GMAC_FCB_LEN of 8, and an additional padding of 8
to handle FSL_GIANFAR_DEV_HAS_TIMER totals to 1534
and nicely fit within the chosen MRBLR.

Alas, if the eTSEC is attached to a DSA enabled switch,
the (E)DSA header extension (4 or 8 bytes) causes every
maximum sized frame to be fragmented by the hardware.

This patch increases the maximum RX buffer size by 8
and rounds up to the next multiple of 64, which the
hardware's defines as RX buffer granularity.
Signed-off-by: Zefir Kurtisi <zefir.kurtisi@neratec.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Laura tracked poll() [and friends] regression caused by commit
e6afc8ac ("udp: remove headers from UDP packets before queueing")

udp_poll() needs to know if there is a valid packet in receive queue,
even if its payload length is 0.

Change first_packet_length() to return an signed int, and use -1
as the indication of an empty queue.

Fixes: e6afc8ac ("udp: remove headers from UDP packets before queueing")
Reported-by: Laura Abbott <labbott@redhat.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Tested-by: Laura Abbott <labbott@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Encoding of the metadata was using the padded length as opposed to
the real length of the data which is a bug per specification.
This has not been an issue todate because all metadatum specified
so far has been 32 bit where aligned and data length are the same width.
This also includes a bug fix for validating the length of a u16 field.
But since there is no metadata of size u16 yes we are fine to include it
here.

While at it get rid of magic numbers.

Fixes: ef6980b6 ("net sched: introduce IFE action")
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Driver never bothered marking the VF's vport with the VF's sw_fid.
As a result, FLR flows are not going to clean those vports.

If the vport was active when FLRed, re-activating it would lead
to a FW assertion.

Fixes: dacd88d6 ("qed: IOV l2 functionality")
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In b8247f09,

   "net: ip_finish_output_gso: If skb_gso_network_seglen exceeds MTU, allow segmentation for local udp tunneled skbs"

gso skbs arriving from an ingress interface that go through UDP
tunneling, are allowed to be fragmented if the resulting encapulated
segments exceed the dst mtu of the egress interface.

This aligned the behavior of gso skbs to non-gso skbs going through udp
encapsulation path.

However the non-gso vs gso anomaly is present also in the following
cases of a GRE tunnel:
 - ip_gre in collect_md mode, where TUNNEL_DONT_FRAGMENT is not set
   (e.g. OvS vport-gre with df_default=false)
 - ip_gre in nopmtudisc mode, where IFLA_GRE_IGNORE_DF is set

In both of the above cases, the non-gso skbs get fragmented, whereas the
gso skbs (having skb_gso_network_seglen that exceeds dst mtu) get dropped,
as they don't go through the segment+fragment code path.

Fix: Setting IPSKB_FRAG_SEGS if the tunnel specified IP_DF bit is NOT set.

Tunnels that do set IP_DF, will not go to fragmentation of segments.
This preserves behavior of ip_gre in (the default) pmtudisc mode.

Fixes: b8247f09 ("net: ip_finish_output_gso: If skb_gso_network_seglen exceeds MTU, allow segmentation for local udp tunneled skbs")
Reported-by: wenxu <wenxu@ucloud.cn>
Cc: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Shmulik Ladkani <shmulik.ladkani@gmail.com>
Tested-by: wenxu <wenxu@ucloud.cn>
Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

If DAD fails with accept_dad set to 2, global addresses and host routes
are incorrectly left in place. Even though disable_ipv6 is set,
contrary to documentation, the addresses are not dynamically deleted
from the interface. It is only on a subsequent link down/up that these
are removed. The fix is not only to set the disable_ipv6 flag, but
also to call addrconf_ifdown(), which is the action to carry out when
disabling IPv6. This results in the addresses and routes being deleted
immediately. The DAD failure for the LL addr is determined as before
via netlink, or by the absence of the LL addr (which also previously
would have had to be checked for in case of an intervening link down
and up). As the call to addrconf_ifdown() requires an rtnl lock, the
logic to disable IPv6 when DAD fails is moved to addrconf_dad_work().

Previous behavior:

root@vm1:/# sysctl net.ipv6.conf.eth3.accept_dad=2
net.ipv6.conf.eth3.accept_dad = 2
root@vm1:/# ip -6 addr add 2000::10/64 dev eth3
root@vm1:/# ip link set up eth3
root@vm1:/# ip -6 addr show dev eth3
5: eth3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000
    inet6 2000::10/64 scope global
       valid_lft forever preferred_lft forever
    inet6 fe80::5054:ff:fe43:dd5a/64 scope link tentative dadfailed
       valid_lft forever preferred_lft forever
root@vm1:/# ip -6 route show dev eth3
2000::/64  proto kernel  metric 256
fe80::/64  proto kernel  metric 256
root@vm1:/# ip link set down eth3
root@vm1:/# ip link set up eth3
root@vm1:/# ip -6 addr show dev eth3
root@vm1:/# ip -6 route show dev eth3
root@vm1:/#

New behavior:

root@vm1:/# sysctl net.ipv6.conf.eth3.accept_dad=2
net.ipv6.conf.eth3.accept_dad = 2
root@vm1:/# ip -6 addr add 2000::10/64 dev eth3
root@vm1:/# ip link set up eth3
root@vm1:/# ip -6 addr show dev eth3
root@vm1:/# ip -6 route show dev eth3
root@vm1:/#
Signed-off-by: Mike Manning <mmanning@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes these compiler warnings via libc-compat.h when glibc netipx/ipx.h is
included before linux/ipx.h:

./linux/ipx.h:9:8: error: redefinition of ‘struct sockaddr_ipx’
./linux/ipx.h:26:8: error: redefinition of ‘struct ipx_route_definition’
./linux/ipx.h:32:8: error: redefinition of ‘struct ipx_interface_definition’
./linux/ipx.h:49:8: error: redefinition of ‘struct ipx_config_data’
./linux/ipx.h:58:8: error: redefinition of ‘struct ipx_route_def’
Signed-off-by: Mikko Rapeli <mikko.rapeli@iki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>

Kernel uapi header are supposed to use them. Fixes userspace compile error:

linux/openvswitch.h:583:2: error: unknown type name ‘uint32_t’
Signed-off-by: Mikko Rapeli <mikko.rapeli@iki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes userspace compile error:

error: field ‘real’ has incomplete type
 struct timeval real;  /* real (wall-clock) time */
Signed-off-by: Mikko Rapeli <mikko.rapeli@iki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes userspace compiler error:

error: unknown type name ‘uint32_t’
Signed-off-by: Mikko Rapeli <mikko.rapeli@iki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes userspace compilation errors:

error: field ‘addr’ has incomplete type
 struct sockaddr_in addr; /* IP address and port to send to */

error: field ‘addr’ has incomplete type
 struct sockaddr_in6 addr; /* IP address and port to send to */
Signed-off-by: Mikko Rapeli <mikko.rapeli@iki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes userspace compilation errors like:

error: field ‘addr’ has incomplete type
 struct sockaddr_in addr; /* IP address and port to send to */
                    ^
error: field ‘addr’ has incomplete type
 struct sockaddr_in6 addr; /* IP address and port to send to */
Signed-off-by: Mikko Rapeli <mikko.rapeli@iki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes userspace compilation errors like:

error: field ‘iph’ has incomplete type
error: field ‘prefix’ has incomplete type
Signed-off-by: Mikko Rapeli <mikko.rapeli@iki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes userspace compilation error:

error: ‘IFNAMSIZ’ undeclared here (not in a function)
Signed-off-by: Mikko Rapeli <mikko.rapeli@iki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>

trivial typo fix in pr_err message
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

If the address configured in the device tree is invalid, the
driver will fallback to using a random address from the locally
administered range.
Signed-off-by: Daniel Romell <daro@hms.se>
Signed-off-by: David S. Miller <davem@davemloft.net>

'Commit 3c8b3efc ("vmxnet3: allow variable length transmit data ring
buffer")' changed the size of the buffers in the tx data ring from a
fixed size of 128 bytes to a variable size.

However, while copying data to the data ring, vmxnet3_copy_hdr continues
to carry the old code that assumes fixed buffer size of 128. This patch
fixes it by adding correct offset based on the actual data ring buffer
size.
Signed-off-by: Guolin Yang <gyang@vmware.com>
Signed-off-by: Shrikrishna Khare <skhare@vmware.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The RAR entry for the SAN MAC address was being cleared when we were
clearing the VMDq pool bits.  In order to prevent this we need to add
an extra check to protect the SAN MAC from being cleared.

Fixes: 6e982aea ("ixgbe: Clear stale pool mappings")
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pool index has to be converted by get_pool helper to work correctly for
egress pool. In mlxsw the egress pool index starts from 0.

Fixes: 0f433fa0 ("mlxsw: spectrum_buffers: Implement shared buffer configuration")
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The sk->sk_state is bits flag, so need use bit operation check
instead of value check.
Signed-off-by: Gao Feng <fgao@ikuai8.com>
Tested-by: Guillaume Nault <g.nault@alphalink.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>

Because otherwise when crc computation is still needed it's way more
expensive than on a linear buffer to the point that it affects
performance.

It's so expensive that netperf test gives a perf output as below:

Overhead  Command         Shared Object       Symbol
  18,62%  netserver       [kernel.vmlinux]    [k] crc32_generic_shift
   2,57%  netserver       [kernel.vmlinux]    [k] __pskb_pull_tail
   1,94%  netserver       [kernel.vmlinux]    [k] fib_table_lookup
   1,90%  netserver       [kernel.vmlinux]    [k] copy_user_enhanced_fast_string
   1,66%  swapper         [kernel.vmlinux]    [k] intel_idle
   1,63%  netserver       [kernel.vmlinux]    [k] _raw_spin_lock
   1,59%  netserver       [sctp]              [k] sctp_packet_transmit
   1,55%  netserver       [kernel.vmlinux]    [k] memcpy_erms
   1,42%  netserver       [sctp]              [k] sctp_rcv

# netperf -H 192.168.10.1 -l 10 -t SCTP_STREAM -cC -- -m 12000
SCTP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 192.168.10.1 () port 0 AF_INET
Recv   Send    Send                          Utilization       Service Demand
Socket Socket  Message  Elapsed              Send     Recv     Send    Recv
Size   Size    Size     Time     Throughput  local    remote   local   remote
bytes  bytes   bytes    secs.    10^6bits/s  % S      % S      us/KB   us/KB

212992 212992  12000    10.00      3016.42   2.88     3.78     1.874   2.462

After patch:
Overhead  Command         Shared Object      Symbol
   2,75%  netserver       [kernel.vmlinux]   [k] memcpy_erms
   2,63%  netserver       [kernel.vmlinux]   [k] copy_user_enhanced_fast_string
   2,39%  netserver       [kernel.vmlinux]   [k] fib_table_lookup
   2,04%  netserver       [kernel.vmlinux]   [k] __pskb_pull_tail
   1,91%  netserver       [kernel.vmlinux]   [k] _raw_spin_lock
   1,91%  netserver       [sctp]             [k] sctp_packet_transmit
   1,72%  netserver       [mlx4_en]          [k] mlx4_en_process_rx_cq
   1,68%  netserver       [sctp]             [k] sctp_rcv

# netperf -H 192.168.10.1 -l 10 -t SCTP_STREAM -cC -- -m 12000
SCTP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 192.168.10.1 () port 0 AF_INET
Recv   Send    Send                          Utilization       Service Demand
Socket Socket  Message  Elapsed              Send     Recv     Send    Recv
Size   Size    Size     Time     Throughput  local    remote   local   remote
bytes  bytes   bytes    secs.    10^6bits/s  % S      % S      us/KB   us/KB

212992 212992  12000    10.00      3681.77   3.83     3.46     2.045   1.849

Fixes: 3acb50c1 ("sctp: delay as much as possible skb_linearize")
Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Saeed Mahameed says:

====================
Mellanox 100G mlx5 fixes 2016-08-16

This series includes some bug fixes for mlx5e driver.

From Saeed and Tariq, Optimize MTU change to not reset when it is not required.

From Paul, Command interface message length check to speedup firmware
command preparation.

From Mohamad, Save pci state when pci error is detected.

From Amir, Flow counters "lastuse" update fix.

From Hadar, Use correct flow dissector key on flower offloading.
Plus a small optimization for switchdev hardware id query.

From Or, three patches to address some E-Switch offloads issues.

For -stable of 4.6.y and 4.7.y:
    net/mlx5e: Use correct flow dissector key on flower offloading
    net/mlx5: Fix pci error recovery flow
    net/mlx5: Added missing check of msg length in verifying its signature
====================
Signed-off-by: David S. Miller <davem@davemloft.net>