When an application needs to force a source IP on an active TCP socket
it has to use bind(IP, port=x).
As most applications do not want to deal with already used ports, x is
often set to 0, meaning the kernel is in charge to find an available
port.
But kernel does not know yet if this socket is going to be a listener or
be connected.
It has very limited choices (no full knowledge of final 4-tuple for a
connect())
With limited ephemeral port range (about 32K ports), it is very easy to
fill the space.
This patch adds a new SOL_IP socket option, asking kernel to ignore
the 0 port provided by application in bind(IP, port=0) and only
remember the given IP address.
The port will be automatically chosen at connect() time, in a way
that allows sharing a source port as long as the 4-tuples are unique.
This new feature is available for both IPv4 and IPv6 (Thanks Neal)
Tested:
Wrote a test program and checked its behavior on IPv4 and IPv6.
strace(1) shows sequences of bind(IP=127.0.0.2, port=0) followed by
connect().
Also getsockname() show that the port is still 0 right after bind()
but properly allocated after connect().
socket(PF_INET, SOCK_STREAM, IPPROTO_IP) = 5
setsockopt(5, SOL_IP, IP_BIND_ADDRESS_NO_PORT, [1], 4) = 0
bind(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("127.0.0.2")}, 16) = 0
getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("127.0.0.2")}, [16]) = 0
connect(5, {sa_family=AF_INET, sin_port=htons(53174), sin_addr=inet_addr("127.0.0.3")}, 16) = 0
getsockname(5, {sa_family=AF_INET, sin_port=htons(38050), sin_addr=inet_addr("127.0.0.2")}, [16]) = 0
IPv6 test :
socket(PF_INET6, SOCK_STREAM, IPPROTO_IP) = 7
setsockopt(7, SOL_IP, IP_BIND_ADDRESS_NO_PORT, [1], 4) = 0
bind(7, {sa_family=AF_INET6, sin6_port=htons(0), inet_pton(AF_INET6, "::1", &sin6_addr), sin6_flowinfo=0, sin6_scope_id=0}, 28) = 0
getsockname(7, {sa_family=AF_INET6, sin6_port=htons(0), inet_pton(AF_INET6, "::1", &sin6_addr), sin6_flowinfo=0, sin6_scope_id=0}, [28]) = 0
connect(7, {sa_family=AF_INET6, sin6_port=htons(57300), inet_pton(AF_INET6, "::1", &sin6_addr), sin6_flowinfo=0, sin6_scope_id=0}, 28) = 0
getsockname(7, {sa_family=AF_INET6, sin6_port=htons(60964), inet_pton(AF_INET6, "::1", &sin6_addr), sin6_flowinfo=0, sin6_scope_id=0}, [28]) = 0
I was able to bind()/connect() a million concurrent IPv4 sockets,
instead of ~32000 before patch.
lpaa23:~# ulimit -n 1000010
lpaa23:~# ./bind --connect --num-flows=1000000 &
1000000 sockets
lpaa23:~# grep TCP /proc/net/sockstat
TCP: inuse 2000063 orphan 0 tw 47 alloc 2000157 mem 66
Check that a given source port is indeed used by many different
connections :
lpaa23:~# ss -t src :40000 | head -10
State Recv-Q Send-Q Local Address:Port Peer Address:Port
ESTAB 0 0 127.0.0.2:40000 127.0.202.33:44983
ESTAB 0 0 127.0.0.2:40000 127.2.27.240:44983
ESTAB 0 0 127.0.0.2:40000 127.2.98.5:44983
ESTAB 0 0 127.0.0.2:40000 127.0.124.196:44983
ESTAB 0 0 127.0.0.2:40000 127.2.139.38:44983
ESTAB 0 0 127.0.0.2:40000 127.1.59.80:44983
ESTAB 0 0 127.0.0.2:40000 127.3.6.228:44983
ESTAB 0 0 127.0.0.2:40000 127.0.38.53:44983
ESTAB 0 0 127.0.0.2:40000 127.1.197.10:44983
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A long standing problem on busy servers is the tiny available TCP port
range (/proc/sys/net/ipv4/ip_local_port_range) and the default
sequential allocation of source ports in connect() system call.
If a host is having a lot of active TCP sessions, chances are
very high that all ports are in use by at least one flow,
and subsequent bind(0) attempts fail, or have to scan a big portion of
space to find a slot.
In this patch, I changed the starting point in __inet_hash_connect()
so that we try to favor even [1] ports, leaving odd ports for bind()
users.
We still perform a sequential search, so there is no guarantee, but
if connect() targets are very different, end result is we leave
more ports available to bind(), and we spread them all over the range,
lowering time for both connect() and bind() to find a slot.
This strategy only works well if /proc/sys/net/ipv4/ip_local_port_range
is even, ie if start/end values have different parity.
Therefore, default /proc/sys/net/ipv4/ip_local_port_range was changed to
32768 - 60999 (instead of 32768 - 61000)
There is no change on security aspects here, only some poor hashing
schemes could be eventually impacted by this change.
[1] : The odd/even property depends on ip_local_port_range values parity
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now that sk_alloc knows when a kernel socket is being allocated modify
it to not reference count the network namespace of kernel sockets.
Keep track of if a socket needs reference counting by adding a flag to
struct sock called sk_net_refcnt.
Update all of the callers of sock_create_kern to stop using
sk_change_net and sk_release_kernel as those hacks are no longer
needed, to avoid reference counting a kernel socket.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In preparation for changing how struct net is refcounted
on kernel sockets pass the knowledge that we are creating
a kernel socket from sock_create_kern through to sk_alloc.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is long overdue, and is part of cleaning up how we allocate kernel
sockets that don't reference count struct net.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The ipv4 code uses a mixture of coding styles. In some instances check
for non-NULL pointer is done as x != NULL and sometimes as x. x is
preferred according to checkpatch and this patch makes the code
consistent by adopting the latter form.
No changes detected by objdiff.
Signed-off-by: Ian Morris <ipm@chirality.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
The ipv4 code uses a mixture of coding styles. In some instances check
for NULL pointer is done as x == NULL and sometimes as !x. !x is
preferred according to checkpatch and this patch makes the code
consistent by adopting the latter form.
No changes detected by objdiff.
Signed-off-by: Ian Morris <ipm@chirality.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
After TIPC doesn't depend on iocb argument in its internal
implementations of sendmsg() and recvmsg() hooks defined in proto
structure, no any user is using iocb argument in them at all now.
Then we can drop the redundant iocb argument completely from kinds of
implementations of both sendmsg() and recvmsg() in the entire
networking stack.
Cc: Christoph Hellwig <hch@lst.de>
Suggested-by: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As part of an effort to move skb->dropcount to skb->cb[] use a common
macro in protocol families using skb->cb[] for ancillary data to
validate available room in skb->cb[].
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Receive Flow Steering is a nice solution but suffers from
hash collisions when a mix of connected and unconnected traffic
is received on the host, when flow hash table is populated.
Also, clearing flow in inet_release() makes RFS not very good
for short lived flows, as many packets can follow close().
(FIN , ACK packets, ...)
This patch extends the information stored into global hash table
to not only include cpu number, but upper part of the hash value.
I use a 32bit value, and dynamically split it in two parts.
For host with less than 64 possible cpus, this gives 6 bits for the
cpu number, and 26 (32-6) bits for the upper part of the hash.
Since hash bucket selection use low order bits of the hash, we have
a full hash match, if /proc/sys/net/core/rps_sock_flow_entries is big
enough.
If the hash found in flow table does not match, we fallback to RPS (if
it is enabled for the rxqueue).
This means that a packet for an non connected flow can avoid the
IPI through a unrelated/victim CPU.
This also means we no longer have to clear the table at socket
close time, and this helps short lived flows performance.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
TCP timestamping introduced MSG_ERRQUEUE handling for TCP sockets.
If the socket is of family AF_INET6, call ipv6_recv_error instead
of ip_recv_error.
This change is more complex than a single branch due to the loadable
ipv6 module. It reuses a pre-existing indirect function call from
ping. The ping code is safe to call, because it is part of the core
ipv6 module and always present when AF_INET6 sockets are active.
Fixes: 4ed2d765 (net-timestamp: TCP timestamping)
Signed-off-by: Willem de Bruijn <willemb@google.com>
----
It may also be worthwhile to add WARN_ON_ONCE(sk->family == AF_INET6)
to ip_recv_error.
Signed-off-by: David S. Miller <davem@davemloft.net>
Device can export MPLS GSO support in dev->mpls_features same way
it export vlan features in dev->vlan_features. So it is safe to
remove NETIF_F_GSO_MPLS redundant flag.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Add a new GSO type, SKB_GSO_TUNNEL_REMCSUM, which indicates remote
checksum offload being done (in this case inner checksum must not
be offloaded to the NIC).
Added logic in __skb_udp_tunnel_segment to handle remote checksum
offload case.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
skb_gso_segment() has a 'features' argument representing offload features
available to the output path.
A few handlers, e.g. GRE, instead re-fetch the features of skb->dev and use
those instead of the provided ones when handing encapsulation/tunnels.
Depending on dev->hw_enc_features of the output device skb_gso_segment() can
then return NULL even when the caller has disabled all GSO feature bits,
as segmentation of inner header thinks device will take care of segmentation.
This e.g. affects the tbf scheduler, which will silently drop GRE-encap GSO skbs
that did not fit the remaining token quota as the segmentation does not work
when device supports corresponding hw offload capabilities.
Cc: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Proper CHECKSUM_COMPLETE support needs to adjust skb->csum
when we remove one header. Its done using skb_gro_postpull_rcsum()
In the case of IPv4, we know that the adjustment is not really needed,
because the checksum over IPv4 header is 0. Lets add a comment to
ease code comprehension and avoid copy/paste errors.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The send_check logic was only interesting in cases of TCP offload and
UDP UFO where the checksum needed to be initialized to the pseudo
header checksum. Now we've moved that logic into the related
gso_segment functions so gso_send_check is no longer needed.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add inet_gro_receive and inet_gro_complete to ipip_offload to
support GRO.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net.ipv4.ip_nonlocal_bind sysctl was global to all network
namespaces. This patch allows to set a different value for each
network namespace.
Signed-off-by: Vincent Bernat <vincent@bernat.im>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fixed a bug that was introduced by my GRE-GRO patch
(bf5a755f5e net-gre-gro: Add GRE
support to the GRO stack) that breaks the forwarding path
because various GSO related fields were not set. The bug will
cause on the egress path either the GSO code to fail, or a
GRE-TSO capable (NETIF_F_GSO_GRE) NICs to choke. The following
fix has been tested for both cases.
Signed-off-by: H.K. Jerry Chu <hkchu@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Call gso_make_checksum. This should have the benefit of using a
checksum that may have been previously computed for the packet.
This also adds NETIF_F_GSO_GRE_CSUM to differentiate devices that
offload GRE GSO with and without the GRE checksum offloaed.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Added a new netif feature for GSO_UDP_TUNNEL_CSUM. This indicates
that a device is capable of computing the UDP checksum in the
encapsulating header of a UDP tunnel.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
It doesn't seem like an protocols are setting anything other
than the default, and allowing to arbitrarily disable checksums
for a whole protocol seems dangerous. This can be done on a per
socket basis.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
ip_local_port_range is already per netns, so should ip_local_reserved_ports
be. And since it is none by default we don't actually need it when we don't
enable CONFIG_SYSCTL.
By the way, rename inet_is_reserved_local_port() to inet_is_local_reserved_port()
Cc: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/ethernet/altera/altera_sgdma.c
net/netlink/af_netlink.c
net/sched/cls_api.c
net/sched/sch_api.c
The netlink conflict dealt with moving to netlink_capable() and
netlink_ns_capable() in the 'net' tree vs. supporting 'tc' operations
in non-init namespaces. These were simple transformations from
netlink_capable to netlink_ns_capable.
The Altera driver conflict was simply code removal overlapping some
void pointer cast cleanups in net-next.
Signed-off-by: David S. Miller <davem@davemloft.net>
Similarly, when CONFIG_SYSCTL is not set, ping_group_range should still
work, just that no one can change it. Therefore we should move it out of
sysctl_net_ipv4.c. And, it should not share the same seqlock with
ip_local_port_range.
BTW, rename it to ->ping_group_range instead.
Cc: David S. Miller <davem@davemloft.net>
Cc: Francois Romieu <romieu@fr.zoreil.com>
Reported-by: Stefan de Konink <stefan@konink.de>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When CONFIG_SYSCTL is not set, ip_local_port_range should still work,
just that no one can change it. Therefore we should move it out of sysctl_inet.c.
Also, rename it to ->ip_local_ports instead.
Cc: David S. Miller <davem@davemloft.net>
Cc: Francois Romieu <romieu@fr.zoreil.com>
Reported-by: Stefan de Konink <stefan@konink.de>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
commit 8f0ea0fe3a (snmp: reduce percpu needs by 50%)
reduced snmp array size to 1, so technically it doesn't have to be
an array any more. What's more, after the following commit:
commit 933393f58f
Date: Thu Dec 22 11:58:51 2011 -0600
percpu: Remove irqsafe_cpu_xxx variants
We simply say that regular this_cpu use must be safe regardless of
preemption and interrupt state. That has no material change for x86
and s390 implementations of this_cpu operations. However, arches that
do not provide their own implementation for this_cpu operations will
now get code generated that disables interrupts instead of preemption.
probably no arch wants to have SNMP_ARRAY_SZ == 2. At least after
almost 3 years, no one complains.
So, just convert the array to a single pointer and remove snmp_mib_init()
and snmp_mib_free() as well.
Cc: Christoph Lameter <cl@linux.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Replace the bh safe variant with the hard irq safe variant.
We need a hard irq safe variant to deal with netpoll transmitting
packets from hard irq context, and we need it in most if not all of
the places using the bh safe variant.
Except on 32bit uni-processor the code is exactly the same so don't
bother with a bh variant, just have a hard irq safe variant that
everyone can use.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently the UFO fragmentation process does not correctly handle inner
UDP frames.
(The following tcpdumps are captured on the parent interface with ufo
disabled while tunnel has ufo enabled, 2000 bytes payload, mtu 1280,
both sit device):
IPv6:
16:39:10.031613 IP (tos 0x0, ttl 64, id 3208, offset 0, flags [DF], proto IPv6 (41), length 1300)
192.168.122.151 > 1.1.1.1: IP6 (hlim 64, next-header Fragment (44) payload length: 1240) 2001::1 > 2001::8: frag (0x00000001:0|1232) 44883 > distinct: UDP, length 2000
16:39:10.031709 IP (tos 0x0, ttl 64, id 3209, offset 0, flags [DF], proto IPv6 (41), length 844)
192.168.122.151 > 1.1.1.1: IP6 (hlim 64, next-header Fragment (44) payload length: 784) 2001::1 > 2001::8: frag (0x00000001:0|776) 58979 > 46366: UDP, length 5471
We can see that fragmentation header offset is not correctly updated.
(fragmentation id handling is corrected by 916e4cf46d ("ipv6: reuse
ip6_frag_id from ip6_ufo_append_data")).
IPv4:
16:39:57.737761 IP (tos 0x0, ttl 64, id 3209, offset 0, flags [DF], proto IPIP (4), length 1296)
192.168.122.151 > 1.1.1.1: IP (tos 0x0, ttl 64, id 57034, offset 0, flags [none], proto UDP (17), length 1276)
192.168.99.1.35961 > 192.168.99.2.distinct: UDP, length 2000
16:39:57.738028 IP (tos 0x0, ttl 64, id 3210, offset 0, flags [DF], proto IPIP (4), length 792)
192.168.122.151 > 1.1.1.1: IP (tos 0x0, ttl 64, id 57035, offset 0, flags [none], proto UDP (17), length 772)
192.168.99.1.13531 > 192.168.99.2.20653: UDP, length 51109
In this case fragmentation id is incremented and offset is not updated.
First, I aligned inet_gso_segment and ipv6_gso_segment:
* align naming of flags
* ipv6_gso_segment: setting skb->encapsulation is unnecessary, as we
always ensure that the state of this flag is left untouched when
returning from upper gso segmenation function
* ipv6_gso_segment: move skb_reset_inner_headers below updating the
fragmentation header data, we don't care for updating fragmentation
header data
* remove currently unneeded comment indicating skb->encapsulation might
get changed by upper gso_segment callback (gre and udp-tunnel reset
encapsulation after segmentation on each fragment)
If we encounter an IPIP or SIT gso skb we now check for the protocol ==
IPPROTO_UDP and that we at least have already traversed another ip(6)
protocol header.
The reason why we have to special case GSO_IPIP and GSO_SIT is that
we reset skb->encapsulation to 0 while skb_mac_gso_segment the inner
protocol of GSO_UDP_TUNNEL or GSO_GRE packets.
Reported-by: Wolfgang Walter <linux@stwm.de>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Tom Herbert <therbert@google.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This new ip_no_pmtu_disc mode only allowes fragmentation-needed errors
to be honored by protocols which do more stringent validation on the
ICMP's packet payload. This knob is useful for people who e.g. want to
run an unmodified DNS server in a namespace where they need to use pmtu
for TCP connections (as they are used for zone transfers or fallback
for requests) but don't want to use possibly spoofed UDP pmtu information.
Currently the whitelisted protocols are TCP, SCTP and DCCP as they check
if the returned packet is in the window or if the association is valid.
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: David Miller <davem@davemloft.net>
Cc: John Heffner <johnwheffner@gmail.com>
Suggested-by: Florian Weimer <fweimer@redhat.com>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch built on top of Commit 299603e837
("net-gro: Prepare GRO stack for the upcoming tunneling support") to add
the support of the standard GRE (RFC1701/RFC2784/RFC2890) to the GRO
stack. It also serves as an example for supporting other encapsulation
protocols in the GRO stack in the future.
The patch supports version 0 and all the flags (key, csum, seq#) but
will flush any pkt with the S (seq#) flag. This is because the S flag
is not support by GSO, and a GRO pkt may end up in the forwarding path,
thus requiring GSO support to break it up correctly.
Currently the "packet_offload" structure only contains L3 (ETH_P_IP/
ETH_P_IPV6) GRO offload support so the encapped pkts are limited to
IP pkts (i.e., w/o L2 hdr). But support for other protocol type can
be easily added, so is the support for GRE variations like NVGRE.
The patch also support csum offload. Specifically if the csum flag is on
and the h/w is capable of checksumming the payload (CHECKSUM_COMPLETE),
the code will take advantage of the csum computed by the h/w when
validating the GRE csum.
Note that commit 60769a5dcd "ipv4: gre:
add GRO capability" already introduces GRO capability to IPv4 GRE
tunnels, using the gro_cells infrastructure. But GRO is done after
GRE hdr has been removed (i.e., decapped). The following patch applies
GRO when pkts first come in (before hitting the GRE tunnel code). There
is some performance advantage for applying GRO as early as possible.
Also this approach is transparent to other subsystem like Open vSwitch
where GRE decap is handled outside of the IP stack hence making it
harder for the gro_cells stuff to apply. On the other hand, some NICs
are still not capable of hashing on the inner hdr of a GRE pkt (RSS).
In that case the GRO processing of pkts from the same remote host will
all happen on the same CPU and the performance may be suboptimal.
I'm including some rough preliminary performance numbers below. Note
that the performance will be highly dependent on traffic load, mix as
usual. Moreover it also depends on NIC offload features hence the
following is by no means a comprehesive study. Local testing and tuning
will be needed to decide the best setting.
All tests spawned 50 copies of netperf TCP_STREAM and ran for 30 secs.
(super_netperf 50 -H 192.168.1.18 -l 30)
An IP GRE tunnel with only the key flag on (e.g., ip tunnel add gre1
mode gre local 10.246.17.18 remote 10.246.17.17 ttl 255 key 123)
is configured.
The GRO support for pkts AFTER decap are controlled through the device
feature of the GRE device (e.g., ethtool -K gre1 gro on/off).
1.1 ethtool -K gre1 gro off; ethtool -K eth0 gro off
thruput: 9.16Gbps
CPU utilization: 19%
1.2 ethtool -K gre1 gro on; ethtool -K eth0 gro off
thruput: 5.9Gbps
CPU utilization: 15%
1.3 ethtool -K gre1 gro off; ethtool -K eth0 gro on
thruput: 9.26Gbps
CPU utilization: 12-13%
1.4 ethtool -K gre1 gro on; ethtool -K eth0 gro on
thruput: 9.26Gbps
CPU utilization: 10%
The following tests were performed on a different NIC that is capable of
csum offload. I.e., the h/w is capable of computing IP payload csum
(CHECKSUM_COMPLETE).
2.1 ethtool -K gre1 gro on (hence will use gro_cells)
2.1.1 ethtool -K eth0 gro off; csum offload disabled
thruput: 8.53Gbps
CPU utilization: 9%
2.1.2 ethtool -K eth0 gro off; csum offload enabled
thruput: 8.97Gbps
CPU utilization: 7-8%
2.1.3 ethtool -K eth0 gro on; csum offload disabled
thruput: 8.83Gbps
CPU utilization: 5-6%
2.1.4 ethtool -K eth0 gro on; csum offload enabled
thruput: 8.98Gbps
CPU utilization: 5%
2.2 ethtool -K gre1 gro off
2.2.1 ethtool -K eth0 gro off; csum offload disabled
thruput: 5.93Gbps
CPU utilization: 9%
2.2.2 ethtool -K eth0 gro off; csum offload enabled
thruput: 5.62Gbps
CPU utilization: 8%
2.2.3 ethtool -K eth0 gro on; csum offload disabled
thruput: 7.69Gbps
CPU utilization: 8%
2.2.4 ethtool -K eth0 gro on; csum offload enabled
thruput: 8.96Gbps
CPU utilization: 5-6%
Signed-off-by: H.K. Jerry Chu <hkchu@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Steffen Klassert says:
====================
pull request (net-next): ipsec-next 2013-12-19
1) Use the user supplied policy index instead of a generated one
if present. From Fan Du.
2) Make xfrm migration namespace aware. From Fan Du.
3) Make the xfrm state and policy locks namespace aware. From Fan Du.
4) Remove ancient sleeping when the SA is in acquire state,
we now queue packets to the policy instead. This replaces the
sleeping code.
5) Remove FLOWI_FLAG_CAN_SLEEP. This was used to notify xfrm about the
posibility to sleep. The sleeping code is gone, so remove it.
6) Check user specified spi for IPComp. Thr spi for IPcomp is only
16 bit wide, so check for a valid value. From Fan Du.
7) Export verify_userspi_info to check for valid user supplied spi ranges
with pfkey and netlink. From Fan Du.
8) RFC3173 states that if the total size of a compressed payload and the IPComp
header is not smaller than the size of the original payload, the IP datagram
must be sent in the original non-compressed form. These packets are dropped
by the inbound policy check because they are not transformed. Document the need
to set 'level use' for IPcomp to receive such packets anyway. From Fan Du.
Please pull or let me know if there are problems.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The other field in ipv4_config, log_martians, was converted to a
per-interface setting, so we can just remove the whole structure.
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch modifies the GRO stack to avoid the use of "network_header"
and associated macros like ip_hdr() and ipv6_hdr() in order to allow
an arbitary number of IP hdrs (v4 or v6) to be used in the
encapsulation chain. This lays the foundation for various IP
tunneling support (IP-in-IP, GRE, VXLAN, SIT,...) to be added later.
With this patch, the GRO stack traversing now is mostly based on
skb_gro_offset rather than special hdr offsets saved in skb (e.g.,
skb->network_header). As a result all but the top layer (i.e., the
the transport layer) must have hdrs of the same length in order for
a pkt to be considered for aggregation. Therefore when adding a new
encap layer (e.g., for tunneling), one must check and skip flows
(e.g., by setting NAPI_GRO_CB(p)->same_flow to 0) that have a
different hdr length.
Note that unlike the network header, the transport header can and
will continue to be set by the GRO code since there will be at
most one "transport layer" in the encap chain.
Signed-off-by: H.K. Jerry Chu <hkchu@google.com>
Suggested-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
FLOWI_FLAG_CAN_SLEEP was used to notify xfrm about the posibility
to sleep until the needed states are resolved. This code is gone,
so FLOWI_FLAG_CAN_SLEEP is not needed anymore.
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Pull core locking changes from Ingo Molnar:
"The biggest changes:
- add lockdep support for seqcount/seqlocks structures, this
unearthed both bugs and required extra annotation.
- move the various kernel locking primitives to the new
kernel/locking/ directory"
* 'core-locking-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (21 commits)
block: Use u64_stats_init() to initialize seqcounts
locking/lockdep: Mark __lockdep_count_forward_deps() as static
lockdep/proc: Fix lock-time avg computation
locking/doc: Update references to kernel/mutex.c
ipv6: Fix possible ipv6 seqlock deadlock
cpuset: Fix potential deadlock w/ set_mems_allowed
seqcount: Add lockdep functionality to seqcount/seqlock structures
net: Explicitly initialize u64_stats_sync structures for lockdep
locking: Move the percpu-rwsem code to kernel/locking/
locking: Move the lglocks code to kernel/locking/
locking: Move the rwsem code to kernel/locking/
locking: Move the rtmutex code to kernel/locking/
locking: Move the semaphore core to kernel/locking/
locking: Move the spinlock code to kernel/locking/
locking: Move the lockdep code to kernel/locking/
locking: Move the mutex code to kernel/locking/
hung_task debugging: Add tracepoint to report the hang
x86/locking/kconfig: Update paravirt spinlock Kconfig description
lockstat: Report avg wait and hold times
lockdep, x86/alternatives: Drop ancient lockdep fixup message
...
While testing virtio_net and skb_segment() changes, Hannes reported
that UFO was sending wrong frames.
It appears this was introduced by a recent commit :
8c3a897bfa ("inet: restore gso for vxlan")
The old condition to perform IP frag was :
tunnel = !!skb->encapsulation;
...
if (!tunnel && proto == IPPROTO_UDP) {
So the new one should be :
udpfrag = !skb->encapsulation && proto == IPPROTO_UDP;
...
if (udpfrag) {
Initialization of udpfrag must be done before call
to ops->callbacks.gso_segment(skb, features), as
skb_udp_tunnel_segment() clears skb->encapsulation
(We want udpfrag to be true for UFO, false for VXLAN)
With help from Alexei Starovoitov
Reported-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to enable lockdep on seqcount/seqlock structures, we
must explicitly initialize any locks.
The u64_stats_sync structure, uses a seqcount, and thus we need
to introduce a u64_stats_init() function and use it to initialize
the structure.
This unfortunately adds a lot of fairly trivial initialization code
to a number of drivers. But the benefit of ensuring correctness makes
this worth while.
Because these changes are required for lockdep to be enabled, and the
changes are quite trivial, I've not yet split this patch out into 30-some
separate patches, as I figured it would be better to get the various
maintainers thoughts on how to best merge this change along with
the seqcount lockdep enablement.
Feedback would be appreciated!
Signed-off-by: John Stultz <john.stultz@linaro.org>
Acked-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>
Cc: James Morris <jmorris@namei.org>
Cc: Jesse Gross <jesse@nicira.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Mirko Lindner <mlindner@marvell.com>
Cc: Patrick McHardy <kaber@trash.net>
Cc: Roger Luethi <rl@hellgate.ch>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Simon Horman <horms@verge.net.au>
Cc: Stephen Hemminger <stephen@networkplumber.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Wensong Zhang <wensong@linux-vs.org>
Cc: netdev@vger.kernel.org
Link: http://lkml.kernel.org/r/1381186321-4906-2-git-send-email-john.stultz@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Alexei reported a performance regression on vxlan, caused
by commit 3347c96029 "ipv4: gso: make inet_gso_segment() stackable"
GSO vxlan packets were not properly segmented, adding IP fragments
while they were not expected.
Rename 'bool tunnel' to 'bool encap', and add a new boolean
to express the fact that UDP should be fragmented.
This fragmentation is triggered by skb->encapsulation being set.
Remove a "skb->encapsulation = 1" added in above commit,
as its not needed, as frags inherit skb->frag from original
GSO skb.
Reported-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Tested-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now ipv6_gso_segment() is stackable, its relatively easy to
implement GSO/TSO support for SIT tunnels
Performance results, when segmentation is done after tunnel
device (as no NIC is yet enabled for TSO SIT support) :
Before patch :
lpq84:~# ./netperf -H 2002:af6:1153:: -Cc
MIGRATED TCP STREAM TEST from ::0 (::) port 0 AF_INET6 to 2002:af6:1153:: () port 0 AF_INET6
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
87380 16384 16384 10.00 3168.31 4.81 4.64 2.988 2.877
After patch :
lpq84:~# ./netperf -H 2002:af6:1153:: -Cc
MIGRATED TCP STREAM TEST from ::0 (::) port 0 AF_INET6 to 2002:af6:1153:: () port 0 AF_INET6
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
87380 16384 16384 10.00 5525.00 7.76 5.17 2.763 1.840
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The code that is implemented is per memory cgroup not per netns, and
having per netns bits is just confusing. Remove the per netns bits to
make it easier to see what is really going on.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Initialize the ehash and ipv6_hash_secrets with net_get_random_once.
Each compilation unit gets its own secret now:
ipv4/inet_hashtables.o
ipv4/udp.o
ipv6/inet6_hashtables.o
ipv6/udp.o
rds/connection.o
The functions still get inlined into the hashing functions. In the fast
path we have at most two (needed in ipv6) if (unlikely(...)).
Cc: Eric Dumazet <edumazet@google.com>
Cc: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now inet_gso_segment() is stackable, its relatively easy to
implement GSO/TSO support for IPIP
Performance results, when segmentation is done after tunnel
device (as no NIC is yet enabled for TSO IPIP support) :
Before patch :
lpq83:~# ./netperf -H 7.7.9.84 -Cc
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.9.84 () 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
87380 16384 16384 10.00 3357.88 5.09 3.70 2.983 2.167
After patch :
lpq83:~# ./netperf -H 7.7.9.84 -Cc
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.9.84 () 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
87380 16384 16384 10.00 7710.19 4.52 6.62 1.152 1.687
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to support GSO on IPIP, we need to make
inet_gso_segment() stackable.
It should not assume network header starts right after mac
header.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
inet_gso_segment() and inet_gso_send_check() are called by
skb_mac_gso_segment() under rcu lock, no need to use
rcu_read_lock() / rcu_read_unlock()
Avoid calling ip_hdr() twice per function.
We can use ip_send_check() helper.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The removal of the routing cache introduced a performance regression for
some UDP workloads since a dst lookup must be done for each packet.
This change caches the dst per socket in a similar manner to what we do
for TCP by implementing early_demux.
For UDP multicast we can only cache the dst if there is only one
receiving socket on the host. Since caching only works when there is
one receiving socket we do the multicast socket lookup using RCU.
For UDP unicast we only demux sockets with an exact match in order to
not break forwarding setups. Additionally since the hash chains may be
long we only check the first socket to see if it is a match and not
waste extra time searching the whole chain when we might not find an
exact match.
Benchmark results from a netperf UDP_RR test:
Before 87961.22 transactions/s
After 89789.68 transactions/s
Benchmark results from a fio 1 byte UDP multicast pingpong test
(Multicast one way unicast response):
Before 12.97us RTT
After 12.63us RTT
Signed-off-by: Shawn Bohrer <sbohrer@rgmadvisors.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A host might need net_secret[] and never open a single socket.
Problem added in commit aebda156a5
("net: defer net_secret[] initialization")
Based on prior patch from Hannes Frederic Sowa.
Reported-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Hannes Frederic Sowa <hannes@strressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fengguang reported:
net/built-in.o: In function `in6_dev_finish_destroy':
(.text+0x4ca7d): undefined reference to `snmp_mib_free'
this is due to snmp_mib_free() is defined when CONFIG_INET is enabled,
but in6_dev_finish_destroy() is now moved to core kernel.
I think snmp_mib_free() is small enough to be inlined, so just make it
static inline.
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Cong Wang <amwang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix following sparse error :
net/ipv4/af_inet.c:1410:59: warning: restricted __be16 degrades to
integer
added in commit db8caf3dbc
("gro: should aggregate frames without DF")
Reported-by: kbuild test robot <fengguang.wu@intel.com>
From: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet
Eric W. Biederman
Ian Morris
Ying Xue
Eyal Birger
David S. Miller
Willem de Bruijn
Pravin B Shelar
Tom Herbert
Florian Westphal
Vincent Bernat
Jerry Chu
WANG Cong
Hannes Frederic Sowa
Steffen Klassert
Linus Torvalds
John Stultz
Shawn Bohrer
Cong Wang