early_demux() handlers should be called in RCU context, and as we
use skb_dst_set_noref(skb, dst), caller must not exit from RCU context
before dst use (skb_dst(skb)) or release (skb_drop(dst))
Therefore, rcu_read_lock()/rcu_read_unlock() pairs around
->early_demux() are confusing and not needed :
Protocol handlers are already in an RCU read lock section.
(__netif_receive_skb() does the rcu_read_lock() )
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When userspace use RTM_GETROUTE to dump route table, with an already
expired route entry, we always got an 'expires' value(2147157)
calculated base on INT_MAX.
The reason of this problem is in the following satement:
rt->dst.expires - jiffies < INT_MAX
gcc promoted the type of both sides of '<' to unsigned long, thus
a small negative value would be considered greater than INT_MAX.
With the help of Eric Dumazet, do the out of bound checks in
rtnl_put_cacheinfo(), _after_ conversion to clock_t.
Signed-off-by: Li Wei <lw@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is the IPv6 missing bits for infrastructure added in commit
41063e9dd1 (ipv4: Early TCP socket demux.)
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
ICMP messages generated in output path if frame length is bigger than
mtu are actually lost because socket is owned by user (doing the xmit)
One example is the ipgre_tunnel_xmit() calling
icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, htonl(mtu));
We had a similar case fixed in commit a34a101e1e (ipv6: disable GSO on
sockets hitting dst_allfrag).
Problem of such fix is that it relied on retransmit timers, so short tcp
sessions paid a too big latency increase price.
This patch uses the tcp_release_cb() infrastructure so that MTU
reduction messages (ICMP messages) are not lost, and no extra delay
is added in TCP transmits.
Reported-by: Maciej Żenczykowski <maze@google.com>
Diagnosed-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Cc: Tom Herbert <therbert@google.com>
Cc: Tore Anderson <tore@fud.no>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a big comment explaining how the field works, and use defines
instead of magic constants for the values assigned to it.
Suggested by Joe Perches.
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch impelements the common code for both the client and server.
1. TCP Fast Open option processing. Since Fast Open does not have an
option number assigned by IANA yet, it shares the experiment option
code 254 by implementing draft-ietf-tcpm-experimental-options
with a 16 bits magic number 0xF989. This enables global experiments
without clashing the scarce(2) experimental options available for TCP.
When the draft status becomes standard (maybe), the client should
switch to the new option number assigned while the server supports
both numbers for transistion.
2. The new sysctl tcp_fastopen
3. A place holder init function
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Introduce ipv6_addr_hash() helper doing a XOR on all bits
of an IPv6 address, with an optimized x86_64 version.
Use it in flow dissector, as suggested by Andrew McGregor,
to reduce hash collision probabilities in fq_codel (and other
users of flow dissector)
Use it in ip6_tunnel.c and use more bit shuffling, as suggested
by David Laight, as existing hash was ignoring most of them.
Use it in sunrpc and use more bit shuffling, using hash_32().
Use it in net/ipv6/addrconf.c, using hash_32() as well.
As a cleanup, use it in net/ipv4/tcp_metrics.c
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Andrew McGregor <andrewmcgr@gmail.com>
Cc: Dave Taht <dave.taht@gmail.com>
Cc: Tom Herbert <therbert@google.com>
Cc: David Laight <David.Laight@ACULAB.COM>
Cc: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We should provide to inet6_csk_route_socket a struct flowi6 pointer,
so that net6_csk_xmit() works correctly instead of sending garbage.
Also add some consts
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
These patches implement the final mechanism necessary to really allow
us to go without the route cache in ipv4.
We need a place to have long-term storage of PMTU/redirect information
which is independent of the routes themselves, yet does not get us
back into a situation where we have to write to metrics or anything
like that.
For this we use an "next-hop exception" table in the FIB nexthops.
The one thing I desperately want to avoid is having to create clone
routes in the FIB trie for this purpose, because that is very
expensive. However, I'm willing to entertain such an idea later
if this current scheme proves to have downsides that the FIB trie
variant would not have.
In order to accomodate this any such scheme, we need to be able to
produce a full flow key at PMTU/redirect time. That required an
adjustment of the interface call-sites used to propagate these events.
For a PMTU/redirect with a fully specified socket, we pass that socket
and use it to produce the flow key.
Otherwise we use a passed in SKB to formulate the key. There are two
cases that need to be distinguished, ICMP message processing (in which
case the IP header is at skb->data) and output packet processing
(mostly tunnels, and in all such cases the IP header is at ip_hdr(skb)).
We also have to make the code able to handle the case where the dst
itself passed into the dst_ops->{update_pmtu,redirect} method is
invalidated. This matters for calls from sockets that have cached
that route. We provide a inet{,6} helper function for this purpose,
and edit SCTP specially since it caches routes at the transport rather
than socket level.
Signed-off-by: David S. Miller <davem@davemloft.net>
This will be used so that we can compose a full flow key.
Even though we have a route in this context, we need more. In the
future the routes will be without destination address, source address,
etc. keying. One ipv4 route will cover entire subnets, etc.
In this environment we have to have a way to possess persistent storage
for redirects and PMTU information. This persistent storage will exist
in the FIB tables, and that's why we'll need to be able to rebuild a
full lookup flow key here. Using that flow key will do a fib_lookup()
and create/update the persistent entry.
Signed-off-by: David S. Miller <davem@davemloft.net>
We need to check the passed in multicast address and return
appropriate errno(EINVAL) if it is not valid. And it's no need
to walk through the ipv6_mc_list in this situation.
Signed-off-by: Li Wei <lw@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Userspace implementations of network routing protocols sometimes need to
tell RA-originated IPv6 routes from other kernel routes to make proper
routing decisions. This makes most sense for RA routes with nexthops,
namely, default routes and Route Information routes.
The intended mean of preserving RA route origin in a netlink message is
through indicating RTPROT_RA as protocol code. Function rt6_fill_node()
tried to do that for default routes, but its test condition was taken
wrong. This change is modeled after the original mailing list posting
by Jeff Haran. It fixes the test condition for default route case and
sets the same behaviour for Route Information case (both types use
nexthops). Handling of the 3rd RA route type, Prefix Information, is
left unchanged, as it stands for interface connected routes (without
nexthops).
Signed-off-by: Denis Ovsienko <infrastation@yandex.ru>
Signed-off-by: David S. Miller <davem@davemloft.net>
We start initializing the struct rt6_info at the first field
behind the struct dst_enty. This is error prone because it
might leave a new field uninitialized. So start initializing
the struct rt6_info right behind the dst_entry.
Suggested-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This sets things up so that we can have the protocol error handlers
call down into the ipv6 route code for redirects just as ipv4 already
does.
Signed-off-by: David S. Miller <davem@davemloft.net>
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix incorrect start markers, wrapped summary lines, missing section
breaks, incorrect separators, and some name mismatches.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
No longer needed. TCP writes metrics, but now in it's own special
cache that does not dirty the route metrics. Therefore there is no
longer any reason to pre-cow metrics in this way.
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix a bug in ip6_dst_lookup_tail(), where typeof(dst) is
"struct dst_entry **", not "struct dst_entry *"
Reported-by: Fengguang Wu <wfg@linux.intel.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
remove redundant declarations, they belong in include/net/tcp.h
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
git commit 97cac082 (ipv6: Store route neighbour in rt6_info struct)
added a neighbour pointer to rt6_info. Currently we don't initialize
this pointer at allocation time. We assume this pointer to be valid
if it is not a null pointer, so initialize it on allocation.
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
opt always equals np->opts, so it is meaningless to define opt, and
check if opt does not equal np->opts and then try to free opt.
Signed-off-by: RongQing.Li <roy.qing.li@gmail.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This makes for a simplified conversion away from dst_get_neighbour*().
All code outside of ipv6 will use neigh lookups via dst_neigh_lookup*().
Signed-off-by: David S. Miller <davem@davemloft.net>
Causes the handler to use the daddr in the ipv4/ipv6 header when
the route gateway is unspecified (local subnet).
Signed-off-by: David S. Miller <davem@davemloft.net>
When a dst_confirm() happens, mark the confirmation as pending in the
dst. Then on the next packet out, when we have the neigh in-hand, do
the update.
This removes the dependency in dst_confirm() of dst's having an
attached neigh.
While we're here, remove the explicit 'dst' NULL check, all except 2
or 3 call sites ensure it's not NULL. So just fix those cases up.
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch generalizes nf_ct_l4proto_net by splitting it into chunks and
moving the corresponding protocol part to where it really belongs to.
To clarify, note that we follow two different approaches to support per-net
depending if it's built-in or run-time loadable protocol tracker.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Acked-by: Gao feng <gaofeng@cn.fujitsu.com>
At Facebook, we do Layer-3 DSR via IP-in-IP tunneling. Our load balancers wrap
an extra IP header on incoming packets so they can be routed to the backend.
In the v4 tunnel driver, when these packets fall on the default tunl0 device,
the behavior is to decapsulate them and drop them back on the stack. So our
setup is that tunl0 has the VIP and eth0 has (obviously) the backend's real
address.
In IPv6 we do the same thing, but the v6 tunnel driver didn't have this same
behavior - if you didn't have an explicit tunnel setup, it would drop the
packet.
This patch brings that v4 feature to the v6 driver.
The same IPv6 address checks are performed as with any normal tunnel,
but as the fallback tunnel endpoint addresses are unspecified, the checks
must be performed on a per-packet basis, rather than at tunnel
configuration time.
[Patch description modified by phil@ipom.com]
Signed-off-by: Ville Nuorvala <ville.nuorvala@gmail.com>
Tested-by: Phil Dibowitz <phil@ipom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The code in tcp_v6_conn_request() was implicitly assuming that
tcp_v6_send_synack() would take care of dst_release(), much as
tcp_v4_send_synack() already does. This resulted in
tcp_v6_conn_request() leaking a dst if sysctl_tw_recycle is enabled.
This commit restructures tcp_v6_send_synack() so that it accepts a dst
pointer and takes care of releasing the dst that is passed in, to plug
the leak and avoid future surprises by bringing the IPv6 behavior in
line with the IPv4 side.
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With the recent change (earlier in this patch series) to set
flowi6_oif to treq->iif in inet6_csk_route_req(), the dst lookup in
these two functions is now identical, so tcp_v6_send_synack() can now
just call inet6_csk_route_req(), to reduce code duplication and keep
things closer to the IPv4 side, which is structured this way.
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit changes inet_csk_route_req() so that it uses a pointer to
a struct flowi6, rather than allocating its own on the stack. This
brings its behavior in line with its IPv4 cousin,
inet_csk_route_req(), and allows a follow-on patch to fix a dst leak.
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix inet6_csk_route_req() to use as the flowi6_oif the treq->iif,
which is correctly fixed up in tcp_v6_conn_request() to handle the
case of link-local addresses. This brings it in line with the
tcp_v6_send_synack() code, which is already correctly using the
treq->iif in this way.
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/caif/caif_hsi.c
drivers/net/usb/qmi_wwan.c
The qmi_wwan merge was trivial.
The caif_hsi.c, on the other hand, was not. It's a conflict between
1c385f1fdf ("caif-hsi: Replace platform
device with ops structure.") in the net-next tree and commit
39abbaef19 ("caif-hsi: Postpone init of
HIS until open()") in the net tree.
I did my best with that one and will ask Sjur to check it out.
Signed-off-by: David S. Miller <davem@davemloft.net>
dropwatch wrongly diagnose all received UDP packets as drops.
This patch removes trace_kfree_skb() done in skb_free_datagram_locked().
Locations calling skb_free_datagram_locked() should do it on their own.
As a result, drops are accounted on the right function.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Split sysctl function into smaller chucks to cleanup code and prepare
patches to reduce ifdef pollution.
Signed-off-by: Gao feng <gaofeng@cn.fujitsu.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
l4proto->init contain quite redundant code. We can simplify this
by adding a new parameter l3proto.
This patch prepares that code simplification.
Signed-off-by: Gao feng <gaofeng@cn.fujitsu.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>