With support from the master key option in the previous commit, it
becomes easy to make frequent updates/exchanges of session keys between
authenticated cluster nodes.
Basically, there are two situations where the key exchange will take in
place:

- When a new node joins the cluster (with the master key), it will need
  to get its peer's TX key, so that be able to decrypt further messages
  from that peer.

- When a new session key is generated (by either user manual setting or
  later automatic rekeying feature), the key will be distributed to all
  peer nodes in the cluster.

A key to be exchanged is encapsulated in the data part of a 'MSG_CRYPTO
/KEY_DISTR_MSG' TIPC v2 message, then xmit-ed as usual and encrypted by
using the master key before sending out. Upon receipt of the message it
will be decrypted in the same way as regular messages, then attached as
the sender's RX key in the receiver node.

In this way, the key exchange is reliable by the link layer, as well as
security, integrity and authenticity by the crypto layer.

Also, the forward security will be easily achieved by user changing the
master key actively but this should not be required very frequently.

The key exchange feature is independent on the presence of a master key
Note however that the master key still is needed for new nodes to be
able to join the cluster. It is also optional, and can be turned off/on
via the sysfs: 'net/tipc/key_exchange_enabled' [default 1: enabled].

Backward compatibility is guaranteed because for nodes that do not have
master key support, key exchange using master key ie. tx_key = 0 if any
will be shortly discarded at the message validation step. In other
words, the key exchange feature will be automatically disabled to those
nodes.

v2: fix the "implicit declaration of function 'tipc_crypto_key_flush'"
error in node.c. The function only exists when built with the TIPC
"CONFIG_TIPC_CRYPTO" option.

v3: use 'info->extack' for a message emitted due to netlink operations
instead (- David's comment).
Reported-by: Nkernel test robot <lkp@intel.com>
Acked-by: NJon Maloy <jmaloy@redhat.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In addition to the supported cluster & per-node encryption keys for the
en/decryption of TIPC messages, we now introduce one option for user to
set a cluster key as 'master key', which is simply a symmetric key like
the former but has a longer life cycle. It has two purposes:

- Authentication of new member nodes in the cluster. New nodes, having
  no knowledge of current session keys in the cluster will still be
  able to join the cluster as long as they know the master key. This is
  because all neighbor discovery (LINK_CONFIG) messages must be
  encrypted with this key.

- Encryption of session encryption keys during automatic exchange and
  update of those.This is a feature we will introduce in a later commit
  in this series.

We insert the new key into the currently unused slot 0 in the key array
and start using it immediately once the user has set it.
After joining, a node only knowing the master key should be fully
communicable to existing nodes in the cluster, although those nodes may
have their own session keys activated (i.e. not the master one). To
support this, we define a 'grace period', starting from the time a node
itself reports having no RX keys, so the existing nodes will use the
master key for encryption instead. The grace period can be extended but
will automatically stop after e.g. 5 seconds without a new report. This
is also the basis for later key exchanging feature as the new node will
be impossible to decrypt anything without the support from master key.

For user to set a master key, we define a new netlink flag -
'TIPC_NLA_NODE_KEY_MASTER', so it can be added to the current 'set key'
netlink command to specify the setting key to be a master key.

Above all, the traditional cluster/per-node key mechanism is guaranteed
to work when user comes not to use this master key option. This is also
compatible to legacy nodes without the feature supported.

Even this master key can be updated without any interruption of cluster
connectivity but is so is needed, this has to be coordinated and set by
the user.
Acked-by: NJon Maloy <jmaloy@redhat.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Make use of the struct_size() helper instead of an open-coded version
in order to avoid any potential type mistakes.

This code was detected with the help of Coccinelle and, audited and
fixed manually.
Signed-off-by: NGustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, updating binding table (add service binding to
name table/withdraw a service binding) is being sent over replicast.
However, if we are scaling up clusters to > 100 nodes/containers this
method is less affection because of looping through nodes in a cluster one
by one.

It is worth to use broadcast to update a binding service. This way, the
binding table can be updated on all peer nodes in one shot.

Broadcast is used when all peer nodes, as indicated by a new capability
flag TIPC_NAMED_BCAST, support reception of this message type.

Four problems need to be considered when introducing this feature.
1) When establishing a link to a new peer node we still update this by a
unicast 'bulk' update. This may lead to race conditions, where a later
broadcast publication/withdrawal bypass the 'bulk', resulting in
disordered publications, or even that a withdrawal may arrive before the
corresponding publication. We solve this by adding an 'is_last_bulk' bit
in the last bulk messages so that it can be distinguished from all other
messages. Only when this message has arrived do we open up for reception
of broadcast publications/withdrawals.

2) When a first legacy node is added to the cluster all distribution
will switch over to use the legacy 'replicast' method, while the
opposite happens when the last legacy node leaves the cluster. This
entails another risk of message disordering that has to be handled. We
solve this by adding a sequence number to the broadcast/replicast
messages, so that disordering can be discovered and corrected. Note
however that we don't need to consider potential message loss or
duplication at this protocol level.

3) Bulk messages don't contain any sequence numbers, and will always
arrive in order. Hence we must exempt those from the sequence number
control and deliver them unconditionally. We solve this by adding a new
'is_bulk' bit in those messages so that they can be recognized.

4) Legacy messages, which don't contain any new bits or sequence
numbers, but neither can arrive out of order, also need to be exempt
from the initial synchronization and sequence number check, and
delivered unconditionally. Therefore, we add another 'is_not_legacy' bit
to all new messages so that those can be distinguished from legacy
messages and the latter delivered directly.

v1->v2:
 - fix warning issue reported by kbuild test robot <lkp@intel.com>
 - add santiy check to drop the publication message with a sequence
number that is lower than the agreed synch point
Signed-off-by: Nkernel test robot <lkp@intel.com>
Signed-off-by: NHoang Huu Le <hoang.h.le@dektech.com.au>
Acked-by: NJon Maloy <jmaloy@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When streaming in Nagle mode, we try to bundle small messages from user
as many as possible if there is one outstanding buffer, i.e. not ACK-ed
by the receiving side, which helps boost up the overall throughput. So,
the algorithm's effectiveness really depends on when Nagle ACK comes or
what the specific network latency (RTT) is, compared to the user's
message sending rate.

In a bad case, the user's sending rate is low or the network latency is
small, there will not be many bundles, so making a Nagle ACK or waiting
for it is not meaningful.
For example: a user sends its messages every 100ms and the RTT is 50ms,
then for each messages, we require one Nagle ACK but then there is only
one user message sent without any bundles.

In a better case, even if we have a few bundles (e.g. the RTT = 300ms),
but now the user sends messages in medium size, then there will not be
any difference at all, that says 3 x 1000-byte data messages if bundled
will still result in 3 bundles with MTU = 1500.

When Nagle is ineffective, the delay in user message sending is clearly
wasted instead of sending directly.

Besides, adding Nagle ACKs will consume some processor load on both the
sending and receiving sides.

This commit adds a test on the effectiveness of the Nagle algorithm for
an individual connection in the network on which it actually runs.
Particularly, upon receipt of a Nagle ACK we will compare the number of
bundles in the backlog queue to the number of user messages which would
be sent directly without Nagle. If the ratio is good (e.g. >= 2), Nagle
mode will be kept for further message sending. Otherwise, we will leave
Nagle and put a 'penalty' on the connection, so it will have to spend
more 'one-way' messages before being able to re-enter Nagle.

In addition, the 'ack-required' bit is only set when really needed that
the number of Nagle ACKs will be reduced during Nagle mode.

Testing with benchmark showed that with the patch, there was not much
difference in throughput for small messages since the tool continuously
sends messages without a break, so Nagle would still take in effect.
Acked-by: NYing Xue <ying.xue@windriver.com>
Acked-by: NJon Maloy <jmaloy@redhat.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This commit enables dumping the statistics of a broadcast-receiver link
like the traditional 'broadcast-link' one (which is for broadcast-
sender). The link dumping can be triggered via netlink (e.g. the
iproute2/tipc tool) by the link flag - 'TIPC_NLA_LINK_BROADCAST' as the
indicator.

The name of a broadcast-receiver link of a specific peer will be in the
format: 'broadcast-link:<peer-id>'.

For example:

Link <broadcast-link:1001002>
  Window:50 packets
  RX packets:7841 fragments:2408/440 bundles:0/0
  TX packets:0 fragments:0/0 bundles:0/0
  RX naks:0 defs:124 dups:0
  TX naks:21 acks:0 retrans:0
  Congestion link:0  Send queue max:0 avg:0

In addition, the broadcast-receiver link statistics can be reset in the
usual way via netlink by specifying that link name in command.

Note: the 'tipc_link_name_ext()' is removed because the link name can
now be retrieved simply via the 'l->name'.
Acked-by: NYing Xue <ying.xue@windriver.com>
Acked-by: NJon Maloy <jmaloy@redhat.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

As achieved through commit 9195948f ("tipc: improve TIPC throughput
by Gap ACK blocks"), we apply the same mechanism for the broadcast link
as well. The 'Gap ACK blocks' data field in a 'PROTOCOL/STATE_MSG' will
consist of two parts built for both the broadcast and unicast types:

 31                       16 15                        0
+-------------+-------------+-------------+-------------+
|  bgack_cnt  |  ugack_cnt  |            len            |
+-------------+-------------+-------------+-------------+  -
|            gap            |            ack            |   |
+-------------+-------------+-------------+-------------+    > bc gacks
:                           :                           :   |
+-------------+-------------+-------------+-------------+  -
|            gap            |            ack            |   |
+-------------+-------------+-------------+-------------+    > uc gacks
:                           :                           :   |
+-------------+-------------+-------------+-------------+  -

which is "automatically" backward-compatible.

We also increase the max number of Gap ACK blocks to 128, allowing upto
64 blocks per type (total buffer size = 516 bytes).

Besides, the 'tipc_link_advance_transmq()' function is refactored which
is applicable for both the unicast and broadcast cases now, so some old
functions can be removed and the code is optimized.

With the patch, TIPC broadcast is more robust regardless of packet loss
or disorder, latency, ... in the underlying network. Its performance is
boost up significantly.
For example, experiment with a 5% packet loss rate results:

$ time tipc-pipe --mc --rdm --data_size 123 --data_num 1500000
real    0m 42.46s
user    0m 1.16s
sys     0m 17.67s

Without the patch:

$ time tipc-pipe --mc --rdm --data_size 123 --data_num 1500000
real    8m 27.94s
user    0m 0.55s
sys     0m 2.38s
Acked-by: NYing Xue <ying.xue@windriver.com>
Acked-by: NJon Maloy <jmaloy@redhat.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In the commit f73b1281
("tipc: improve throughput between nodes in netns"), we're missing a check
to handle TIPC_DIRECT_MSG type, it's still using old sending mechanism for
this message type. So, throughput improvement is not significant as
expected.

Besides that, when sending a large message with that type, we're also
handle wrong receiving queue, it should be enqueued in socket receiving
instead of multicast messages.

Fix this by adding the missing case for TIPC_DIRECT_MSG.

Fixes: f73b1281 ("tipc: improve throughput between nodes in netns")
Reported-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NHoang Le <hoang.h.le@dektech.com.au>
Acked-by: NJon Maloy <jmaloy@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This commit offers an option to encrypt and authenticate all messaging,
including the neighbor discovery messages. The currently most advanced
algorithm supported is the AEAD AES-GCM (like IPSec or TLS). All
encryption/decryption is done at the bearer layer, just before leaving
or after entering TIPC.

Supported features:
- Encryption & authentication of all TIPC messages (header + data);
- Two symmetric-key modes: Cluster and Per-node;
- Automatic key switching;
- Key-expired revoking (sequence number wrapped);
- Lock-free encryption/decryption (RCU);
- Asynchronous crypto, Intel AES-NI supported;
- Multiple cipher transforms;
- Logs & statistics;

Two key modes:
- Cluster key mode: One single key is used for both TX & RX in all
nodes in the cluster.
- Per-node key mode: Each nodes in the cluster has one specific TX key.
For RX, a node requires its peers' TX key to be able to decrypt the
messages from those peers.

Key setting from user-space is performed via netlink by a user program
(e.g. the iproute2 'tipc' tool).

Internal key state machine:

                                 Attach    Align(RX)
                                     +-+   +-+
                                     | V   | V
        +---------+      Attach     +---------+
        |  IDLE   |---------------->| PENDING |(user = 0)
        +---------+                 +---------+
           A   A                   Switch|  A
           |   |                         |  |
           |   | Free(switch/revoked)    |  |
     (Free)|   +----------------------+  |  |Timeout
           |              (TX)        |  |  |(RX)
           |                          |  |  |
           |                          |  v  |
        +---------+      Switch     +---------+
        | PASSIVE |<----------------| ACTIVE  |
        +---------+       (RX)      +---------+
        (user = 1)                  (user >= 1)

The number of TFMs is 10 by default and can be changed via the procfs
'net/tipc/max_tfms'. At this moment, as for simplicity, this file is
also used to print the crypto statistics at runtime:

echo 0xfff1 > /proc/sys/net/tipc/max_tfms

The patch defines a new TIPC version (v7) for the encryption message (-
backward compatibility as well). The message is basically encapsulated
as follows:

   +----------------------------------------------------------+
   | TIPCv7 encryption  | Original TIPCv2    | Authentication |
   | header             | packet (encrypted) | Tag            |
   +----------------------------------------------------------+

The throughput is about ~40% for small messages (compared with non-
encryption) and ~9% for large messages. With the support from hardware
crypto i.e. the Intel AES-NI CPU instructions, the throughput increases
upto ~85% for small messages and ~55% for large messages.

By default, the new feature is inactive (i.e. no encryption) until user
sets a key for TIPC. There is however also a new option - "TIPC_CRYPTO"
in the kernel configuration to enable/disable the new code when needed.

MAINTAINERS | add two new files 'crypto.h' & 'crypto.c' in tipc
Acked-by: NYing Xue <ying.xue@windreiver.com>
Acked-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

As mentioned in commit e95584a8 ("tipc: fix unlimited bundling of
small messages"), the current message bundling algorithm is inefficient
that can generate bundles of only one payload message, that causes
unnecessary overheads for both the sender and receiver.

This commit re-designs the 'tipc_msg_make_bundle()' function (now named
as 'tipc_msg_try_bundle()'), so that when a message comes at the first
place, we will just check & keep a reference to it if the message is
suitable for bundling. The message buffer will be put into the link
backlog queue and processed as normal. Later on, when another one comes
we will make a bundle with the first message if possible and so on...
This way, a bundle if really needed will always consist of at least two
payload messages. Otherwise, we let the first buffer go its way without
any need of bundling, so reduce the overheads to zero.

Moreover, since now we have both the messages in hand, we can even
optimize the 'tipc_msg_bundle()' function, make bundle of a very large
(size ~ MSS) and small messages which is not with the current algorithm
e.g. [1400-byte message] + [10-byte message] (MTU = 1500).
Acked-by: NYing Xue <ying.xue@windreiver.com>
Acked-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We introduce a feature that works like a combination of TCP_NAGLE and
TCP_CORK, but without some of the weaknesses of those. In particular,
we will not observe long delivery delays because of delayed acks, since
the algorithm itself decides if and when acks are to be sent from the
receiving peer.

- The nagle property as such is determined by manipulating a new
  'maxnagle' field in struct tipc_sock. If certain conditions are met,
  'maxnagle' will define max size of the messages which can be bundled.
  If it is set to zero no messages are ever bundled, implying that the
  nagle property is disabled.
- A socket with the nagle property enabled enters nagle mode when more
  than 4 messages have been sent out without receiving any data message
  from the peer.
- A socket leaves nagle mode whenever it receives a data message from
  the peer.

In nagle mode, messages smaller than 'maxnagle' are accumulated in the
socket write queue. The last buffer in the queue is marked with a new
'ack_required' bit, which forces the receiving peer to send a CONN_ACK
message back to the sender upon reception.

The accumulated contents of the write queue is transmitted when one of
the following events or conditions occur.

- A CONN_ACK message is received from the peer.
- A data message is received from the peer.
- A SOCK_WAKEUP pseudo message is received from the link level.
- The write queue contains more than 64 1k blocks of data.
- The connection is being shut down.
- There is no CONN_ACK message to expect. I.e., there is currently
  no outstanding message where the 'ack_required' bit was set. As a
  consequence, the first message added after we enter nagle mode
  is always sent directly with this bit set.

This new feature gives a 50-100% improvement of throughput for small
(i.e., less than MTU size) messages, while it might add up to one RTT
to latency time when the socket is in nagle mode.
Acked-by: NYing Xue <ying.xue@windreiver.com>
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, TIPC transports intra-node user data messages directly
socket to socket, hence shortcutting all the lower layers of the
communication stack. This gives TIPC very good intra node performance,
both regarding throughput and latency.

We now introduce a similar mechanism for TIPC data traffic across
network namespaces located in the same kernel. On the send path, the
call chain is as always accompanied by the sending node's network name
space pointer. However, once we have reliably established that the
receiving node is represented by a namespace on the same host, we just
replace the namespace pointer with the receiving node/namespace's
ditto, and follow the regular socket receive patch though the receiving
node. This technique gives us a throughput similar to the node internal
throughput, several times larger than if we let the traffic go though
the full network stacks. As a comparison, max throughput for 64k
messages is four times larger than TCP throughput for the same type of
traffic.

To meet any security concerns, the following should be noted.

- All nodes joining a cluster are supposed to have been be certified
and authenticated by mechanisms outside TIPC. This is no different for
nodes/namespaces on the same host; they have to auto discover each
other using the attached interfaces, and establish links which are
supervised via the regular link monitoring mechanism. Hence, a kernel
local node has no other way to join a cluster than any other node, and
have to obey to policies set in the IP or device layers of the stack.

- Only when a sender has established with 100% certainty that the peer
node is located in a kernel local namespace does it choose to let user
data messages, and only those, take the crossover path to the receiving
node/namespace.

- If the receiving node/namespace is removed, its namespace pointer
is invalidated at all peer nodes, and their neighbor link monitoring
will eventually note that this node is gone.

- To ensure the "100% certainty" criteria, and prevent any possible
spoofing, received discovery messages must contain a proof that the
sender knows a common secret. We use the hash mix of the sending
node/namespace for this purpose, since it can be accessed directly by
all other namespaces in the kernel. Upon reception of a discovery
message, the receiver checks this proof against all the local
namespaces'hash_mix:es. If it finds a match, that, along with a
matching node id and cluster id, this is deemed sufficient proof that
the peer node in question is in a local namespace, and a wormhole can
be opened.

- We should also consider that TIPC is intended to be a cluster local
IPC mechanism (just like e.g. UNIX sockets) rather than a network
protocol, and hence we think it can justified to allow it to shortcut the
lower protocol layers.

Regarding traceability, we should notice that since commit 6c9081a3
("tipc: add loopback device tracking") it is possible to follow the node
internal packet flow by just activating tcpdump on the loopback
interface. This will be true even for this mechanism; by activating
tcpdump on the involved nodes' loopback interfaces their inter-name
space messaging can easily be tracked.

v2:
- update 'net' pointer when node left/rejoined
v3:
- grab read/write lock when using node ref obj
v4:
- clone traffics between netns to loopback
Suggested-by: NJon Maloy <jon.maloy@ericsson.com>
Acked-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NHoang Le <hoang.h.le@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This commit eliminates the use of the link 'stale_limit' & 'prev_from'
(besides the already removed - 'stale_cnt') variables in the detection
of repeated retransmit failures as there is no proper way to initialize
them to avoid a false detection, i.e. it is not really a retransmission
failure but due to a garbage values in the variables.

Instead, a jiffies variable will be added to individual skbs (like the
way we restrict the skb retransmissions) in order to mark the first skb
retransmit time. Later on, at the next retransmissions, the timestamp
will be checked to see if the skb in the link transmq is "too stale",
that is, the link tolerance time has passed, so that a link reset will
be ordered. Note, just checking on the first skb in the queue is fine
enough since it must be the oldest one.
A counter is also added to keep track the actual skb retransmissions'
number for later checking when the failure happens.

The downside of this approach is that the skb->cb[] buffer is about to
be exhausted, however it is always able to allocate another memory area
and keep a reference to it when needed.

Fixes: 77cf8edb ("tipc: simplify stale link failure criteria")
Reported-by: NHoang Le <hoang.h.le@dektech.com.au>
Acked-by: NYing Xue <ying.xue@windriver.com>
Acked-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In conjunction with changing the interfaces' MTU (e.g. especially in
the case of a bonding) where the TIPC links are brought up and down
in a short time, a couple of issues were detected with the current link
changeover mechanism:

1) When one link is up but immediately forced down again, the failover
procedure will be carried out in order to failover all the messages in
the link's transmq queue onto the other working link. The link and node
state is also set to FAILINGOVER as part of the process. The message
will be transmited in form of a FAILOVER_MSG, so its size is plus of 40
bytes (= the message header size). There is no problem if the original
message size is not larger than the link's MTU - 40, and indeed this is
the max size of a normal payload messages. However, in the situation
above, because the link has just been up, the messages in the link's
transmq are almost SYNCH_MSGs which had been generated by the link
synching procedure, then their size might reach the max value already!
When the FAILOVER_MSG is built on the top of such a SYNCH_MSG, its size
will exceed the link's MTU. As a result, the messages are dropped
silently and the failover procedure will never end up, the link will
not be able to exit the FAILINGOVER state, so cannot be re-established.

2) The same scenario above can happen more easily in case the MTU of
the links is set differently or when changing. In that case, as long as
a large message in the failure link's transmq queue was built and
fragmented with its link's MTU > the other link's one, the issue will
happen (there is no need of a link synching in advance).

3) The link synching procedure also faces with the same issue but since
the link synching is only started upon receipt of a SYNCH_MSG, dropping
the message will not result in a state deadlock, but it is not expected
as design.

The 1) & 3) issues are resolved by the last commit that only a dummy
SYNCH_MSG (i.e. without data) is generated at the link synching, so the
size of a FAILOVER_MSG if any then will never exceed the link's MTU.

For the 2) issue, the only solution is trying to fragment the messages
in the failure link's transmq queue according to the working link's MTU
so they can be failovered then. A new function is made to accomplish
this, it will still be a TUNNEL PROTOCOL/FAILOVER MSG but if the
original message size is too large, it will be fragmented & reassembled
at the receiving side.
Acked-by: NYing Xue <ying.xue@windriver.com>
Acked-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This commit along with the next one are to resolve the issues with the
link changeover mechanism. See that commit for details.

Basically, for the link synching, from now on, we will send only one
single ("dummy") SYNCH message to peer. The SYNCH message does not
contain any data, just a header conveying the synch point to the peer.

A new node capability flag ("TIPC_TUNNEL_ENHANCED") is introduced for
backward compatible!
Acked-by: NYing Xue <ying.xue@windriver.com>
Acked-by: NJon Maloy <jon.maloy@ericsson.com>
Suggested-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We rename the inline function msg_get_wrapped() to the more
comprehensible msg_inner_hdr().
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

For unicast transmission, the current NACK sending althorithm is over-
active that forces the sending side to retransmit a packet that is not
really lost but just arrived at the receiving side with some delay, or
even retransmit same packets that have already been retransmitted
before. As a result, many duplicates are observed also under normal
condition, ie. without packet loss.

One example case is: node1 transmits 1 2 3 4 10 5 6 7 8 9, when node2
receives packet #10, it puts into the deferdq. When the packet #5 comes
it sends NACK with gap [6 - 9]. However, shortly after that, when
packet #6 arrives, it pulls out packet #10 from the deferfq, but it is
still out of order, so it makes another NACK with gap [7 - 9] and so on
... Finally, node1 has to retransmit the packets 5 6 7 8 9 a number of
times, but in fact all the packets are not lost at all, so duplicates!

This commit reduces duplicates by changing the condition to send NACK,
also restricting the retransmissions on individual packets via a timer
of about 1ms. However, it also needs to say that too tricky condition
for NACKs or too long timeout value for retransmissions will result in
performance reducing! The criterias in this commit are found to be
effective for both the requirements to reduce duplicates but not affect
performance.

The tipc_link_rcv() is also improved to only dequeue skb from the link
deferdq if it is expected (ie. its seqno <= rcv_nxt).
Acked-by: NYing Xue <ying.xue@windriver.com>
Acked-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

During unicast link transmission, it's observed very often that because
of one or a few lost/dis-ordered packets, the sending side will fastly
reach the send window limit and must wait for the packets to be arrived
at the receiving side or in the worst case, a retransmission must be
done first. The sending side cannot release a lot of subsequent packets
in its transmq even though all of them might have already been received
by the receiving side.
That is, one or two packets dis-ordered/lost and dozens of packets have
to wait, this obviously reduces the overall throughput!

This commit introduces an algorithm to overcome this by using "Gap ACK
blocks". Basically, a Gap ACK block will consist of <ack, gap> numbers
that describes the link deferdq where packets have been got by the
receiving side but with gaps, for example:

      link deferdq: [1 2 3 4      10 11      13 14 15       20]
--> Gap ACK blocks:       <4, 5>,   <11, 1>,      <15, 4>, <20, 0>

The Gap ACK blocks will be sent to the sending side along with the
traditional ACK or NACK message. Immediately when receiving the message
the sending side will now not only release from its transmq the packets
ack-ed by the ACK but also by the Gap ACK blocks! So, more packets can
be enqueued and transmitted.
In addition, the sending side can now do "multi-retransmissions"
according to the Gaps reported in the Gap ACK blocks.

The new algorithm as verified helps greatly improve the TIPC throughput
especially under packet loss condition.

So far, a maximum of 32 blocks is quite enough without any "Too few Gap
ACK blocks" reports with a 5.0% packet loss rate, however this number
can be increased in the furture if needed.

Also, the patch is backward compatible.
Acked-by: NYing Xue <ying.xue@windriver.com>
Acked-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, a multicast stream may start out using replicast, because
there are few destinations, and then it should ideally switch to
L2/broadcast IGMP/multicast when the number of destinations grows beyond
a certain limit. The opposite should happen when the number decreases
below the limit.

To eliminate the risk of message reordering caused by method change,
a sending socket must stick to a previously selected method until it
enters an idle period of 5 seconds. Means there is a 5 seconds pause
in the traffic from the sender socket.

If the sender never makes such a pause, the method will never change,
and transmission may become very inefficient as the cluster grows.

With this commit, we allow such a switch between replicast and
broadcast without any need for a traffic pause.

Solution is to send a dummy message with only the header, also with
the SYN bit set, via broadcast or replicast. For the data message,
the SYN bit is set and sending via replicast or broadcast (inverse
method with dummy).

Then, at receiving side any messages follow first SYN bit message
(data or dummy message), they will be held in deferred queue until
another pair (dummy or data message) arrived in other link.

v2: reverse christmas tree declaration
Acked-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NHoang Le <hoang.h.le@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When a link endpoint is re-created (e.g. after a node reboot or
interface reset), the link session number is varied by random, the peer
endpoint will be synced with this new session number before the link is
re-established.

However, there is a shortcoming in this mechanism that can lead to the
link never re-established or faced with a failure then. It happens when
the peer endpoint is ready in ESTABLISHING state, the 'peer_session' as
well as the 'in_session' flag have been set, but suddenly this link
endpoint leaves. When it comes back with a random session number, there
are two situations possible:

1/ If the random session number is larger than (or equal to) the
previous one, the peer endpoint will be updated with this new session
upon receipt of a RESET_MSG from this endpoint, and the link can be re-
established as normal. Otherwise, all the RESET_MSGs from this endpoint
will be rejected by the peer. In turn, when this link endpoint receives
one ACTIVATE_MSG from the peer, it will move to ESTABLISHED and start
to send STATE_MSGs, but again these messages will be dropped by the
peer due to wrong session.
The peer link endpoint can still become ESTABLISHED after receiving a
traffic message from this endpoint (e.g. a BCAST_PROTOCOL or
NAME_DISTRIBUTOR), but since all the STATE_MSGs are invalid, the link
will be forced down sooner or later!

Even in case the random session number is larger than the previous one,
it can be that the ACTIVATE_MSG from the peer arrives first, and this
link endpoint moves quickly to ESTABLISHED without sending out any
RESET_MSG yet. Consequently, the peer link will not be updated with the
new session number, and the same link failure scenario as above will
happen.

2/ Another situation can be that, the peer link endpoint was reset due
to any reasons in the meantime, its link state was set to RESET from
ESTABLISHING but still in session, i.e. the 'in_session' flag is not
reset...
Now, if the random session number from this endpoint is less than the
previous one, all the RESET_MSGs from this endpoint will be rejected by
the peer. In the other direction, when this link endpoint receives a
RESET_MSG from the peer, it moves to ESTABLISHING and starts to send
ACTIVATE_MSGs, but all these messages will be rejected by the peer too.
As a result, the link cannot be re-established but gets stuck with this
link endpoint in state ESTABLISHING and the peer in RESET!

Solution:

===========

This link endpoint should not go directly to ESTABLISHED when getting
ACTIVATE_MSG from the peer which may belong to the old session if the
link was re-created. To ensure the session to be correct before the
link is re-established, the peer endpoint in ESTABLISHING state will
send back the last session number in ACTIVATE_MSG for a verification at
this endpoint. Then, if needed, a new and more appropriate session
number will be regenerated to force a re-synch first.

In addition, when a link in ESTABLISHING state is reset, its state will
move to RESET according to the link FSM, along with resetting the
'in_session' flag (and the other data) as a normal link reset, it will
also be deleted if requested.

The solution is backward compatible.
Acked-by: NJon Maloy <jon.maloy@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, the broadcast retransmission algorithm is using the
'prev_retr' field in struct tipc_link to time stamp the latest broadcast
retransmission occasion. This helps to restrict retransmission of
individual broadcast packets to max once per 10 milliseconds, even
though all other criteria for retransmission are met.

We now move this time stamp to the control block of each individual
packet, and remove other limiting criteria. This simplifies the
retransmission algorithm, and eliminates any risk of logical errors
in selecting which packets can be retransmitted.
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NLUU Duc Canh <canh.d.luu@dektech.com.au>
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Default socket receive buffer size for a listener socket is 2Mb. For
each arriving empty SYN, the linux kernel allocates a 768 bytes buffer.
This means that a listener socket can serve maximum 2700 simultaneous
empty connection setup requests before it hits a receive buffer
overflow, and much fewer if the SYN is carrying any significant
amount of data.

When this happens the setup request is rejected, and the client
receives an ECONNREFUSED error.

This commit mitigates this problem by letting the client socket try to
retransmit the SYN message multiple times when it sees it rejected with
the code TIPC_ERR_OVERLOAD. Retransmission is done at random intervals
in the range of [100 ms, setup_timeout / 4], as many times as there is
room for within the setup timeout limit.
Signed-off-by: NTung Nguyen <tung.q.nguyen@dektech.com.au>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Messages intended for intitating a connection are currently
indistinguishable from regular datagram messages. The TIPC
protocol specification defines bit 17 in word 0 as a SYN bit
to allow sanity check of such messages in the listening socket,
but this has so far never been implemented.

We do that in this commit.
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When a 32-bit node address is generated from a 128-bit identifier,
there is a risk of collisions which must be discovered and handled.

We do this as follows:
- We don't apply the generated address immediately to the node, but do
  instead initiate a 1 sec trial period to allow other cluster members
  to discover and handle such collisions.

- During the trial period the node periodically sends out a new type
  of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast,
  to all the other nodes in the cluster.

- When a node is receiving such a message, it must check that the
  presented 32-bit identifier either is unused, or was used by the very
  same peer in a previous session. In both cases it accepts the request
  by not responding to it.

- If it finds that the same node has been up before using a different
  address, it responds with a DSC_TRIAL_FAIL_MSG containing that
  address.

- If it finds that the address has already been taken by some other
  node, it generates a new, unused address and returns it to the
  requester.

- During the trial period the requesting node must always be prepared
  to accept a failure message, i.e., a message where a peer suggests a
  different (or equal)  address to the one tried. In those cases it
  must apply the suggested value as trial address and restart the trial
  period.

This algorithm ensures that in the vast majority of cases a node will
have the same address before and after a reboot. If a legacy user
configures the address explicitly, there will be no trial period and
messages, so this protocol addition is completely backwards compatible.
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When sending node local messages the code is using an 'mtu' of 66060
bytes to avoid unnecessary fragmentation. During situations of low
memory tipc_msg_build() may sometimes fail to allocate such large
buffers, resulting in unnecessary send failures. This can easily be
remedied by falling back to a smaller MTU, and then reassemble the
buffer chain as if the message were arriving from a remote node.

At the same time, we change the initial MTU setting of the broadcast
link to a lower value, so that large messages always are fragmented
into smaller buffers even when we run in single node mode. Apart from
obtaining the same advantage as for the 'fallback' solution above, this
turns out to give a significant performance improvement. This can
probably be explained with the __pskb_copy() operation performed on the
buffer for each recipient during reception. We found the optimal value
for this, considering the most relevant skb pool, to be 3744 bytes.
Acked-by: NYing Xue <ying.xue@ericsson.com>
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The socket level flow control is based on the assumption that incoming
buffers meet the condition (skb->truesize / roundup(skb->len) <= 4),
where the latter value is rounded off upwards to the nearest 1k number.
This does empirically hold true for the device drivers we know, but we
cannot trust that it will always be so, e.g., in a system with jumbo
frames and very small packets.

We now introduce a check for this condition at packet arrival, and if
we find it to be false, we copy the packet to a new, smaller buffer,
where the condition will be true. We expect this to affect only a small
fraction of all incoming packets, if at all.
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, the TIPC RPS dissector is based only on the incoming packets'
source node address, hence steering all traffic from a node to the same
core. We have seen that this makes the links vulnerable to starvation
and unnecessary resets when we turn down the link tolerance to very low
values.

To reduce the risk of this happening, we exempt probe and probe replies
packets from the convergence to one core per source node. Instead, we do
the opposite, - we try to diverge those packets across as many cores as
possible, by randomizing the flow selector key.

To make such packets identifiable to the dissector, we add a new
'is_keepalive' bit to word 0 of the LINK_PROTOCOL header. This bit is
set both for PROBE and PROBE_REPLY messages, and only for those.

It should be noted that these packets are not part of any flow anyway,
and only constitute a minuscule fraction of all packets sent across a
link. Hence, there is no risk that this will affect overall performance.
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We already have point-to-multipoint flow control within a group. But
we even need the opposite; -a scheme which can handle that potentially
hundreds of sources may try to send messages to the same destination
simultaneously without causing buffer overflow at the recipient. This
commit adds such a mechanism.

The algorithm works as follows:

- When a member detects a new, joining member, it initially set its
  state to JOINED and advertises a minimum window to the new member.
  This window is chosen so that the new member can send exactly one
  maximum sized message, or several smaller ones, to the recipient
  before it must stop and wait for an additional advertisement. This
  minimum window ADV_IDLE is set to 65 1kB blocks.

- When a member receives the first data message from a JOINED member,
  it changes the state of the latter to ACTIVE, and advertises a larger
  window ADV_ACTIVE = 12 x ADV_IDLE blocks to the sender, so it can
  continue sending with minimal disturbances to the data flow.

- The active members are kept in a dedicated linked list. Each time a
  message is received from an active member, it will be moved to the
  tail of that list. This way, we keep a record of which members have
  been most (tail) and least (head) recently active.

- There is a maximum number (16) of permitted simultaneous active
  senders per receiver. When this limit is reached, the receiver will
  not advertise anything immediately to a new sender, but instead put
  it in a PENDING state, and add it to a corresponding queue. At the
  same time, it will pick the least recently active member, send it an
  advertisement RECLAIM message, and set this member to state
  RECLAIMING.

- The reclaimee member has to respond with a REMIT message, meaning that
  it goes back to a send window of ADV_IDLE, and returns its unused
  advertised blocks beyond that value to the reclaiming member.

- When the reclaiming member receives the REMIT message, it unlinks
  the reclaimee from its active list, resets its state to JOINED, and
  notes that it is now back at ADV_IDLE advertised blocks to that
  member. If there are still unread data messages sent out by
  reclaimee before the REMIT, the member goes into an intermediate
  state REMITTED, where it stays until the said messages have been
  consumed.

- The returned advertised blocks can now be re-advertised to the
  pending member, which is now set to state ACTIVE and added to
  the active member list.

- To be proactive, i.e., to minimize the risk that any member will
  end up in the pending queue, we start reclaiming resources already
  when the number of active members exceeds 3/4 of the permitted
  maximum.
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We need a mechanism guaranteeing that group unicasts sent out from a
socket are not bypassed by later sent broadcasts from the same socket.
We do this as follows:

- Each time a unicast is sent, we set a the broadcast method for the
  socket to "replicast" and "mandatory". This forces the first
  subsequent broadcast message to follow the same network and data path
  as the preceding unicast to a destination, hence preventing it from
  overtaking the latter.

- In order to make the 'same data path' statement above true, we let
  group unicasts pass through the multicast link input queue, instead
  of as previously through the unicast link input queue.

- In the first broadcast following a unicast, we set a new header flag,
  requiring all recipients to immediately acknowledge its reception.

- During the period before all the expected acknowledges are received,
  the socket refuses to accept any more broadcast attempts, i.e., by
  blocking or returning EAGAIN. This period should typically not be
  longer than a few microseconds.

- When all acknowledges have been received, the sending socket will
  open up for subsequent broadcasts, this time giving the link layer
  freedom to itself select the best transmission method.

- The forced and/or abrupt transmission method changes described above
  may lead to broadcasts arriving out of order to the recipients. We
  remedy this by introducing code that checks and if necessary
  re-orders such messages at the receiving end.
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The previously introduced message transport to all group members is
based on the tipc multicast service, but is logically a broadcast
service within the group, and that is what we call it.

We now add functionality for sending messages to all group members
having a certain identity. Correspondingly, we call this feature 'group
multicast'. The service is using unicast when only one destination is
found, otherwise it will use the bearer broadcast service to transfer
the messages. In the latter case, the receiving members filter arriving
messages by looking at the intended destination instance. If there is
no match, the message will be dropped, while still being considered
received and read as seen by the flow control mechanism.
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We now make it possible to send connectionless unicast messages
within a communication group. To send a message, the sender can use
either a direct port address, aka port identity, or an indirect port
name to be looked up.

This type of messages are subject to the same start synchronization
and flow control mechanism as group broadcast messages.
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We introduce an end-to-end flow control mechanism for group broadcast
messages. This ensures that no messages are ever lost because of
destination receive buffer overflow, with minimal impact on performance.
For now, the algorithm is based on the assumption that there is only one
active transmitter at any moment in time.
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Like with any other service, group members' availability can be
subscribed for by connecting to be topology server. However, because
the events arrive via a different socket than the member socket, there
is a real risk that membership events my arrive out of synch with the
actual JOIN/LEAVE action. I.e., it is possible to receive the first
messages from a new member before the corresponding JOIN event arrives,
just as it is possible to receive the last messages from a leaving
member after the LEAVE event has already been received.

Since each member socket is internally also subscribing for membership
events, we now fix this problem by passing those events on to the user
via the member socket. We leverage the already present member synch-
ronization protocol to guarantee correct message/event order. An event
is delivered to the user as an empty message where the two source
addresses identify the new/lost member. Furthermore, we set the MSG_OOB
bit in the message flags to mark it as an event. If the event is an
indication about a member loss we also set the MSG_EOR bit, so it can
be distinguished from a member addition event.
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

With group communication, it becomes important for a message receiver to
identify not only from which socket (identfied by a node:port tuple) the
message was sent, but also the logical identity (type:instance) of the
sending member.

We fix this by adding a second instance of struct sockaddr_tipc to the
source address area when a message is read. The extra address struct
is filled in with data found in the received message header (type,) and
in the local member representation struct (instance.)
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

As a preparation for introducing flow control for multicast and datagram
messaging we need a more strictly defined framework than we have now. A
socket must be able keep track of exactly how many and which other
sockets it is allowed to communicate with at any moment, and keep the
necessary state for those.

We therefore introduce a new concept we have named Communication Group.
Sockets can join a group via a new setsockopt() call TIPC_GROUP_JOIN.
The call takes four parameters: 'type' serves as group identifier,
'instance' serves as an logical member identifier, and 'scope' indicates
the visibility of the group (node/cluster/zone). Finally, 'flags' makes
it possible to set certain properties for the member. For now, there is
only one flag, indicating if the creator of the socket wants to receive
a copy of broadcast or multicast messages it is sending via the socket,
and if wants to be eligible as destination for its own anycasts.

A group is closed, i.e., sockets which have not joined a group will
not be able to send messages to or receive messages from members of
the group, and vice versa.

Any member of a group can send multicast ('group broadcast') messages
to all group members, optionally including itself, using the primitive
send(). The messages are received via the recvmsg() primitive. A socket
can only be member of one group at a time.
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In the following commits we will need to handle multiple incoming and
rejected/returned buffers in the function socket.c::filter_rcv().
As a preparation for this, we generalize the function by handling
buffer queues instead of individual buffers. We also introduce a
help function tipc_skb_reject(), and rename filter_rcv() to
tipc_sk_filter_rcv() in line with other functions in socket.c.
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

As preparation for introducing communication groups, we add the ability
to issue topology subscriptions and receive topology events from kernel
space. This will make it possible for group member sockets to keep track
of other group members.
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

TIPC multicast messages are currently carried over a reliable
'broadcast link', making use of the underlying media's ability to
transport packets as L2 broadcast or IP multicast to all nodes in
the cluster.

When the used bearer is lacking that ability, we can instead emulate
the broadcast service by replicating and sending the packets over as
many unicast links as needed to reach all identified destinations.
We now introduce a new TIPC link-level 'replicast' service that does
this.
Reviewed-by: NParthasarathy Bhuvaragan <parthasarathy.bhuvaragan@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Until now, we allocate memory always with GFP_ATOMIC flag.
When the system is under memory pressure and a user tries to send,
the send fails due to low memory. However, the user application
can wait for free memory if we allocate it using GFP_KERNEL flag.

In this commit, we use allocate memory with GFP_KERNEL for all user
allocation.
Reported-by: NRune Torgersen <runet@innovsys.com>
Acked-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NParthasarathy Bhuvaragan <parthasarathy.bhuvaragan@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The socket code currently handles link congestion by either blocking
and trying to send again when the congestion has abated, or just
returning to the user with -EAGAIN and let him re-try later.

This mechanism is prone to starvation, because the wakeup algorithm is
non-atomic. During the time the link issues a wakeup signal, until the
socket wakes up and re-attempts sending, other senders may have come
in between and occupied the free buffer space in the link. This in turn
may lead to a socket having to make many send attempts before it is
successful. In extremely loaded systems we have observed latency times
of several seconds before a low-priority socket is able to send out a
message.

In this commit, we simplify this mechanism and reduce the risk of the
described scenario happening. When a message is attempted sent via a
congested link, we now let it be added to the link's backlog queue
anyway, thus permitting an oversubscription of one message per source
socket. We still create a wakeup item and return an error code, hence
instructing the sender to block or stop sending. Only when enough space
has been freed up in the link's backlog queue do we issue a wakeup event
that allows the sender to continue with the next message, if any.

The fact that a socket now can consider a message sent even when the
link returns a congestion code means that the sending socket code can
be simplified. Also, since this is a good opportunity to get rid of the
obsolete 'mtu change' condition in the three socket send functions, we
now choose to refactor those functions completely.
Signed-off-by: NParthasarathy Bhuvaragan <parthasarathy.bhuvaragan@ericsson.com>
Acked-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>