TIPC based clusters are by default set up with full-mesh link
connectivity between all nodes. Those links are expected to provide
a short failure detection time, by default set to 1500 ms. Because
of this, the background load for neighbor monitoring in an N-node
cluster increases with a factor N on each node, while the overall
monitoring traffic through the network infrastructure increases at
a ~(N * (N - 1)) rate. Experience has shown that such clusters don't
scale well beyond ~100 nodes unless we significantly increase failure
discovery tolerance.
This commit introduces a framework and an algorithm that drastically
reduces this background load, while basically maintaining the original
failure detection times across the whole cluster. Using this algorithm,
background load will now grow at a rate of ~(2 * sqrt(N)) per node, and
at ~(2 * N * sqrt(N)) in traffic overhead. As an example, each node will
now have to actively monitor 38 neighbors in a 400-node cluster, instead
of as before 399.
This "Overlapping Ring Supervision Algorithm" is completely distributed
and employs no centralized or coordinated state. It goes as follows:
- Each node makes up a linearly ascending, circular list of all its N
known neighbors, based on their TIPC node identity. This algorithm
must be the same on all nodes.
- The node then selects the next M = sqrt(N) - 1 nodes downstream from
itself in the list, and chooses to actively monitor those. This is
called its "local monitoring domain".
- It creates a domain record describing the monitoring domain, and
piggy-backs this in the data area of all neighbor monitoring messages
(LINK_PROTOCOL/STATE) leaving that node. This means that all nodes in
the cluster eventually (default within 400 ms) will learn about
its monitoring domain.
- Whenever a node discovers a change in its local domain, e.g., a node
has been added or has gone down, it creates and sends out a new
version of its node record to inform all neighbors about the change.
- A node receiving a domain record from anybody outside its local domain
matches this against its own list (which may not look the same), and
chooses to not actively monitor those members of the received domain
record that are also present in its own list. Instead, it relies on
indications from the direct monitoring nodes if an indirectly
monitored node has gone up or down. If a node is indicated lost, the
receiving node temporarily activates its own direct monitoring towards
that node in order to confirm, or not, that it is actually gone.
- Since each node is actively monitoring sqrt(N) downstream neighbors,
each node is also actively monitored by the same number of upstream
neighbors. This means that all non-direct monitoring nodes normally
will receive sqrt(N) indications that a node is gone.
- A major drawback with ring monitoring is how it handles failures that
cause massive network partitionings. If both a lost node and all its
direct monitoring neighbors are inside the lost partition, the nodes in
the remaining partition will never receive indications about the loss.
To overcome this, each node also chooses to actively monitor some
nodes outside its local domain. Those nodes are called remote domain
"heads", and are selected in such a way that no node in the cluster
will be more than two direct monitoring hops away. Because of this,
each node, apart from monitoring the member of its local domain, will
also typically monitor sqrt(N) remote head nodes.
- As an optimization, local list status, domain status and domain
records are marked with a generation number. This saves senders from
unnecessarily conveying unaltered domain records, and receivers from
performing unneeded re-adaptations of their node monitoring list, such
as re-assigning domain heads.
- As a measure of caution we have added the possibility to disable the
new algorithm through configuration. We do this by keeping a threshold
value for the cluster size; a cluster that grows beyond this value
will switch from full-mesh to ring monitoring, and vice versa when
it shrinks below the value. This means that if the threshold is set to
a value larger than any anticipated cluster size (default size is 32)
the new algorithm is effectively disabled. A patch set for altering the
threshold value and for listing the table contents will follow shortly.
- This change is fully backwards compatible.
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The node keepalive interval is recalculated at each timer expiration
to catch any changes in the link tolerance, and stored in a field in
struct tipc_node. We use jiffies as unit for the stored value.
This is suboptimal, because it makes the calculation unnecessary
complex, including two unit conversions. The conversions also lead to
a rounding error that causes the link "abort limit" to be 3 in the
normal case, instead of 4, as intended. This again leads to unnecessary
link resets when the network is pushed close to its limit, e.g., in an
environment with hundreds of nodes or namesapces.
In this commit, we do instead let the keepalive value be calculated and
stored in milliseconds, so that there is only one conversion and the
rounding error is eliminated.
We also remove a redundant "keepalive" field in struct tipc_link. This
is remnant from the previous implementation.
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
commit 88e8ac7000 ("tipc: reduce transmission rate of reset messages
when link is down") revealed a flaw in the node FSM, as defined in
the log of commit 66996b6c47 ("tipc: extend node FSM").
We see the following scenario:
1: Node B receives a RESET message from node A before its link endpoint
is fully up, i.e., the node FSM is in state SELF_UP_PEER_COMING. This
event will not change the node FSM state, but the (distinct) link FSM
will move to state RESETTING.
2: As an effect of the previous event, the local endpoint on B will
declare node A lost, and post the event SELF_DOWN to the its node
FSM. This moves the FSM state to SELF_DOWN_PEER_LEAVING, meaning
that no messages will be accepted from A until it receives another
RESET message that confirms that A's endpoint has been reset. This
is wasteful, since we know this as a fact already from the first
received RESET, but worse is that the link instance's FSM has not
wasted this information, but instead moved on to state ESTABLISHING,
meaning that it repeatedly sends out ACTIVATE messages to the reset
peer A.
3: Node A will receive one of the ACTIVATE messages, move its link FSM
to state ESTABLISHED, and start repeatedly sending out STATE messages
to node B.
4: Node B will consistently drop these messages, since it can only accept
accept a RESET according to its node FSM.
5: After four lost STATE messages node A will reset its link and start
repeatedly sending out RESET messages to B.
6: Because of the reduced send rate for RESET messages, it is very
likely that A will receive an ACTIVATE (which is sent out at a much
higher frequency) before it gets the chance to send a RESET, and A
may hence quickly move back to state ESTABLISHED and continue sending
out STATE messages, which will again be dropped by B.
7: GOTO 5.
8: After having repeated the cycle 5-7 a number of times, node A will
by chance get in between with sending a RESET, and the situation is
resolved.
Unfortunately, we have seen that it may take a substantial amount of
time before this vicious loop is broken, sometimes in the order of
minutes.
We correct this by making a small correction to the node FSM: When a
node in state SELF_UP_PEER_COMING receives a SELF_DOWN event, it now
moves directly back to state SELF_DOWN_PEER_DOWN, instead of as now
SELF_DOWN_PEER_LEAVING. This is logically consistent, since we don't
need to wait for RESET confirmation from of an endpoint that we alread
know has been reset. It also means that node B in the scenario above
will not be dropping incoming STATE messages, and the link can come up
immediately.
Finally, a symmetry comparison reveals that the FSM has a similar
error when receiving the event PEER_DOWN in state PEER_UP_SELF_COMING.
Instead of moving to PERR_DOWN_SELF_LEAVING, it should move directly
to SELF_DOWN_PEER_DOWN. Although we have never seen any negative effect
of this logical error, we choose fix this one, too.
The node FSM looks as follows after those changes:
+----------------------------------------+
| PEER_DOWN_EVT|
| |
+------------------------+----------------+ |
|SELF_DOWN_EVT | | |
| | | |
| +-----------+ +-----------+ |
| |NODE_ | |NODE_ | |
| +----------|FAILINGOVER|<---------|SYNCHING |-----------+ |
| |SELF_ +-----------+ FAILOVER_+-----------+ PEER_ | |
| |DOWN_EVT | A BEGIN_EVT A | DOWN_EVT| |
| | | | | | | |
| | | | | | | |
| | |FAILOVER_ |FAILOVER_ |SYNCH_ |SYNCH_ | |
| | |END_EVT |BEGIN_EVT |BEGIN_EVT|END_EVT | |
| | | | | | | |
| | | | | | | |
| | | +--------------+ | | |
| | +-------->| SELF_UP_ |<-------+ | |
| | +-----------------| PEER_UP |----------------+ | |
| | |SELF_DOWN_EVT +--------------+ PEER_DOWN_EVT| | |
| | | A A | | |
| | | | | | | |
| | | PEER_UP_EVT| |SELF_UP_EVT | | |
| | | | | | | |
V V V | | V V V
+------------+ +-----------+ +-----------+ +------------+
|SELF_DOWN_ | |SELF_UP_ | |PEER_UP_ | |PEER_DOWN |
|PEER_LEAVING| |PEER_COMING| |SELF_COMING| |SELF_LEAVING|
+------------+ +-----------+ +-----------+ +------------+
| | A A | |
| | | | | |
| SELF_ | |SELF_ |PEER_ |PEER_ |
| DOWN_EVT| |UP_EVT |UP_EVT |DOWN_EVT |
| | | | | |
| | | | | |
| | +--------------+ | |
|PEER_DOWN_EVT +--->| SELF_DOWN_ |<---+ SELF_DOWN_EVT|
+------------------->| PEER_DOWN |<--------------------+
+--------------+
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When an ACTIVATE or data packet is received in a link in state
ESTABLISHING, the link does not immediately change state to
ESTABLISHED, but does instead return a LINK_UP event to the caller,
which will execute the state change in a different lock context.
This non-atomic approach incurs a low risk that we may have two
LINK_UP events pending simultaneously for the same link, resulting
in the final part of the setup procedure being executed twice. The
only potential harm caused by this it that we may see two LINK_UP
events issued to subsribers of the topology server, something that
may cause confusion.
This commit eliminates this risk by checking if the link is already
up before proceeding with the second half of the setup.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
net/ipv4/ip_gre.c
Minor conflicts between tunnel bug fixes in net and
ipv6 tunnel cleanups in net-next.
Signed-off-by: David S. Miller <davem@davemloft.net>
During neighbor discovery, nodes advertise their capabilities as a bit
map in a dedicated 16-bit field in the discovery message header. This
bit map has so far only be stored in the node structure on the peer
nodes, but we now see the need to keep a copy even in the socket
structure.
This commit adds this functionality.
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We have observed complete lock up of broadcast-link transmission due to
unacknowledged packets never being removed from the 'transmq' queue. This
is traced to nodes having their ack field set beyond the sequence number
of packets that have actually been transmitted to them.
Consider an example where node 1 has sent 10 packets to node 2 on a
link and node 3 has sent 20 packets to node 2 on another link. We
see examples of an ack from node 2 destined for node 3 being treated as
an ack from node 2 at node 1. This leads to the ack on the node 1 to node
2 link being increased to 20 even though we have only sent 10 packets.
When node 1 does get around to sending further packets, none of the
packets with sequence numbers less than 21 are actually removed from the
transmq.
To resolve this we reinstate some code lost in commit d999297c3d ("tipc:
reduce locking scope during packet reception") which ensures that only
messages destined for the receiving node are processed by that node. This
prevents the sequence numbers from getting out of sync and resolves the
packet leakage, thereby resolving the broadcast-link transmission
lock-ups we observed.
While we are aware that this change only patches over a root problem that
we still haven't identified, this is a sanity test that it is always
legitimate to do. It will remain in the code even after we identify and
fix the real problem.
Reviewed-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Reviewed-by: John Thompson <john.thompson@alliedtelesis.co.nz>
Signed-off-by: Hamish Martin <hamish.martin@alliedtelesis.co.nz>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When we are displaying statistics for the first link established between
two peers, it will always be presented as STANDBY although it in reality
is ACTIVE.
This happens because we forget to set the 'active' flag in the link
instance at the moment it is established. Although this is a bug, it only
has impact on the presentation view of the link, not on its actual
functionality.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
According to the link FSM, a received traffic packet can take a link
from state ESTABLISHING to ESTABLISHED, but the link can still not be
fully set up in one atomic operation. This means that even if the the
very first packet on the link is a traffic packet with sequence number
1 (one), it has to be dropped and retransmitted.
This can be avoided if we let the mentioned packet be preceded by a
LINK_PROTOCOL/STATE message, which takes up the endpoint before the
arrival of the traffic.
We add this small feature in this commit.
This is a fully compatible change.
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In some link establishment scenarios we see that packet #2 may be sent
out before packet #1, forcing the receiver to demand retransmission of
the missing packet. This is harmless, but may cause confusion among
people tracing the packet flow.
Since this is extremely easy to fix, we do so by adding en extra send
call to the bearer immediately after the link has come up.
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Make the c files less cluttered and enable netlink attributes to be
shared between files.
Signed-off-by: Richard Alpe <richard.alpe@ericsson.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Parthasarathy Bhuvaragan <parthasarathy.bhuvaragan@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Make sure we have a link before checking if it has been reset or not.
Prior to this patch tipc_link_is_reset() could be called with a non
existing link, resulting in a null pointer dereference.
Signed-off-by: Richard Alpe <richard.alpe@ericsson.com>
Acked-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When the TIPC module is unloaded, we have identified a race condition
that allows a node reference counter to go to zero and the node instance
being freed before the node timer is finished with accessing it. This
leads to occasional crashes, especially in multi-namespace environments.
The scenario goes as follows:
CPU0:(node_stop) CPU1:(node_timeout) // ref == 2
1: if(!mod_timer())
2: if (del_timer())
3: tipc_node_put() // ref -> 1
4: tipc_node_put() // ref -> 0
5: kfree_rcu(node);
6: tipc_node_get(node)
7: // BOOM!
We now clean up this functionality as follows:
1) We remove the node pointer from the node lookup table before we
attempt deactivating the timer. This way, we reduce the risk that
tipc_node_find() may obtain a valid pointer to an instance marked
for deletion; a harmless but undesirable situation.
2) We use del_timer_sync() instead of del_timer() to safely deactivate
the node timer without any risk that it might be reactivated by the
timeout handler. There is no risk of deadlock here, since the two
functions never touch the same spinlocks.
3: We remove a pointless tipc_node_get() + tipc_node_put() from the
timeout handler.
Reported-by: Zhijiang Hu <huzhijiang@gmail.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Although we have never seen it happen, we have identified the
following problematic scenario when nodes are stopped and deleted:
CPU0: CPU1:
tipc_node_xxx() //ref == 1
tipc_node_put() //ref -> 0
tipc_node_find() // node still in table
tipc_node_delete()
list_del_rcu(n. list)
tipc_node_get() //ref -> 1, bad
kfree_rcu()
tipc_node_put() //ref to 0 again.
kfree_rcu() // BOOM!
We fix this by introducing use of the conditional kref_get_if_not_zero()
instead of kref_get() in the function tipc_node_find(). This eliminates
any risk of post-mortem access.
Reported-by: Zhijiang Hu <huzhijiang@gmail.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/phy/bcm7xxx.c
drivers/net/phy/marvell.c
drivers/net/vxlan.c
All three conflicts were cases of simple overlapping changes.
Signed-off-by: David S. Miller <davem@davemloft.net>
Refactor tipc_node_xmit() to fail fast and fail early. Fix several
potential memory leaks in unexpected error paths.
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: Richard Alpe <richard.alpe@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In commit 5266698661 ("tipc: let broadcast packet reception
use new link receive function") we introduced a new per-node
broadcast reception link instance. This link is created at the
moment the node itself is created. Unfortunately, the allocation
is done after the node instance has already been added to the node
lookup hash table. This creates a potential race condition, where
arriving broadcast packets are able to find and access the node
before it has been fully initialized, and before the above mentioned
link has been created. The result is occasional crashes in the function
tipc_bcast_rcv(), which is trying to access the not-yet existing link.
We fix this by deferring the addition of the node instance until after
it has been fully initialized in the function tipc_node_create().
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Changing certain link attributes (link tolerance and link priority)
from the TIPC management tool is supposed to automatically take
effect at both endpoints of the affected link.
Currently the media address is not instantiated for the link and is
used uninstantiated when crafting protocol messages designated for the
peer endpoint. This means that changing a link property currently
results in the property being changed on the local machine but the
protocol message designated for the peer gets lost. Resulting in
property discrepancy between the endpoints.
In this patch we resolve this by using the media address from the
link entry and using the bearer transmit function to send it. Hence,
we can now eliminate the redundant function tipc_link_prot_xmit() and
the redundant field tipc_link::media_addr.
Fixes: 2af5ae372a (tipc: clean up unused code and structures)
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Reported-by: Jason Hu <huzhijiang@gmail.com>
Signed-off-by: Richard Alpe <richard.alpe@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 5405ff6e15 ("tipc: convert node lock to rwlock")
introduced a bug to the node reference counter handling. When a
message is successfully sent in the function tipc_node_xmit(),
we return directly after releasing the node lock, instead of
continuing and decrementing the node reference counter as we
should do.
This commit fixes this bug.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The number of variables with Hungarian notation (l_ptr, n_ptr etc.)
has been significantly reduced over the last couple of years.
We now root out the last traces of this practice.
There are no functional changes in this commit.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We move the definition of struct tipc_link from link.h to link.c in
order to minimize its exposure to the rest of the code.
When needed, we define new functions to make it possible for external
entities to access and set data in the link.
Apart from the above, there are no functional changes.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In our effort to have less code and include dependencies between
entities such as node, link and bearer, we try to narrow down
the exposed interface towards the node as much as possible.
In this commit, we move the definition of struct tipc_node, along
with many of its associated function declarations, from node.h to
node.c. We also move some function definitions from link.c and
name_distr.c to node.c, since they access fields in struct tipc_node
that should not be externally visible. The moved functions are renamed
according to new location, and made static whenever possible.
There are no functional changes in this commit.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
According to the node FSM a node in state SELF_UP_PEER_UP cannot
change state inside a lock context, except when a TUNNEL_PROTOCOL
(SYNCH or FAILOVER) packet arrives. However, the node's individual
links may still change state.
Since each link now is protected by its own spinlock, we finally have
the conditions in place to convert the node spinlock to an rwlock_t.
If the node state and arriving packet type are rigth, we can let the
link directly receive the packet under protection of its own spinlock
and the node lock in read mode. In all other cases we use the node
lock in write mode. This enables full concurrent execution between
parallel links during steady-state traffic situations, i.e., 99+ %
of the time.
This commit implements this change.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As a preparation to allow parallel links to work more independently
from each other we introduce a per-link spinlock, to be stored in the
struct nodes's link entry area. Since the node lock still is a regular
spinlock there is no increase in parallellism at this stage.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The file name_distr.c currently contains three functions,
named_cluster_distribute(), tipc_publ_subcscribe() and
tipc_publ_unsubscribe() that all directly access fields in
struct tipc_node. We want to eliminate such dependencies, so
we move those functions to the file node.c and rename them to
tipc_node_broadcast(), tipc_node_subscribe() and tipc_node_unsubscribe()
respectively.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The function tipc_node_check_state() contains the core logics
for handling link synchronization and failover. For this reason,
it is important to keep it as comprehensible as possible.
In this commit, we make three small cleanups.
1) If the node is in state SELF_DOWN_PEER_LEAVING and the received
packet confirms that the peer has lost contact, there will be no
further action in this function. To make this clearer, we return
from the function directly after the state change.
2) Since commit 0f8b8e28fb ("tipc: eliminate risk of stalled
link synchronization") only the logically first TUNNEL_PROTO/SYNCH
packet can alter the link state and set the synch point,
independently of arrival order. Hence, there is not any longer any
need to adjust the synch value in case such packets arrive in
disorder. We remove this adjustment.
3) It is the intention that any message arriving on any of the links
may trig a check for and possible termination of a node SYNCH state.
A redundant and unnoticed check for tipc_link_is_synching() obviously
beats this purpose, with the effect that only packets arriving on the
synching link may currently end the synch state. We remove this check.
This change will further shorten the synchronization period between
parallel links.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
TO: "David S. Miller" <davem@davemloft.net>
CC: netdev@vger.kernel.org
CC: Jon Maloy <jon.maloy@ericsson.com>
CC: Ying Xue <ying.xue@windriver.com>
CC: tipc-discussion@lists.sourceforge.net
CC: linux-kernel@vger.kernel.org
Signed-off-by: Fengguang Wu <fengguang.wu@intel.com>
Acked-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
After the previous changes in this series, we can now remove some
unused code and structures, both in the broadcast, link aggregation
and link code.
There are no functional changes in this commit.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Correct synchronization of the broadcast link at first contact between
two nodes is dependent on the assumption that the binding table "bulk"
update passes via the same link as the initial broadcast syncronization
message, i.e., via the first link that is established.
This is not guaranteed in the current implementation. If two link
come up very close to each other in time, the "bulk" may quite well
pass via the second link, and hence void the guarantee of a correct
initial synchronization before the broadcast link is opened.
This commit makes two small changes to strengthen this guarantee.
1) We let the second established link occupy slot 1 of the
"active_links" array, while the first link will retain slot 0.
(This is in reality a cosmetic change, we could just as well keep
the current, opposite order)
2) We let the name distributor always use link selector/slot 0 when
it sends it binding table updates.
The extra traffic bias on the first link caused by this change should
be negligible, since binding table updates constitutes a very small
fraction of the total traffic.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With the recent commit series, we have established a one-way dependency
between the link aggregation (struct tipc_node) instances and their
pertaining tipc_link instances. This has enabled quite significant code
and structure simplifications.
In this commit, we eliminate the field 'owner', which points to an
instance of struct tipc_node, from struct tipc_link, and replace it with
a pointer to struct net, which is the only external reference now needed
by a link instance.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Until now, we have been keeping track of the exact set of broadcast
destinations though the help structure tipc_node_map. This leads us to
have to maintain a whole infrastructure for supporting this, including
a pseudo-bearer and a number of functions to manipulate both the bearers
and the node map correctly. Apart from the complexity, this approach is
also limiting, as struct tipc_node_map only can support cluster local
broadcast if we want to avoid it becoming excessively large. We want to
eliminate this limitation, in order to enable introduction of scoped
multicast in the future.
A closer analysis reveals that it is unnecessary maintaining this "full
set" overview; it is sufficient to keep a counter per bearer, indicating
how many nodes can be reached via this bearer at the moment. The protocol
is now robust enough to handle transitional discrepancies between the
nominal number of reachable destinations, as expected by the broadcast
protocol itself, and the number which is actually reachable at the
moment. The initial broadcast synchronization, in conjunction with the
retransmission mechanism, ensures that all packets will eventually be
acknowledged by the correct set of destinations.
This commit introduces these changes.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The code path for receiving broadcast packets is currently distinct
from the unicast path. This leads to unnecessary code and data
duplication, something that can be avoided with some effort.
We now introduce separate per-peer tipc_link instances for handling
broadcast packet reception. Each receive link keeps a pointer to the
common, single, broadcast link instance, and can hence handle release
and retransmission of send buffers as if they belonged to the own
instance.
Furthermore, we let each unicast link instance keep a reference to both
the pertaining broadcast receive link, and to the common send link.
This makes it possible for the unicast links to easily access data for
broadcast link synchronization, as well as for carrying acknowledges for
received broadcast packets.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Until now, we have tried to support both the newer, dedicated broadcast
synchronization mechanism along with the older, less safe, RESET_MSG/
ACTIVATE_MSG based one. The latter method has turned out to be a hazard
in a highly dynamic cluster, so we find it safer to disable it completely
when we find that the former mechanism is supported by the peer node.
For this purpose, we now introduce a new capabability bit,
TIPC_BCAST_SYNCH, to inform any peer nodes that dedicated broadcast
syncronization is supported by the present node. The new bit is conveyed
between peers in the 'capabilities' field of neighbor discovery messages.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In reality, the link implementation is already independent from
struct tipc_bearer, in that it doesn't store any reference to it.
However, we still pass on a pointer to a bearer instance in the
function tipc_link_create(), just to have it extract some
initialization information from it.
I later commits, we need to create instances of tipc_link without
having any associated struct tipc_bearer. To facilitate this, we
want to extract the initialization data already in the creator
function in node.c, before calling tipc_link_create(), and pass
this info on as individual parameters in the call.
This commit introduces this change.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/usb/asix_common.c
net/ipv4/inet_connection_sock.c
net/switchdev/switchdev.c
In the inet_connection_sock.c case the request socket hashing scheme
is completely different in net-next.
The other two conflicts were overlapping changes.
Signed-off-by: David S. Miller <davem@davemloft.net>
The change made in the previous commit revealed a small flaw in the way
the node FSM is updated. When the function tipc_node_link_down() is
called for the last link to a node, we should check whether this was
caused by a local reset or by a received RESET message from the peer.
In the latter case, we can directly issue a PEER_LOST_CONTACT_EVT to
the node FSM, so that it is ready to re-establish contact. If this is
not done, the peer node will sometimes have to go through a second
establish cycle before the link becomes stable.
We fix this in this commit by conditionally issuing the mentioned
event in the function tipc_node_link_down(). We also move LINK_RESET
FSM even away from the link_reset() function and into the caller
function, partially because it is easier to follow the code when state
changes are gathered at a limited number of locations, partially
because there will be cases in future commits where we don't want the
link to go RESET mode when link_reset() is called.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a link is taken down because of a node local event, such as
disabling of a bearer or an interface, we currently leave it to the
peer node to discover the broken communication. The default time for
such failure discovery is 1.5-2 seconds.
If we instead allow the terminating link endpoint to send out a RESET
message at the moment it is reset, we can achieve the impression that
both endpoints are going down instantly. Since this is a very common
scenario, we find it worthwhile to make this small modification.
Apart from letting the link produce the said message, we also have to
ensure that the interface is able to transmit it before TIPC is
detached. We do this by performing the disabling of a bearer in three
steps:
1) Disable reception of TIPC packets from the interface in question.
2) Take down the links, while allowing them so send out a RESET message.
3) Disable transmission of TIPC packets on the interface.
Apart from this, we now have to react on the NETDEV_GOING_DOWN event,
instead of as currently the NEDEV_DOWN event, to ensure that such
transmission is possible during the teardown phase.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Link establishing, just like link teardown, is a non-atomic action, in
the sense that discovering that conditions are right to establish a link,
and the actual adding of the link to one of the node's send slots is done
in two different lock contexts. The link FSM is designed to help bridging
the gap between the two contexts in a safe manner.
We have now discovered a weakness in the implementaton of this FSM.
Because we directly let the link go from state LINK_ESTABLISHING to
state LINK_ESTABLISHED already in the first lock context, we are unable
to distinguish between a fully established link, i.e., a link that has
been added to its slot, and a link that has not yet reached the second
lock context. It may hence happen that a manual intervention, e.g., when
disabling an interface, causes the function tipc_node_link_down() to try
removing the link from the node slots, decrementing its active link
counter etc, although the link was never added there in the first place.
We solve this by delaying the actual state change until we reach the
second lock context, inside the function tipc_node_link_up(). This
makes it possible for potentail callers of __tipc_node_link_down() to
know if they should proceed or not, and the problem is solved.
Unforunately, the situation described above also has a second problem.
Since there by necessity is a tipc_node_link_up() call pending once
the node lock has been released, we must defuse that call by setting
the link back from LINK_ESTABLISHING to LINK_RESET state. This forces
us to make a slight modification to the link FSM, which will now look
as follows.
+------------------------------------+
|RESET_EVT |
| |
| +--------------+
| +-----------------| SYNCHING |-----------------+
| |FAILURE_EVT +--------------+ PEER_RESET_EVT|
| | A | |
| | | | |
| | | | |
| | |SYNCH_ |SYNCH_ |
| | |BEGIN_EVT |END_EVT |
| | | | |
| V | V V
| +-------------+ +--------------+ +------------+
| | RESETTING |<---------| ESTABLISHED |--------->| PEER_RESET |
| +-------------+ FAILURE_ +--------------+ PEER_ +------------+
| | EVT | A RESET_EVT |
| | | | |
| | +----------------+ | |
| RESET_EVT| |RESET_EVT | |
| | | | |
| | | |ESTABLISH_EVT |
| | | +-------------+ | |
| | | | RESET_EVT | | |
| | | | | | |
| V V V | | |
| +-------------+ +--------------+ RESET_EVT|
+--->| RESET |--------->| ESTABLISHING |<----------------+
+-------------+ PEER_ +--------------+
| A RESET_EVT |
| | |
| | |
|FAILOVER_ |FAILOVER_ |FAILOVER_
|BEGIN_EVT |END_EVT |BEGIN_EVT
| | |
V | |
+-------------+ |
| FAILINGOVER |<----------------+
+-------------+
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In commit 6e498158a8 ("tipc: move link synch and failover to link aggregation level")
we introduced a new mechanism for performing link failover and
synchronization. We have now detected a bug in this mechanism.
During link synchronization we use the arrival of any packet on
the tunnel link to trig a check for whether it has reached the
synchronization point or not. This has turned out to be too
permissive, since it may cause an arriving non-last SYNCH packet to
end the synch state, just to see the next SYNCH packet initiate a
new synch state with a new, higher synch point. This is not fatal,
but should be avoided, because it may significantly extend the
synchronization period, while at the same time we are not allowed
to send NACKs if packets are lost. In the worst case, a low-traffic
user may see its traffic stall until a LINK_PROTOCOL state message
trigs the link to leave synchronization state.
At the same time, LINK_PROTOCOL packets which happen to have a (non-
valid) sequence number lower than the tunnel link's rcv_nxt value will
be consistently dropped, and will never be able to resolve the situation
described above.
We fix this by exempting LINK_PROTOCOL packets from the sequence number
check, as they should be. We also reduce (but don't completely
eliminate) the risk of entering multiple synchronization states by only
allowing the (logically) first SYNCH packet to initiate a synchronization
state. This works independently of actual packet arrival order.
Fixes: commit 6e498158a8 ("tipc: move link synch and failover to link aggregation level")
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Recent changes to the link synchronization means that we can now just
drop packets arriving on the synchronizing link before the synch point
is reached. This has lead to significant simplifications to the
implementation, but also turns out to have a flip side that we need
to consider.
Under unlucky circumstances, the two endpoints may end up
repeatedly dropping each other's packets, while immediately
asking for retransmission of the same packets, just to drop
them once more. This pattern will eventually be broken when
the synch point is reached on the other link, but before that,
the endpoints may have arrived at the retransmission limit
(stale counter) that indicates that the link should be broken.
We see this happen at rare occasions.
The fix for this is to not ask for retransmissions when a link is in
state LINK_SYNCHING. The fact that the link has reached this state
means that it has already received the first SYNCH packet, and that it
knows the synch point. Hence, it doesn't need any more packets until the
other link has reached the synch point, whereafter it can go ahead and
ask for the missing packets.
However, because of the reduced traffic on the synching link that
follows this change, it may now take longer to discover that the
synch point has been reached. We compensate for this by letting all
packets, on any of the links, trig a check for synchronization
termination. This is possible because the packets themselves don't
contain any information that is needed for discovering this condition.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When we introduced the new link failover/synch mechanism
in commit 6e498158a8
("tipc: move link synch and failover to link aggregation level"),
we missed the case when the non-tunnel link goes down during the link
synchronization period. In this case the tunnel link will remain in
state LINK_SYNCHING, something leading to unpredictable behavior when
the failover procedure is initiated.
In this commit, we ensure that the node and remaining link goes
back to regular communication state (SELF_UP_PEER_UP/LINK_ESTABLISHED)
when one of the parallel links goes down. We also ensure that we don't
re-enter synch mode if subsequent SYNCH packets arrive on the remaining
link.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a link goes down, and there is still a working link towards its
destination node, a failover is initiated, and the failed link is not
allowed to re-establish until that procedure is finished. To ensure
this, the concerned link endpoints are set to state LINK_FAILINGOVER,
and the node endpoints to NODE_FAILINGOVER during the failover period.
However, if the link reset is due to a disabled bearer, the corres-
ponding link endpoint is deleted, and only the node endpoint knows
about the ongoing failover. Now, if the disabled bearer is re-enabled
during the failover period, the discovery mechanism may create a new
link endpoint that is ready to be established, despite that this is not
permitted. This situation may cause both the ongoing failover and any
subsequent link synchronization to fail.
In this commit, we ensure that a newly created link goes directly to
state LINK_FAILINGOVER if the corresponding node state is
NODE_FAILINGOVER. This eliminates the problem described above.
Furthermore, we tighten the criteria for which packets are allowed
to end a failover state in the function tipc_node_check_state().
By checking that the receiving link is up and running, instead of just
checking that it is not in failover mode, we eliminate the risk that
protocol packets from the re-created link may cause the failover to
be prematurely terminated.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We simplify the link creation function tipc_link_create() and the way
the link struct it is connected to the node struct. In particular, we
remove the duplicate initialization of some fields which are anyway set
in tipc_link_reset().
Tested-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
After the most recent changes, all access calls to a link which
may entail addition of messages to the link's input queue are
postpended by an explicit call to tipc_sk_rcv(), using a reference
to the correct queue.
This means that the potentially hazardous implicit delivery, using
tipc_node_unlock() in combination with a binary flag and a cached
queue pointer, now has become redundant.
This commit removes this implicit delivery mechanism both for regular
data messages and for binding table update messages.
Tested-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to facilitate future improvements to the locking structure, we
want to make resetting and establishing of links non-atomic. I.e., the
functions tipc_node_link_up() and tipc_node_link_down() should be called
from outside the node lock context, and grab/release the node lock
themselves. This requires that we can freeze the link state from the
moment it is set to RESETTING or PEER_RESET in one lock context until
it is set to RESET or ESTABLISHING in a later context. The recently
introduced link FSM makes this possible, so we are now ready to introduce
the above change.
This commit implements this.
Tested-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The node lock is currently grabbed and and released in the function
tipc_disc_rcv() in the file discover.c. As a preparation for the next
commits, we need to move this node lock handling, along with the code
area it is covering, to node.c.
This commit introduces this change.
Tested-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Until now, we have been handling link failover and synchronization
by using an additional link state variable, "exec_mode". This variable
is not independent of the link FSM state, something causing a risk of
inconsistencies, apart from the fact that it clutters the code.
The conditions are now in place to define a new link FSM that covers
all existing use cases, including failover and synchronization, and
eliminate the "exec_mode" field altogether. The FSM must also support
non-atomic resetting of links, which will be introduced later.
The new link FSM is shown below, with 7 states and 8 events.
Only events leading to state change are shown as edges.
+------------------------------------+
|RESET_EVT |
| |
| +--------------+
| +-----------------| SYNCHING |-----------------+
| |FAILURE_EVT +--------------+ PEER_RESET_EVT|
| | A | |
| | | | |
| | | | |
| | |SYNCH_ |SYNCH_ |
| | |BEGIN_EVT |END_EVT |
| | | | |
| V | V V
| +-------------+ +--------------+ +------------+
| | RESETTING |<---------| ESTABLISHED |--------->| PEER_RESET |
| +-------------+ FAILURE_ +--------------+ PEER_ +------------+
| | EVT | A RESET_EVT |
| | | | |
| | | | |
| | +--------------+ | |
| RESET_EVT| |RESET_EVT |ESTABLISH_EVT |
| | | | |
| | | | |
| V V | |
| +-------------+ +--------------+ RESET_EVT|
+--->| RESET |--------->| ESTABLISHING |<----------------+
+-------------+ PEER_ +--------------+
| A RESET_EVT |
| | |
| | |
|FAILOVER_ |FAILOVER_ |FAILOVER_
|BEGIN_EVT |END_EVT |BEGIN_EVT
| | |
V | |
+-------------+ |
| FAILINGOVER |<----------------+
+-------------+
These changes are fully backwards compatible.
Tested-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The implementation of the link FSM currently takes decisions about and
sends out link protocol messages. This is unnecessary, since such
actions are not the result of any link state change, and are even
decided based on non-FSM state information ("silent_intv_cnt").
We now move the sending of unicast link protocol messages to the
function tipc_link_timeout(), and the initial broadcast synchronization
message to tipc_node_link_up(). The latter is done because a link
instance should not need to know whether it is the first or second
link to a destination. Such information is now restricted to and
handled by the link aggregation layer in node.c
Tested-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Link failover and synchronization have until now been handled by the
links themselves, forcing them to have knowledge about and to access
parallel links in order to make the two algorithms work correctly.
In this commit, we move the control part of this functionality to the
link aggregation level in node.c, which is the right location for this.
As a result, the two algorithms become easier to follow, and the link
implementation becomes simpler.
Tested-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jon Paul Maloy
David S. Miller
Hamish Martin
Richard Alpe
Wu Fengguang