| Age | Commit message (Collapse) | Author |
|---|
|
Until now, we have kept a pre-allocated protocol message header
aggregated into struct tipc_link. Apart from adding unnecessary
footprint to the link instances, this requires extra code both to
initialize and re-initialize it.
We now remove this sub-optimization. This change also makes it
possible to clean up the function tipc_build_proto_msg() and remove
a couple of small functions that were accessing the mentioned header.
In particular, we can replace all occurrences of the local function
call link_own_addr(link) with the generic tipc_own_addr(net).
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 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>
|
|
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>
|
|
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 only been supporting a fix MTU size of 1500 bytes
for all broadcast media, irrespective of their actual capability.
We now make the broadcast MTU adaptable to the carrying media, i.e.,
we use the smallest MTU supported by any of the interfaces attached
to TIPC.
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>
|
|
This commit simplifies the broadcast link transmission function, by
leveraging previous changes to the link transmission function and the
broadcast transmission link life cycle.
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 broadcast link instance (struct tipc_link) used for sending is
currently aggregated into struct tipc_bclink. This means that we cannot
use the regular tipc_link_create() function for initiating the link, but
do instead have to initiate numerous fields directly from the
bcast_init() function.
We want to reduce dependencies between the broadcast functionality
and the inner workings of tipc_link. In this commit, we introduce
a new function tipc_bclink_create() to link.c, and allocate the
instance of the link separately using this function.
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 broadcast transmission link is currently instantiated when the
network subsystem is started, i.e., on order from user space via netlink.
This forces the broadcast transmission code to do unnecessary tests for
the existence of the transmission link, as well in single mode node as
in network mode.
In this commit, we do instead create the link during initialization of
the name space, and remove it when it is stopped. The fact that the
transmission link now has a guaranteed longer life cycle than any of its
potential clients paves the way for further code simplifcations
and optimizations.
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 broadcast lock will need to be acquired outside bcast.c in a later
commit. For this reason, we move the lock to struct tipc_net. Consistent
with the changes in the previous commit, we also introducee two new
functions tipc_bcast_lock() and tipc_bcast_unlock(). The code that is
currently using tipc_bclink_lock()/unlock() will be phased out during
the coming commits in this series.
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>
|
|
Currently, a number of structure and function definitions related
to the broadcast functionality are unnecessarily exposed in the file
bcast.h. This obscures the fact that the external interface towards
the broadcast link in fact is very narrow, and causes unnecessary
recompilations of other files when anything changes in those
definitions.
In this commit, we move as many of those definitions as is currently
possible to the file bcast.c.
We also rename the structure 'tipc_bclink' to 'tipc_bc_base', both
since the name does not correctly describe the contents of this
struct, and will do so even less in the future, and because we want
to use the term 'link' more appropriately in the functionality
introduced later in this series.
Finally, we rename a couple of functions, such as tipc_bclink_xmit()
and others that will be kept in the future, to include the term 'bcast'
instead.
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>
|
|
The default fix broadcast window size is currently set to 20 packets.
This is a very low value, set at a time when we were still testing on
10 Mb/s hubs, and a change to it is long overdue.
Commit 7845989cb4b3da1db ("net: tipc: fix stall during bclink wakeup procedure")
revealed a problem with this low value. For messages of importance LOW,
the backlog queue limit will be calculated to 30 packets, while a
single, maximum sized message of 66000 bytes, carried across a 1500 MTU
network consists of 46 packets.
This leads to the following scenario (among others leading to the same
situation):
1: Msg 1 of 46 packets is sent. 20 packets go to the transmit queue, 26
packets to the backlog queue.
2: Msg 2 of 46 packets is attempted sent, but rejected because there is
no more space in the backlog queue at this level. The sender is added
to the wakeup queue with a "pending packets chain size" number of 46.
3: Some packets in the transmit queue are acked and released. We try to
wake up the sender, but the pending size of 46 is bigger than the LOW
wakeup limit of 30, so this doesn't happen.
5: Subsequent acks releases all the remaining buffers. Each time we test
for the wakeup criteria and find that 46 still is larger than 30,
even after both the transmit and the backlog queues are empty.
6: The sender is never woken up and given a chance to send its message.
He is stuck.
We could now loosen the wakeup criteria (used by link_prepare_wakeup())
to become equal to the send criteria (used by tipc_link_xmit()), i.e.,
by ignoring the "pending packets chain size" value altogether, or we can
just increase the queue limits so that the criteria can be satisfied
anyway. There are good reasons (potentially multiple waiting senders) to
not opt for the former solution, so we choose the latter one.
This commit fixes the problem by giving the broadcast link window a
default value of 50 packets. We also introduce a new minimum link
window size BCLINK_MIN_WIN of 32, which is enough to always avoid the
described situation. Finally, in order to not break any existing users
which may set the window explicitly, we enforce that the window is set
to the new minimum value in case the user is trying to set it to
anything lower.
Fixes: 7845989cb4b3da1db ("net: tipc: fix stall during bclink wakeup procedure")
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>
|
|
If an attempt to wake up users of broadcast link is made when there is
no enough place in send queue than it may hang up inside the
tipc_sk_rcv() function since the loop breaks only after the wake up
queue becomes empty. This can lead to complete CPU stall with the
following message generated by RCU:
INFO: rcu_sched self-detected stall on CPU { 0} (t=2101 jiffies
g=54225 c=54224 q=11465)
Task dump for CPU 0:
tpch R running task 0 39949 39948 0x0000000a
ffffffff818536c0 ffff88181fa037a0 ffffffff8106a4be 0000000000000000
ffffffff818536c0 ffff88181fa037c0 ffffffff8106d8a8 ffff88181fa03800
0000000000000001 ffff88181fa037f0 ffffffff81094a50 ffff88181fa15680
Call Trace:
<IRQ> [<ffffffff8106a4be>] sched_show_task+0xae/0x120
[<ffffffff8106d8a8>] dump_cpu_task+0x38/0x40
[<ffffffff81094a50>] rcu_dump_cpu_stacks+0x90/0xd0
[<ffffffff81097c3b>] rcu_check_callbacks+0x3eb/0x6e0
[<ffffffff8106e53f>] ? account_system_time+0x7f/0x170
[<ffffffff81099e64>] update_process_times+0x34/0x60
[<ffffffff810a84d1>] tick_sched_handle.isra.18+0x31/0x40
[<ffffffff810a851c>] tick_sched_timer+0x3c/0x70
[<ffffffff8109a43d>] __run_hrtimer.isra.34+0x3d/0xc0
[<ffffffff8109aa95>] hrtimer_interrupt+0xc5/0x1e0
[<ffffffff81030d52>] ? native_smp_send_reschedule+0x42/0x60
[<ffffffff81032f04>] local_apic_timer_interrupt+0x34/0x60
[<ffffffff810335bc>] smp_apic_timer_interrupt+0x3c/0x60
[<ffffffff8165a3fb>] apic_timer_interrupt+0x6b/0x70
[<ffffffff81659129>] ? _raw_spin_unlock_irqrestore+0x9/0x10
[<ffffffff8107eb9f>] __wake_up_sync_key+0x4f/0x60
[<ffffffffa313ddd1>] tipc_write_space+0x31/0x40 [tipc]
[<ffffffffa313dadf>] filter_rcv+0x31f/0x520 [tipc]
[<ffffffffa313d699>] ? tipc_sk_lookup+0xc9/0x110 [tipc]
[<ffffffff81659259>] ? _raw_spin_lock_bh+0x19/0x30
[<ffffffffa314122c>] tipc_sk_rcv+0x2dc/0x3e0 [tipc]
[<ffffffffa312e7ff>] tipc_bclink_wakeup_users+0x2f/0x40 [tipc]
[<ffffffffa313ce26>] tipc_node_unlock+0x186/0x190 [tipc]
[<ffffffff81597c1c>] ? kfree_skb+0x2c/0x40
[<ffffffffa313475c>] tipc_rcv+0x2ac/0x8c0 [tipc]
[<ffffffffa312ff58>] tipc_l2_rcv_msg+0x38/0x50 [tipc]
[<ffffffff815a76d3>] __netif_receive_skb_core+0x5a3/0x950
[<ffffffff815a98d3>] __netif_receive_skb+0x13/0x60
[<ffffffff815a993e>] netif_receive_skb_internal+0x1e/0x90
[<ffffffff815aa138>] napi_gro_receive+0x78/0xa0
[<ffffffffa07f93f4>] tg3_poll_work+0xc54/0xf40 [tg3]
[<ffffffff81597c8c>] ? consume_skb+0x2c/0x40
[<ffffffffa07f9721>] tg3_poll_msix+0x41/0x160 [tg3]
[<ffffffff815ab0f2>] net_rx_action+0xe2/0x290
[<ffffffff8104b92a>] __do_softirq+0xda/0x1f0
[<ffffffff8104bc26>] irq_exit+0x76/0xa0
[<ffffffff81004355>] do_IRQ+0x55/0xf0
[<ffffffff8165a12b>] common_interrupt+0x6b/0x6b
<EOI>
The issue occurs only when tipc_sk_rcv() is used to wake up postponed
senders:
tipc_bclink_wakeup_users()
// wakeupq - is a queue which consists of special
// messages with SOCK_WAKEUP type.
tipc_sk_rcv(wakeupq)
...
while (skb_queue_len(inputq)) {
filter_rcv(skb)
// Here the type of message is checked
// and if it is SOCK_WAKEUP then
// it tries to wake up a sender.
tipc_write_space(sk)
wake_up_interruptible_sync_poll()
}
After the sender thread is woke up it can gather control and perform
an attempt to send a message. But if there is no enough place in send
queue it will call link_schedule_user() function which puts a message
of type SOCK_WAKEUP to the wakeup queue and put the sender to sleep.
Thus the size of the queue actually is not changed and the while()
loop never exits.
The approach I proposed is to wake up only senders for which there is
enough place in send queue so the described issue can't occur.
Moreover the same approach is already used to wake up senders on
unicast links.
I have got into the issue on our product code but to reproduce the
issue I changed a benchmark test application (from
tipcutils/demos/benchmark) to perform the following scenario:
1. Run 64 instances of test application (nodes). It can be done
on the one physical machine.
2. Each application connects to all other using TIPC sockets in
RDM mode.
3. When setup is done all nodes start simultaneously send
broadcast messages.
4. Everything hangs up.
The issue is reproducible only when a congestion on broadcast link
occurs. For example, when there are only 8 nodes it works fine since
congestion doesn't occur. Send queue limit is 40 in my case (I use a
critical importance level) and when 64 nodes send a message at the
same moment a congestion occurs every time.
Signed-off-by: Dmitry S Kolmakov <kolmakov.dmitriy@huawei.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
We convert packet/message reception according to the same principle
we have been using for message sending and timeout handling:
We move the function tipc_rcv() to node.c, hence handling the initial
packet reception at the link aggregation level. The function grabs
the node lock, selects the receiving link, and accesses it via a new
call tipc_link_rcv(). This function appends buffers to the input
queue for delivery upwards, but it may also append outgoing packets
to the xmit queue, just as we do during regular message sending. The
latter will happen when buffers are forwarded from the link backlog,
or when retransmission is requested.
Upon return of this function, and after having released the node lock,
tipc_rcv() delivers/tranmsits the contents of those queues, but it may
also perform actions such as link activation or reset, as indicated by
the return flags from the link.
This reduces the number of cpu cycles spent inside the node spinlock,
and reduces contention on that lock.
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 status flag LINK_STOPPED is not needed any more, since the
mechanism for delayed deletion of links has been removed.
Likewise, LINK_STARTED and LINK_START_EVT are unnecessary,
because we can just as well start the link timer directly from
inside tipc_link_create().
We eliminate these flags in this commit.
Instead of the above flags, we now introduce three new link modes,
TIPC_LINK_OPEN, TIPC_LINK_BLOCKED and TIPC_LINK_TUNNEL. The values
indicate whether, and in the case of TIPC_LINK_TUNNEL, which, messages
the link is allowed to receive in this state. TIPC_LINK_BLOCKED also
blocks timer-driven protocol messages to be sent out, and any change
to the link FSM. Since the modes are mutually exclusive, we convert
them to state values, and rename the 'flags' field in struct tipc_link
to 'exec_mode'.
Finally, we move the #defines for link FSM states and events from link.h
into enums inside the file link.c, which is the real usage scope of
these definitions.
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 the function tipc_link_xmit() is given a buffer list for
transmission, it currently consumes the list both when transmission
is successful and when it fails, except for the special case when
it encounters link congestion.
This behavior is inconsistent, and needs to be corrected if we want
to avoid problems in later commits in this series.
In this commit, we change this to let the function consume the list
only when transmission is successful, and leave the list with the
sender in all other cases. We also modifiy the socket code so that
it adapts to this change, i.e., purges the list when a non-congestion
error code is returned.
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>
|
|
struct 'tipc_node' currently contains two arrays for link attributes,
one for the link pointers, and one for the usable link MTUs.
We now group those into a new struct 'tipc_link_entry', and intoduce
one single array consisting of such enties. Apart from being a cosmetic
improvement, this is a starting point for the strict master-slave
relation between node and link that we will introduce in the following
commits.
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 commit 1f66d161ab3d8b518903fa6c3f9c1f48d6919e74
("tipc: introduce starvation free send algorithm")
we introduced a counter per priority level for buffers
in the link backlog queue. We also introduced a new
function tipc_link_purge_backlog(), to reset these
counters to zero when the link is reset.
Unfortunately, we missed to call this function when
the broadcast link is reset, with the result that the
values of these counters might be permanently skewed
when new nodes are attached. This may in the worst case
lead to permananent, but spurious, broadcast link
congestion, where no broadcast packets can be sent at
all.
We fix this bug with this commit.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Currently, the packet sequence number is updated and added to each
packet at the moment a packet is added to the link backlog queue.
This is wasteful, since it forces the code to traverse the send
packet list packet by packet when adding them to the backlog queue.
It would be better to just splice the whole packet list into the
backlog queue when that is the right action to do.
In this commit, we do this change. Also, since the sequence numbers
cannot now be assigned to the packets at the moment they are added
the backlog queue, we do instead calculate and add them at the moment
of transmission, when the backlog queue has to be traversed anyway.
We do this in the function tipc_link_push_packet().
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
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 rename some fields in struct tipc_link, in order to give them more
descriptive names:
next_in_no -> rcv_nxt
next_out_no-> snd_nxt
fsm_msg_cnt-> silent_intv_cnt
cont_intv -> keepalive_intv
last_retransmitted -> last_retransm
There are no functional changes in this commit.
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
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>
|
|
Add the ability to get or set the broadcast link window through the
new netlink API. The functionality was unintentionally missing from
the new netlink API. Adding this means that we also fix the breakage
in the old API when coming through the compat layer.
Fixes: 37e2d4843f9e (tipc: convert legacy nl link prop set to nl compat)
Reported-by: Tomi Ollila <tomi.ollila@iki.fi>
Signed-off-by: Richard Alpe <richard.alpe@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When a link is being established, the two endpoints advertise their
respective interface MTU in the transmitted RESET and ACTIVATE messages.
If there is any difference, the lower of the two MTUs will be selected
for use by both endpoints.
However, as a remnant of earlier attempts to introduce TIPC level
routing. there also exists an MTU discovery mechanism. If an intermediate
node has a lower MTU than the two endpoints, they will discover this
through a bisectional approach, and finally adopt this MTU for common use.
Since there is no TIPC level routing, and probably never will be,
this mechanism doesn't make any sense, and only serves to make the
link level protocol unecessarily complex.
In this commit, we eliminate the MTU discovery algorithm,and fall back
to the simple MTU advertising approach. This change is fully backwards
compatible.
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>
|
|
TIPC node hash node table is protected with rcu lock on read side.
tipc_node_find() is used to look for a node object with node address
through iterating the hash node table. As the entire process of what
tipc_node_find() traverses the table is guarded with rcu read lock,
it's safe for us. However, when callers use the node object returned
by tipc_node_find(), there is no rcu read lock applied. Therefore,
this is absolutely unsafe for callers of tipc_node_find().
Now we introduce a reference counter for node structure. Before
tipc_node_find() returns node object to its caller, it first increases
the reference counter. Accordingly, after its caller used it up,
it decreases the counter again. This can prevent a node being used by
one thread from being freed by another thread.
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Jon Maloy <jon.maloy@ericson.com>
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
[ 60.988363] ======================================================
[ 60.988754] [ INFO: possible circular locking dependency detected ]
[ 60.989152] 3.19.0+ #194 Not tainted
[ 60.989377] -------------------------------------------------------
[ 60.989781] swapper/3/0 is trying to acquire lock:
[ 60.990079] (&(&n_ptr->lock)->rlock){+.-...}, at: [<ffffffffa0006dca>] tipc_link_retransmit+0x1aa/0x240 [tipc]
[ 60.990743]
[ 60.990743] but task is already holding lock:
[ 60.991106] (&(&bclink->lock)->rlock){+.-...}, at: [<ffffffffa00004be>] tipc_bclink_lock+0x8e/0xa0 [tipc]
[ 60.991738]
[ 60.991738] which lock already depends on the new lock.
[ 60.991738]
[ 60.992174]
[ 60.992174] the existing dependency chain (in reverse order) is:
[ 60.992174]
-> #1 (&(&bclink->lock)->rlock){+.-...}:
[ 60.992174] [<ffffffff810a9c0c>] lock_acquire+0x9c/0x140
[ 60.992174] [<ffffffff8179c41f>] _raw_spin_lock_bh+0x3f/0x50
[ 60.992174] [<ffffffffa00004be>] tipc_bclink_lock+0x8e/0xa0 [tipc]
[ 60.992174] [<ffffffffa0000f57>] tipc_bclink_add_node+0x97/0xf0 [tipc]
[ 60.992174] [<ffffffffa0011815>] tipc_node_link_up+0xf5/0x110 [tipc]
[ 60.992174] [<ffffffffa0007783>] link_state_event+0x2b3/0x4f0 [tipc]
[ 60.992174] [<ffffffffa00193c0>] tipc_link_proto_rcv+0x24c/0x418 [tipc]
[ 60.992174] [<ffffffffa0008857>] tipc_rcv+0x827/0xac0 [tipc]
[ 60.992174] [<ffffffffa0002ca3>] tipc_l2_rcv_msg+0x73/0xd0 [tipc]
[ 60.992174] [<ffffffff81646e66>] __netif_receive_skb_core+0x746/0x980
[ 60.992174] [<ffffffff816470c1>] __netif_receive_skb+0x21/0x70
[ 60.992174] [<ffffffff81647295>] netif_receive_skb_internal+0x35/0x130
[ 60.992174] [<ffffffff81648218>] napi_gro_receive+0x158/0x1d0
[ 60.992174] [<ffffffff81559e05>] e1000_clean_rx_irq+0x155/0x490
[ 60.992174] [<ffffffff8155c1b7>] e1000_clean+0x267/0x990
[ 60.992174] [<ffffffff81647b60>] net_rx_action+0x150/0x360
[ 60.992174] [<ffffffff8105ec43>] __do_softirq+0x123/0x360
[ 60.992174] [<ffffffff8105f12e>] irq_exit+0x8e/0xb0
[ 60.992174] [<ffffffff8179f9f5>] do_IRQ+0x65/0x110
[ 60.992174] [<ffffffff8179da6f>] ret_from_intr+0x0/0x13
[ 60.992174] [<ffffffff8100de9f>] arch_cpu_idle+0xf/0x20
[ 60.992174] [<ffffffff8109dfa6>] cpu_startup_entry+0x2f6/0x3f0
[ 60.992174] [<ffffffff81033cda>] start_secondary+0x13a/0x150
[ 60.992174]
-> #0 (&(&n_ptr->lock)->rlock){+.-...}:
[ 60.992174] [<ffffffff810a8f7d>] __lock_acquire+0x163d/0x1ca0
[ 60.992174] [<ffffffff810a9c0c>] lock_acquire+0x9c/0x140
[ 60.992174] [<ffffffff8179c41f>] _raw_spin_lock_bh+0x3f/0x50
[ 60.992174] [<ffffffffa0006dca>] tipc_link_retransmit+0x1aa/0x240 [tipc]
[ 60.992174] [<ffffffffa0001e11>] tipc_bclink_rcv+0x611/0x640 [tipc]
[ 60.992174] [<ffffffffa0008646>] tipc_rcv+0x616/0xac0 [tipc]
[ 60.992174] [<ffffffffa0002ca3>] tipc_l2_rcv_msg+0x73/0xd0 [tipc]
[ 60.992174] [<ffffffff81646e66>] __netif_receive_skb_core+0x746/0x980
[ 60.992174] [<ffffffff816470c1>] __netif_receive_skb+0x21/0x70
[ 60.992174] [<ffffffff81647295>] netif_receive_skb_internal+0x35/0x130
[ 60.992174] [<ffffffff81648218>] napi_gro_receive+0x158/0x1d0
[ 60.992174] [<ffffffff81559e05>] e1000_clean_rx_irq+0x155/0x490
[ 60.992174] [<ffffffff8155c1b7>] e1000_clean+0x267/0x990
[ 60.992174] [<ffffffff81647b60>] net_rx_action+0x150/0x360
[ 60.992174] [<ffffffff8105ec43>] __do_softirq+0x123/0x360
[ 60.992174] [<ffffffff8105f12e>] irq_exit+0x8e/0xb0
[ 60.992174] [<ffffffff8179f9f5>] do_IRQ+0x65/0x110
[ 60.992174] [<ffffffff8179da6f>] ret_from_intr+0x0/0x13
[ 60.992174] [<ffffffff8100de9f>] arch_cpu_idle+0xf/0x20
[ 60.992174] [<ffffffff8109dfa6>] cpu_startup_entry+0x2f6/0x3f0
[ 60.992174] [<ffffffff81033cda>] start_secondary+0x13a/0x150
[ 60.992174]
[ 60.992174] other info that might help us debug this:
[ 60.992174]
[ 60.992174] Possible unsafe locking scenario:
[ 60.992174]
[ 60.992174] CPU0 CPU1
[ 60.992174] ---- ----
[ 60.992174] lock(&(&bclink->lock)->rlock);
[ 60.992174] lock(&(&n_ptr->lock)->rlock);
[ 60.992174] lock(&(&bclink->lock)->rlock);
[ 60.992174] lock(&(&n_ptr->lock)->rlock);
[ 60.992174]
[ 60.992174] *** DEADLOCK ***
[ 60.992174]
[ 60.992174] 3 locks held by swapper/3/0:
[ 60.992174] #0: (rcu_read_lock){......}, at: [<ffffffff81646791>] __netif_receive_skb_core+0x71/0x980
[ 60.992174] #1: (rcu_read_lock){......}, at: [<ffffffffa0002c35>] tipc_l2_rcv_msg+0x5/0xd0 [tipc]
[ 60.992174] #2: (&(&bclink->lock)->rlock){+.-...}, at: [<ffffffffa00004be>] tipc_bclink_lock+0x8e/0xa0 [tipc]
[ 60.992174]
The correct the sequence of grabbing n_ptr->lock and bclink->lock
should be that the former is first held and the latter is then taken,
which exactly happened on CPU1. But especially when the retransmission
of broadcast link is failed, bclink->lock is first held in
tipc_bclink_rcv(), and n_ptr->lock is taken in link_retransmit_failure()
called by tipc_link_retransmit() subsequently, which is demonstrated on
CPU0. As a result, deadlock occurs.
If the order of holding the two locks happening on CPU0 is reversed, the
deadlock risk will be relieved. Therefore, the node lock taken in
link_retransmit_failure() originally is moved to tipc_bclink_rcv()
so that it's obtained before bclink lock. But the precondition of
the adjustment of node lock is that responding to bclink reset event
must be moved from tipc_bclink_unlock() to tipc_node_unlock().
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Currently, we only use a single counter; the length of the backlog
queue, to determine whether a message should be accepted to the queue
or not. Each time a message is being sent, the queue length is compared
to a threshold value for the message's importance priority. If the queue
length is beyond this threshold, the message is rejected. This algorithm
implies a risk of starvation of low importance senders during very high
load, because it may take a long time before the backlog queue has
decreased enough to accept a lower level message.
We now eliminate this risk by introducing a counter for each importance
priority. When a message is sent, we check only the queue level for that
particular message's priority. If that is ok, the message can be added
to the backlog, irrespective of the queue level for other priorities.
This way, each level is guaranteed a certain portion of the total
bandwidth, and any risk of starvation is eliminated.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When a node joins a cluster while we are transmitting a fragment
stream over the broadcast link, it's missing the preceding fragments
needed to build a meaningful message. As a result, the node has to
drop it. However, as the fragment message is not acknowledged to
its sender before it's dropped, it accidentally causes link reset
of retransmission failure on the node.
Reported-by: Erik Hugne <erik.hugne@ericsson.com>
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Tested-by: Erik Hugne <erik.hugne@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Messages transferred by TIPC are assigned an "importance priority", -an
integer value indicating how to treat the message when there is link or
destination socket congestion.
There is no separate header field for this value. Instead, the message
user values have been chosen in ascending order according to perceived
importance, so that the message user field can be used for this.
This is not a good solution. First, we have many more users than the
needed priority levels, so we end up with treating more priority
levels than necessary. Second, the user field cannot always
accurately reflect the priority of the message. E.g., a message
fragment packet should really have the priority of the enveloped
user data message, and not the priority of the MSG_FRAGMENTER user.
Until now, we have been working around this problem in different ways,
but it is now time to implement a consistent way of handling such
priorities, although still within the constraint that we cannot
allocate any more bits in the regular data message header for this.
In this commit, we define a new priority level, TIPC_SYSTEM_IMPORTANCE,
that will be the only one used apart from the four (lower) user data
levels. All non-data messages map down to this priority. Furthermore,
we take some free bits from the MSG_FRAGMENTER header and allocate
them to store the priority of the enveloped message. We then adjust
the functions msg_importance()/msg_set_importance() so that they
read/set the correct header fields depending on user type.
This small protocol change is fully compatible, because the code at
the receiving end of a link currently reads the importance level
only from user data messages, where there is no change.
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
struct tipc_link contains one single queue for outgoing packets,
where both transmitted and waiting packets are queued.
This infrastructure is hard to maintain, because we need
to keep a number of fields to keep track of which packets are
sent or unsent, and the number of packets in each category.
A lot of code becomes simpler if we split this queue into a transmission
queue, where sent/unacknowledged packets are kept, and a backlog queue,
where we keep the not yet sent packets.
In this commit we do this separation.
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
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 unicast packet header contains a broadcast acknowledge sequence
number, that may need to be conveyed to the broadcast link for proper
treatment. Currently, the function tipc_rcv(), which is on the most
critical data path, calls the function tipc_bclink_acknowledge() to
have this done. This call is made for each received packet, and results
in the unconditional grabbing of the broadcast link spinlock.
This is unnecessary, since we can see directly from tipc_rcv() if
the acknowledged number differs from what has been previously acked
from the node in question. In the vast majority of cases the numbers
won't differ, and there is nothing to update.
We now make the call to tipc_bclink_acknowledge() conditional
to that the two ack values differ.
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
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>
|
|
Add functionality for safely appending string data to a TLV without
keeping write count in the caller.
Convert TIPC_CMD_SHOW_LINK_STATS to compat dumpit.
Signed-off-by: Richard Alpe <richard.alpe@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The new netlink API is no longer "v2" but rather the standard API and
the legacy API is now "nl compat". We split them into separate
start/stop and put them in different files in order to further
distinguish them.
Signed-off-by: Richard Alpe <richard.alpe@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
In a previous commit in this series we resolved a race problem during
unicast message reception.
Here, we resolve the same problem at multicast reception. We apply the
same technique: an input queue serializing the delivery of arriving
buffers. The main difference is that here we do it in two steps.
First, the broadcast link feeds arriving buffers into the tail of an
arrival queue, which head is consumed at the socket level, and where
destination lookup is performed. Second, if the lookup is successful,
the resulting buffer clones are fed into a second queue, the input
queue. This queue is consumed at reception in the socket just like
in the unicast case. Both queues are protected by the same lock, -the
one of the input queue.
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 structure 'tipc_port_list' is used to collect port numbers
representing multicast destination socket on a receiving node.
The list is not based on a standard linked list, and is in reality
optimized for the uncommon case that there are more than one
multicast destinations per node. This makes the list handling
unecessarily complex, and as a consequence, even the socket
multicast reception becomes more complex.
In this commit, we replace 'tipc_port_list' with a new 'struct
tipc_plist', which is based on a standard list. We give the new
list stack (push/pop) semantics, someting that simplifies
the implementation of the function tipc_sk_mcast_rcv().
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>
|
|
TIPC handles message cardinality and sequencing at the link layer,
before passing messages upwards to the destination sockets. During the
upcall from link to socket no locks are held. It is therefore possible,
and we see it happen occasionally, that messages arriving in different
threads and delivered in sequence still bypass each other before they
reach the destination socket. This must not happen, since it violates
the sequentiality guarantee.
We solve this by adding a new input buffer queue to the link structure.
Arriving messages are added safely to the tail of that queue by the
link, while the head of the queue is consumed, also safely, by the
receiving socket. Sequentiality is secured per socket by only allowing
buffers to be dequeued inside the socket lock. Since there may be multiple
simultaneous readers of the queue, we use a 'filter' parameter to reduce
the risk that they peek the same buffer from the queue, hence also
reducing the risk of contention on the receiving socket locks.
This solves the sequentiality problem, and seems to cause no measurable
performance degradation.
A nice side effect of this change is that lock handling in the functions
tipc_rcv() and tipc_bcast_rcv() now becomes uniform, something that
will enable future simplifications of those functions.
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 most common usage of namespace information is when we fetch the
own node addess from the net structure. This leads to a lot of
passing around of a parameter of type 'struct net *' between
functions just to make them able to obtain this address.
However, in many cases this is unnecessary. The own node address
is readily available as a member of both struct tipc_sock and
tipc_link, and can be fetched from there instead.
The fact that the vast majority of functions in socket.c and link.c
anyway are maintaining a pointer to their respective base structures
makes this option even more compelling.
In this commit, we introduce the inline functions tsk_own_node()
and link_own_node() to make it easy for functions to fetch the node
address from those structs instead of having to pass along and
dereference the namespace struct.
In particular, we make calls to the msg_xx() functions in msg.{h,c}
context independent by directly passing them the own node address
as parameter when needed. Those functions should be regarded as
leaves in the code dependency tree, and it is hence desirable to
keep them namspace unaware.
Apart from a potential positive effect on cache behavior, these
changes make it easier to introduce the changes that will follow
later in this series.
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>
|
|
If net namespace is supported in tipc, each namespace will be treated
as a separate tipc node. Therefore, every namespace must own its
private tipc node address. This means the "tipc_own_addr" global
variable of node address must be moved to tipc_net structure to
satisfy the requirement. It's turned out that users also can assign
node address for every namespace.
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Tested-by: Tero Aho <Tero.Aho@coriant.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
TIPC broadcast link is statically established and its relevant states
are maintained with the global variables: "bcbearer", "bclink" and
"bcl". Allowing different namespace to own different broadcast link
instances, these variables must be moved to tipc_net structure and
broadcast link instances would be allocated and initialized when
namespace is created.
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Tested-by: Tero Aho <Tero.Aho@coriant.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Bearer list defined as a global variable is used to store bearer
instances. When tipc supports net namespace, bearers created in
one namespace must be isolated with others allocated in other
namespaces, which requires us that the bearer list(bearer_list)
must be moved to tipc_net structure. As a result, a net namespace
pointer has to be passed to functions which access the bearer list.
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Tested-by: Tero Aho <Tero.Aho@coriant.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Global variables associated with node table are below:
- node table list (node_htable)
- node hash table list (tipc_node_list)
- node table lock (node_list_lock)
- node number counter (tipc_num_nodes)
- node link number counter (tipc_num_links)
To make node table support namespace, above global variables must be
moved to tipc_net structure in order to keep secret for different
namespaces. As a consequence, these variables are allocated and
initialized when namespace is created, and deallocated when namespace
is destroyed. After the change, functions associated with these
variables have to utilize a namespace pointer to access them. So
adding namespace pointer as a parameter of these functions is the
major change made in the commit.
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Tested-by: Tero Aho <Tero.Aho@coriant.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Involve namespace infrastructure, make the "tipc_net_id" global
variable aware of per namespace, and rename it to "net_id". In
order that the conversion can be successfully done, an instance
of networking namespace must be passed to relevant functions,
allowing them to access the "net_id" variable of per namespace.
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Tested-by: Tero Aho <Tero.Aho@coriant.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
In commit 58dc55f25631178ee74cd27185956a8f7dcb3e32 ("tipc: use generic
SKB list APIs to manage link transmission queue") we replace all list
traversal loops with the macros skb_queue_walk() or
skb_queue_walk_safe(). While the previous loops were based on the
assumption that the list was NULL-terminated, the standard macros
stop when the iterator reaches the list head, which is non-NULL.
In the function bclink_retransmit_pkt() this macro replacement has
lead to a bug. When we receive a BCAST STATE_MSG we unconditionally
call the function bclink_retransmit_pkt(), whether there really is
anything to retransmit or not, assuming that the sequence number
comparisons will lead to the correct behavior. However, if the
transmission queue is empty, or if there are no eligible buffers in
the transmission queue, we will by mistake pass the list head pointer
to the function tipc_link_retransmit(). Since the list head is not a
valid sk_buff, this leads to a crash.
In this commit we fix this by only calling tipc_link_retransmit()
if we actually found eligible buffers in the transmission queue.
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>
|
|
commit 908344cdda80 ("tipc: fix bug in multicast congestion
handling") introduced two bugs with the bclink wakeup
function. This commit fixes the missing spinlock init for the
waiting_sks list. We also eliminate the race condition
between the waiting_sks length check/dequeue operations in
tipc_bclink_wakeup_users by simply removing the redundant
length check.
Signed-off-by: Erik Hugne <erik.hugne@ericsson.com>
Acked-by: Tero Aho <Tero.Aho@coriant.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use standard SKB list APIs associated with struct sk_buff_head to
manage socket outgoing packet chain and name table outgoing packet
chain, having relevant code simpler and more readable.
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use standard SKB list APIs associated with struct sk_buff_head to
manage link's deferred queue, simplifying relevant code.
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use standard SKB list APIs associated with struct sk_buff_head to
manage link transmission queue, having relevant code more clean.
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
In original tipc_link_push_packet(), it pushes messages from protocol
message queue, retransmission queue and next_out queue. But as the two
first queues are removed, we can simplify its relevant code through
deleting tipc_link_push_queue().
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Fix sparse warnings about non-static declaration of static functions
in the new tipc netlink API.
Signed-off-by: Richard Alpe <richard.alpe@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add TIPC_NL_LINK_GET command to the new tipc netlink API.
This command supports dumping all information about all links
(including the broadcast link) or getting all information about a
specific link (not the broadcast link).
The information about a link includes name, transmission info,
properties and link statistics.
As the tipc broadcast link is special we unfortunately have to treat
it specially. It is a deliberate decision not to abstract the
broadcast link on this (API) level.
Netlink logical layout of link response message:
-> port
-> name
-> MTU
-> RX
-> TX
-> up flag
-> active flag
-> properties
-> priority
-> tolerance
-> window
-> statistics
-> rx_info
-> rx_fragments
-> rx_fragmented
-> rx_bundles
-> rx_bundled
-> tx_info
-> tx_fragments
-> tx_fragmented
-> tx_bundles
-> tx_bundled
-> msg_prof_tot
-> msg_len_cnt
-> msg_len_tot
-> msg_len_p0
-> msg_len_p1
-> msg_len_p2
-> msg_len_p3
-> msg_len_p4
-> msg_len_p5
-> msg_len_p6
-> rx_states
-> rx_probes
-> rx_nacks
-> rx_deferred
-> tx_states
-> tx_probes
-> tx_nacks
-> tx_acks
-> retransmitted
-> duplicates
-> link_congs
-> max_queue
-> avg_queue
Signed-off-by: Richard Alpe <richard.alpe@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
