The policy for handling the skb list locks on the send and receive paths
is simple.
- On the send path we never need to grab the lock on the 'xmitq' list
when the destination is an exernal node.
- On the receive path we always need to grab the lock on the 'inputq'
list, irrespective of source node.
However, when transmitting node local messages those will eventually
end up on the receive path of a local socket, meaning that the argument
'xmitq' in tipc_node_xmit() will become the 'ínputq' argument in the
function tipc_sk_rcv(). This has been handled by always initializing
the spinlock of the 'xmitq' list at message creation, just in case it
may end up on the receive path later, and despite knowing that the lock
in most cases never will be used.
This approach is inaccurate and confusing, and has also concealed the
fact that the stated 'no lock grabbing' policy for the send path is
violated in some cases.
We now clean up this by never initializing the lock at message creation,
instead doing this at the moment we find that the message actually will
enter the receive path. At the same time we fix the four locations
where we incorrectly access the spinlock on the send/error path.
This patch also reverts commit d12cffe932 ("tipc: ensure head->lock
is initialised") which has now become redundant.
CC: Eric Dumazet <edumazet@google.com>
Reported-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Syzbot reported a memleak caused by grp members' deferredq list not
purged when the grp is be deleted.
The issue occurs when more(msg_grp_bc_seqno(hdr), m->bc_rcv_nxt) in
tipc_group_filter_msg() and the skb will stay in deferredq.
So fix it by calling __skb_queue_purge for each member's deferredq
in tipc_group_delete() when a tipc sk leaves the grp.
Fixes: b87a5ea31c ("tipc: guarantee group unicast doesn't bypass group broadcast")
Reported-by: syzbot+78fbe679c8ca8d264a8d@syzkaller.appspotmail.com
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Even if the NLA_F_NESTED flag was introduced more than 11 years ago, most
netlink based interfaces (including recently added ones) are still not
setting it in kernel generated messages. Without the flag, message parsers
not aware of attribute semantics (e.g. wireshark dissector or libmnl's
mnl_nlmsg_fprintf()) cannot recognize nested attributes and won't display
the structure of their contents.
Unfortunately we cannot just add the flag everywhere as there may be
userspace applications which check nlattr::nla_type directly rather than
through a helper masking out the flags. Therefore the patch renames
nla_nest_start() to nla_nest_start_noflag() and introduces nla_nest_start()
as a wrapper adding NLA_F_NESTED. The calls which add NLA_F_NESTED manually
are rewritten to use nla_nest_start().
Except for changes in include/net/netlink.h, the patch was generated using
this semantic patch:
@@ expression E1, E2; @@
-nla_nest_start(E1, E2)
+nla_nest_start_noflag(E1, E2)
@@ expression E1, E2; @@
-nla_nest_start_noflag(E1, E2 | NLA_F_NESTED)
+nla_nest_start(E1, E2)
Signed-off-by: Michal Kubecek <mkubecek@suse.cz>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
nla_nest_start could fail and requires a check. The fix returns
-EMSGSIZE if it fails.
Signed-off-by: Kangjie Lu <kjlu@umn.edu>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes the following sparse warnings:
net/tipc/link.c:376:5: warning: symbol 'link_bc_rcv_gap' was not declared. Should it be static?
net/tipc/link.c:823:6: warning: symbol 'link_prepare_wakeup' was not declared. Should it be static?
net/tipc/link.c:959:6: warning: symbol 'tipc_link_advance_backlog' was not declared. Should it be static?
net/tipc/link.c:1009:5: warning: symbol 'tipc_link_retrans' was not declared. Should it be static?
net/tipc/monitor.c:687:5: warning: symbol '__tipc_nl_add_monitor_peer' was not declared. Should it be static?
net/tipc/group.c:230:20: warning: symbol 'tipc_group_find_member' was not declared. Should it be static?
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
All of the conflicts were cases of overlapping changes.
In net/core/devlink.c, we have to make care that the
resouce size_params have become a struct member rather
than a pointer to such an object.
Signed-off-by: David S. Miller <davem@davemloft.net>
In commit 60c2530696 ("tipc: fix race between poll() and
setsockopt()") we introduced a pointer from struct tipc_group to the
'group_is_connected' flag in struct tipc_sock, so that this field can
be checked without dereferencing the group pointer of the latter struct.
The initial value for this flag is correctly set to 'false' when a
group is created, but we miss the case when no group is created at
all, in which case the initial value should be 'true'. This has the
effect that SOCK_RDM/DGRAM sockets sending datagrams never receive
POLLOUT if they request so.
This commit corrects this bug.
Fixes: 60c2530696 ("tipc: fix race between poll() and setsockopt()")
Reported-by: Hoang Le <hoang.h.le@dektek.com.au>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We rename struct tipc_server to struct tipc_topsrv. This reflect its now
specialized role as topology server. Accoringly, we change or add function
prefixes to make it clearer which functionality those belong to.
There are no functional changes in this commit.
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>
Letting tipc_poll() dereference a socket's pointer to struct tipc_group
entails a race risk, as the group item may be deleted in a concurrent
tipc_sk_join() or tipc_sk_leave() thread.
We now move the 'open' flag in struct tipc_group to struct tipc_sock,
and let the former retain only a pointer to the moved field. This will
eliminate the race risk.
Reported-by: syzbot+799dafde0286795858ac@syzkaller.appspotmail.com
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The current criteria for returning POLLOUT from a group member socket is
too simplistic. It basically returns POLLOUT as soon as the group has
external destinations, something obviously leading to a lot of spinning
during destination congestion situations. At the same time, the internal
congestion handling is unnecessarily complex.
We now change this as follows.
- We introduce an 'open' flag in struct tipc_group. This flag is used
only to help poll() get the setting of POLLOUT right, and *not* for
congeston handling as such. This means that a user can choose to
ignore an EAGAIN for a destination and go on sending messages to
other destinations in the group if he wants to.
- The flag is set to false every time we return EAGAIN on a send call.
- The flag is set to true every time any member, i.e., not necessarily
the member that caused EAGAIN, is removed from the small_win list.
- We remove the group member 'usr_pending' flag. The size of the send
window and presence in the 'small_win' list is sufficient criteria
for recognizing congestion.
This solution seems to be a reasonable compromise between 'anycast',
which is normally not waiting for POLLOUT for a specific destination,
and the other three send modes, which are.
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 a member joins a group, it also indicates a binding scope. This
makes it possible to create both node local groups, invisible to other
nodes, as well as cluster global groups, visible everywhere.
In order to avoid that different members end up having permanently
differing views of group size and memberhip, we must inhibit locally
and globally bound members from joining the same group.
We do this by using the binding scope as an additional separator between
groups. I.e., a member must ignore all membership events from sockets
using a different scope than itself, and all lookups for message
destinations must require an exact match between the message's lookup
scope and the potential target's binding scope.
Apart from making it possible to create local groups using the same
identity on different nodes, a side effect of this is that it now also
becomes possible to create a cluster global group with the same identity
across the same nodes, without interfering with the local groups.
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>
Currently, when a user is subscribing for binding table publications,
he will receive a PUBLISH event for all already existing matching items
in the binding table.
However, a group socket making a subscriptions doesn't need this initial
status update from the binding table, because it has already scanned it
during the join operation. Worse, the multiplicatory effect of issuing
mutual events for dozens or hundreds group members within a short time
frame put a heavy load on the topology server, with the end result that
scale out operations on a big group tend to take much longer than needed.
We now add a new filter option, TIPC_SUB_NO_STATUS, for topology server
subscriptions, so that this initial avalanche of events is suppressed.
This change, along with the previous commit, significantly improves the
range and speed of group scale out operations.
We keep the new option internal for the tipc driver, at least for now.
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 a socket is joining a group, we look up in the binding table to
find if there are already other members of the group present. This is
used for being able to return EAGAIN instead of EHOSTUNREACH if the
user proceeds directly to a send attempt.
However, the information in the binding table can be used to directly
set the created member in state MBR_PUBLISHED and send a JOIN message
to the peer, instead of waiting for a topology PUBLISH event to do this.
When there are many members in a group, the propagation time for such
events can be significant, and we can save time during the join
operation if we use the initial lookup result fully.
In this commit, we eliminate the member state MBR_DISCOVERED which has
been the result of the initial lookup, and do instead go directly to
MBR_PUBLISHED, which initiates the setup.
After this change, the tipc_member FSM looks as follows:
+-----------+
---->| PUBLISHED |-----------------------------------------------+
PUB- +-----------+ LEAVE/WITHRAW |
LISH |JOIN |
| +-------------------------------------------+ |
| | LEAVE/WITHDRAW | |
| | +------------+ | |
| | +----------->| PENDING |---------+ | |
| | |msg/maxactv +-+---+------+ LEAVE/ | | |
| | | | | WITHDRAW | | |
| | | +----------+ | | | |
| | | |revert/maxactv| | | |
| | | V V V V V
| +----------+ msg +------------+ +-----------+
+-->| JOINED |------>| ACTIVE |------>| LEAVING |--->
| +----------+ +--- -+------+ LEAVE/+-----------+DOWN
| A A | WITHDRAW A A A EVT
| | | |RECLAIM | | |
| | |REMIT V | | |
| | |== adv +------------+ | | |
| | +---------| RECLAIMING |--------+ | |
| | +-----+------+ LEAVE/ | |
| | |REMIT WITHDRAW | |
| | |< adv | |
| |msg/ V LEAVE/ | |
| |adv==ADV_IDLE+------------+ WITHDRAW | |
| +-------------| REMITTED |------------+ |
| +------------+ |
|PUBLISH |
JOIN +-----------+ LEAVE/WITHDRAW |
---->| JOINING |-----------------------------------------------+
+-----------+
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>
After the changes in the previous commit the group LEAVE sequence
can be simplified.
We now let the arrival of a LEAVE message unconditionally issue a group
DOWN event to the user. When a topology WITHDRAW event is received, the
member, if it still there, is set to state LEAVING, but we only issue a
group DOWN event when the link to the peer node is gone, so that no
LEAVE message is to be expected.
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 the current implementation, a group socket receiving topology
events about other members just converts the topology event message
into a group event message and stores it until it reaches the right
state to issue it to the user. This complicates the code unnecessarily,
and becomes impractical when we in the coming commits will need to
create and issue membership events independently.
In this commit, we change this so that we just notice the type and
origin of the incoming topology event, and then drop the buffer. Only
when it is time to actually send a group event to the user do we
explicitly create a new message and send it upwards.
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>
Analysis reveals that the member state MBR_QURANTINED in reality is
unnecessary, and can be replaced by the state MBR_JOINING at all
occurrencs.
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 handle a corner case in the function tipc_group_update_rcv_win().
During extreme pessure it might happen that a message receiver has all
its active senders in RECLAIMING or REMITTED mode, meaning that there
is nobody to reclaim advertisements from if an additional sender tries
to go active.
Currently we just set the new sender to ACTIVE anyway, hence at least
theoretically opening up for a receiver queue overflow by exceeding the
MAX_ACTIVE limit. The correct solution to this is to instead add the
member to the pending queue, while letting the oldest member in that
queue revert to JOINED state.
In this commit we refactor the code for handling message arrival from
a JOINED member, both to make it more comprehensible and to cover the
case described above.
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 remove the 'reclaiming' member list in struct tipc_group, since
it doesn't serve any purpose.
- We simplify the GRP_REMIT_MSG branch of tipc_group_protocol_rcv().
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 simplify the sorting algorithm in tipc_update_member(). We also make
the remaining conditional call to this function unconditional, since the
same condition now is tested for inside the said function.
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 rename some functions and variables, to make their purpose clearer.
- tipc_group::congested -> tipc_group::small_win. Members in this list
are not necessarily (and typically) congested. Instead, they may
*potentially* be subject to congestion because their send window is
less than ADV_IDLE, and therefore need to be checked during message
transmission.
- tipc_group_is_receiver() -> tipc_group_is_sender(). This socket will
accept messages coming from members fulfilling this condition, i.e.,
they are senders from this member's viewpoint.
- tipc_group_is_enabled() -> tipc_group_is_receiver(). Members
fulfilling this condition will accept messages sent from the current
socket, i.e., they are receivers from its viewpoint.
There are no functional changes in this commit.
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 commit 04d7b574b2 ("tipc: add multipoint-to-point flow control") we
introduced a protocol for preventing buffer overflow when many group
members try to simultaneously send messages to the same receiving member.
Stress test of this mechanism has revealed a couple of related bugs:
- When the receiving member receives an advertisement REMIT message from
one of the senders, it will sometimes prematurely activate a pending
member and send it the remitted advertisement, although the upper
limit for active senders has been reached. This leads to accumulation
of illegal advertisements, and eventually to messages being dropped
because of receive buffer overflow.
- When the receiving member leaves REMITTED state while a received
message is being read, we miss to look at the pending queue, to
activate the oldest pending peer. This leads to some pending senders
being starved out, and never getting the opportunity to profit from
the remitted advertisement.
We fix the former in the function tipc_group_proto_rcv() by returning
directly from the function once it becomes clear that the remitting
peer cannot leave REMITTED state at that point.
We fix the latter in the function tipc_group_update_rcv_win() by looking
up and activate the longest pending peer when it becomes clear that the
remitting peer now can leave REMITTED state.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
