linux-toradex.git/net/tipc, branch v3.0.98

tipc: fix lockdep warning during bearer initialization

2013-09-14T12:50:48+00:00

[ Upstream commit 4225a398c1352a7a5c14dc07277cb5cc4473983b ]

When the lockdep validator is enabled, it will report the below
warning when we enable a TIPC bearer:

[ INFO: possible irq lock inversion dependency detected ]
---------------------------------------------------------
Possible interrupt unsafe locking scenario:

        CPU0                    CPU1
        ----                    ----
   lock(ptype_lock);
                                local_irq_disable();
                                lock(tipc_net_lock);
                                lock(ptype_lock);
   
   lock(tipc_net_lock);

  *** DEADLOCK ***

the shortest dependencies between 2nd lock and 1st lock:
  -> (ptype_lock){+.+...} ops: 10 {
[...]
SOFTIRQ-ON-W at:
                      [] __lock_acquire+0x528/0x13e0
                      [] lock_acquire+0x90/0x100
                      [] _raw_spin_lock+0x38/0x50
                      [] dev_add_pack+0x3a/0x60
                      [] arp_init+0x1a/0x48
                      [] inet_init+0x181/0x27e
                      [] do_one_initcall+0x34/0x170
                      [] kernel_init+0x110/0x1b2
                      [] kernel_thread_helper+0x6/0x10
[...]
   ... key      at: [] ptype_lock+0x10/0x20
   ... acquired at:
    [] lock_acquire+0x90/0x100
    [] _raw_spin_lock+0x38/0x50
    [] dev_add_pack+0x3a/0x60
    [] enable_bearer+0xf2/0x140 [tipc]
    [] tipc_enable_bearer+0x1ba/0x450 [tipc]
    [] tipc_cfg_do_cmd+0x5c4/0x830 [tipc]
    [] handle_cmd+0x42/0xd0 [tipc]
    [] genl_rcv_msg+0x232/0x280
    [] netlink_rcv_skb+0x86/0xb0
    [] genl_rcv+0x1c/0x30
    [] netlink_unicast+0x174/0x1f0
    [] netlink_sendmsg+0x1eb/0x2d0
    [] sock_aio_write+0x161/0x170
    [] do_sync_write+0xac/0xf0
    [] vfs_write+0x156/0x170
    [] sys_write+0x42/0x70
    [] sysenter_do_call+0x12/0x38
[...]
}
  -> (tipc_net_lock){+..-..} ops: 4 {
[...]
    IN-SOFTIRQ-R at:
                     [] __lock_acquire+0x64a/0x13e0
                     [] lock_acquire+0x90/0x100
                     [] _raw_read_lock_bh+0x3d/0x50
                     [] tipc_recv_msg+0x1d/0x830 [tipc]
                     [] recv_msg+0x3f/0x50 [tipc]
                     [] __netif_receive_skb+0x22a/0x590
                     [] netif_receive_skb+0x2b/0xf0
                     [] pcnet32_poll+0x292/0x780
                     [] net_rx_action+0xfa/0x1e0
                     [] __do_softirq+0xae/0x1e0
[...]
}

>From the log, we can see three different call chains between
CPU0 and CPU1:

Time 0 on CPU0:

  kernel_init()->inet_init()->dev_add_pack()

At time 0, the ptype_lock is held by CPU0 in dev_add_pack();

Time 1 on CPU1:

  tipc_enable_bearer()->enable_bearer()->dev_add_pack()

At time 1, tipc_enable_bearer() first holds tipc_net_lock, and then
wants to take ptype_lock to register TIPC protocol handler into the
networking stack.  But the ptype_lock has been taken by dev_add_pack()
on CPU0, so at this time the dev_add_pack() running on CPU1 has to be
busy looping.

Time 2 on CPU0:

  netif_receive_skb()->recv_msg()->tipc_recv_msg()

At time 2, an incoming TIPC packet arrives at CPU0, hence
tipc_recv_msg() will be invoked. In tipc_recv_msg(), it first wants
to hold tipc_net_lock.  At the moment, below scenario happens:

On CPU0, below is our sequence of taking locks:

  lock(ptype_lock)->lock(tipc_net_lock)

On CPU1, our sequence of taking locks looks like:

  lock(tipc_net_lock)->lock(ptype_lock)

Obviously deadlock may happen in this case.

But please note the deadlock possibly doesn't occur at all when the
first TIPC bearer is enabled.  Before enable_bearer() -- running on
CPU1 does not hold ptype_lock, so the TIPC receive handler (i.e.
recv_msg()) is not registered successfully via dev_add_pack(), so
the tipc_recv_msg() cannot be called by recv_msg() even if a TIPC
message comes to CPU0. But when the second TIPC bearer is
registered, the deadlock can perhaps really happen.

To fix it, we will push the work of registering TIPC protocol
handler into workqueue context. After the change, both paths taking
ptype_lock are always in process contexts, thus, the deadlock should
never occur.

Signed-off-by: Ying Xue 
Signed-off-by: Jon Maloy 
Signed-off-by: Paul Gortmaker 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

tipc: fix info leaks via msg_name in recv_msg/recv_stream

2013-05-01T15:56:40+00:00

[ Upstream commit 60085c3d009b0df252547adb336d1ccca5ce52ec ]

The code in set_orig_addr() does not initialize all of the members of
struct sockaddr_tipc when filling the sockaddr info -- namely the union
is only partly filled. This will make recv_msg() and recv_stream() --
the only users of this function -- leak kernel stack memory as the
msg_name member is a local variable in net/socket.c.

Additionally to that both recv_msg() and recv_stream() fail to update
the msg_namelen member to 0 while otherwise returning with 0, i.e.
"success". This is the case for, e.g., non-blocking sockets. This will
lead to a 128 byte kernel stack leak in net/socket.c.

Fix the first issue by initializing the memory of the union with
memset(0). Fix the second one by setting msg_namelen to 0 early as it
will be updated later if we're going to fill the msg_name member.

Signed-off-by: Mathias Krause 
Cc: Jon Maloy 
Cc: Allan Stephens 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

tipc: Revise timings used when sending link request messages

2011-05-10T20:04:02+00:00

Revises the algorithm governing the sending of link request messages
to take into account the number of nodes each bearer is currently in
contact with, and to ensure more rapid rediscovery of neighboring nodes
if a bearer fails and then recovers.

The discovery object now sends requests at least once a second if it
is not in contact with any other nodes, and at least once a minute if
it has at least one neighbor; if contact with the only neighbor is
lost, the object immediately reverts to its initial rapid-fire search
timing to accelerate the rediscovery process.

In addition, the discovery object now stops issuing link request
messages if it is in contact with the only neighboring node it is
configured to communicate with, since further searching is unnecessary.

Signed-off-by: Allan Stephens 
Signed-off-by: Paul Gortmaker

tipc: Add monitoring of number of nodes discovered by bearer

2011-05-10T20:04:01+00:00

Augments TIPC's discovery object to track the number of neighboring nodes
having an active link to the associated bearer.

This means tipc_disc_update_link_req() becomes either one of:

       tipc_disc_add_dest()
or:
       tipc_disc_remove_dest()

depending on the code flow direction of things.

Signed-off-by: Allan Stephens 
Signed-off-by: Paul Gortmaker

tipc: Enhance sending of discovery object link request messages

2011-05-10T20:04:00+00:00

Augments TIPC's discovery object to send its initial neighbor discovery
request message as soon as the associated bearer is created, rather than
waiting for its first periodic timeout to occur, thereby speeding up the
discovery process. Also adds a check to suppress the initial request or
subsequent requests if the bearer is blocked at the time the request is
scheduled for transmission.

Signed-off-by: Allan Stephens 
Signed-off-by: Paul Gortmaker

tipc: Enhance handling of discovery object creation failures

2011-05-10T20:03:59+00:00

Modifies bearer creation and deletion code to improve handling of
scenarios when a neighbor discovery object cannot be created. The
creation routine now aborts the creation of a bearer if its discovery
object cannot be created, and deletes the newly created bearer, rather
than failing quietly and leaving an unusable bearer hanging around.

Since the exit via the goto label really isn't a definitive failure
in all cases, relabel it appropriately.

Signed-off-by: Allan Stephens 
Signed-off-by: Paul Gortmaker

tipc: Introduce routine to enqueue a chain of messages on link tx queue

2011-05-10T20:03:58+00:00

Create a helper routine to enqueue a chain of sk_buffs to a link's
transmit queue.  It improves readability and the new function is
anticipated to be used more than just once in the future as well.

Signed-off-by: Allan Stephens 
Signed-off-by: Paul Gortmaker

tipc: Avoid recomputation of outgoing message length

2011-05-10T20:03:57+00:00

Rework TIPC's message sending routines to take advantage of the total
amount of data value passed to it by the kernel socket infrastructure.
This change eliminates the need for TIPC to compute the size of outgoing
messages itself, as well as the check for an oversize message in
tipc_msg_build().  In addition, this change warrants an explanation:

   -     res = send_packet(NULL, sock, &my_msg, 0);
   +     res = send_packet(NULL, sock, &my_msg, bytes_to_send);

Previously, the final argument to send_packet() was ignored (since the
amount of data being sent was recalculated by a lower-level routine)
and we could just pass in a dummy value (0). Now that the
recalculation is being eliminated, the argument value being passed to
send_packet() is significant and we have to supply the actual amount
of data we want to send.

Signed-off-by: Allan Stephens 
Signed-off-by: Paul Gortmaker

tipc: Abort excessive send requests as early as possible

2011-05-10T20:03:56+00:00

Adds checks to TIPC's socket send routines to promptly detect and
abort attempts to send more than 66,000 bytes in a single TIPC
message or more than 2**31-1 bytes in a single TIPC byte stream request.
In addition, this ensures that the number of iovecs in a send request
does not exceed the limits of a standard integer variable.

Signed-off-by: Allan Stephens 
Signed-off-by: Paul Gortmaker

tipc: Strengthen checks for neighboring node discovery

2011-05-10T20:03:55+00:00

Enhances existing checks on the discovery domain associated with a TIPC
bearer. A bearer can no longer be configured to accept links from itself
only (which would be pointless), or to nodes outside its own cluster
(since multi-cluster support has now been removed from TIPC). Also, the
neighbor discovery routine now validates link setup requests against the
configured discovery domain for the bearer, rather than simply ensuring
the requesting node belongs to the node's own cluster.

Signed-off-by: Allan Stephens 
Signed-off-by: Paul Gortmaker