This adds support for generating and distributing all the keys
specified in the third phase of SMP.

This will make possible to re-establish secure connections, resolve
private addresses and sign commands.

For now, the values generated are random.
Signed-off-by: NVinicius Costa Gomes <vinicius.gomes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

The ERTM receive buffer is now handled in a way that does not require
the busy queue and the associated polling code.
Signed-off-by: NMat Martineau <mathewm@codeaurora.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

This change moves most L2CAP ERTM receive buffer handling out of the
L2CAP core and in to the socket code.  It's up to the higher layer
(the socket code, in this case) to tell the core when its buffer is
full or has space available.  The recv op should always accept
incoming ERTM data or else the connection will go down.

Within the socket layer, an skb that does not fit in the socket
receive buffer will be temporarily stored.  When the socket is read
from, that skb will be placed in the receive buffer if possible.  Once
adequate buffer space becomes available, the L2CAP core is informed
and the ERTM local busy state is cleared.

Receive buffer management for non-ERTM modes is unchanged.
Signed-off-by: NMat Martineau <mathewm@codeaurora.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

The local busy state is entered and exited based on buffer status in
the socket layer (or other upper layer).  This change is in
preparation for general buffer status reports from the socket layer,
which will then be used to change the local busy status.
Signed-off-by: NMat Martineau <mathewm@codeaurora.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Even when the received tx_seq is expected, the frame still needs to be
dropped if the TX window is exceeded or the receiver is in the local
busy state.
Signed-off-by: NMat Martineau <mathewm@codeaurora.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Make it easier to use more normal logging styles later.
Signed-off-by: NJoe Perches <joe@perches.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

ERTM timeouts are defined in milliseconds, but need to be converted
to jiffies when passed to mod_timer().
Signed-off-by: NMat Martineau <mathewm@codeaurora.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Signed-off-by: NMat Martineau <mathewm@codeaurora.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

When the connection is ready we should set the connection
to CONNECTED so userspace can use it.
Signed-off-by: NVinicius Costa Gomes <vinicius.gomes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

We already have access to the chan, we don't have to access the
socket to get its imtu.
Signed-off-by: NVinicius Costa Gomes <vinicius.gomes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

We should not try to do any other type of configuration for
LE links when they become ready.
Signed-off-by: NVinicius Costa Gomes <vinicius.gomes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Instead of setting bits manually we use set_bit, test_bit, etc.
Also remove L2CAP_ prefix from macros.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Instead of making the bit operations manually, we now use set_bit,
test_bit, etc.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

This patch adds support for disconnecting the link when SMP procedure
takes more than 30 seconds.

SMP begins when either the Pairing Request command is sent or the
Pairing Response is received, and it ends when the link is encrypted
(or terminated). Vol 3, Part H Section 3.4.
Signed-off-by: NVinicius Costa Gomes <vinicius.gomes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

When authentication completes we shouldn't blindly accept any pending
L2CAP connect requests. If the socket has the defer_setup feature
enabled it should still wait for user space acceptance of the connect
request. The issue only happens for non-SSP connections since with SSP
the L2CAP Connect request may not be sent for non-SDP PSMs before
authentication has completed successfully.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

This adds support for resuming the user space traffic when SMP
negotiation is complete.
Signed-off-by: NVinicius Costa Gomes <vinicius.gomes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Start SMP procedure for LE connections. This modification intercepts
l2cap received frames and call proper SMP functions to start the SMP
procedure. By now, no keys are being used.
Signed-off-by: NVinicius Costa Gomes <vinicius.gomes@openbossa.org>
Signed-off-by: NAnderson Briglia <anderson.briglia@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Otherwise the wrong error can be returned.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

ERTM use the generic L2CAP timer functions to keep a reference to the
channel. This is useful for avoiding crashes.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

We now plan to use l2cap_set_timer and l2cap_clear_timer in ERTM timers.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

struct l2cap_chan has now its own refcnt that is compatible with the
socket refcnt, i.e., we won't see sk_refcnt = 0 and chan->refcnt > 0.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Now socket state is tracked by struct sock and channel state is tracked by
chan->state. At this point both says the same, but this is going to change
when we add AMP Support for example.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

close() calls l2cap_sock_kill() on l2cap_sock.c
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

This abstracts the call to sock_queue_recv_skb() into
l2cap_chan_ops->recv().
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Add an abstraction layer between L2CAP core and its users (only
l2cap_sock.c now). The first function implemented is new_connection() that
replaces calls to l2cap_sock_alloc() in l2cap_core.c
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

As a first step to remove l2cap_sock_alloc() and l2cap_sock_init() from
l2cap_core.c
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Local busy is encoded in a bitfield, but was not masked out correctly.
Signed-off-by: NMat Martineau <mathewm@codeaurora.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

When code was moved from l2cap_core.c to l2cap_sock.c in commit
6de0702b, one line was dropped
from the old __l2cap_sock_close() implementation. This sk_state
change should still be in l2cap_chan_close().
Signed-off-by: NMat Martineau <mathewm@codeaurora.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Add BT_POWER socket option used to control the power
characteristics of the underlying ACL link. When the remote end
has put the link in sniff mode and the host stack wants to send
data we need need to explicitly exit sniff mode to work well with
certain devices (For example, A2DP on Plantronics Voyager 855).
However, this causes problems with HID devices.

Hence, moving into active mode when sending data, irrespective
of who set the sniff mode has been made as a socket option. By
default, we will move into active mode. HID devices can set the
L2CAP socket option to prevent this from happening.

Currently, this has been implemented for L2CAP sockets. This has been
tested with incoming and outgoing L2CAP sockets for HID and A2DP.

Based on discussions on linux-bluetooth and patches submitted by
Andrei Emeltchenko.
Signed-off-by: NJaikumar Ganesh <jaikumar@google.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

sparse complains about a few things that should
be static.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

To make it consistent with the rest of the API.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

The call to l2cap_chan_clear_timer() is not really needed in l2cap_sock.c.
This patch also adds a call to l2cap_chan_clear_timer() to the only place
in __l2cap_sock_close() that wasn't calling it. It's safe call it there
because l2cap_chan_clear_timer() check first for timer_peding().
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

The new timer does not belong to struct sock, tought it still touch some
sock things, but this will be sorted out soon.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

chan_type says if our chan is raw(direclty access to HCI),
connection less or connection oriented.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

This move all the sending logic to l2cap_core.c, but we still have a
socket dependence there, struct msghdr. It will be removed in some of the
further commits.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

This is actually __l2cap_sock_close() renamed to __l2cap_chan_close().
At a first look it may not make sense, but with the further cleanups that
will come it will.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

In L2CAP_SDU_UNSEGMENTED case, if sock_queue_rcv_skb returns error,
l2cap_ertm_reassembly_sdu should not return 0 so as to insert the
skb into BUSY_QUEUE for later retries.
Signed-off-by: NRuiyi Zhang <Ruiyi.Zhang@Atheros.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

read_lock() ... read_unlock_bh() is clearly bogus.
This was broken by

commit 23691d75
Author: Gustavo F. Padovan <padovan@profusion.mobi>
Date:   Wed Apr 27 18:26:32 2011 -0300

    Bluetooth: Remove l2cap_sk_list
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>