As HCI rx path is now done in process context it makes sense to do all the
timer in process context as well.
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Now we run everything in HCI in process context, so it's a better idea use
mutex instead spin_lock. The macro remains hci_dev_lock() (and I got rid
of hci_dev_lock_bh()), of course.
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Run recv process in workqueue helps a lot with our processing as the recv
path will also be in the process context, i.e., now all our tx and rx are
in process context.
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Only the list member of the struct was used, so we now fold it into
hci_conn.
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Remove old code not touched for several years.
Signed-off-by: NAndrei Emeltchenko <andrei.emeltchenko@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

This patch adds the dev_flags field to struct hci_dev. This new
flags variable should be used to define flags related to BR/EDR
and/or LE controller itself. It should be used to define flags
which represents states from the controller. The dev_flags is
cleared in case the controller sends a Reset Command Complete
Event to the host.

Also, this patch adds the HCI_LE_SCAN flag which was created to
track if the controller is performing LE scan or not. The flag
is set/cleared when the controller starts/stops scanning.

This is an initial effort to stop using hdev->flags to define
internal flags since it is exported to userspace by an ioctl.
Signed-off-by: NAndre Guedes <andre.guedes@openbossa.org>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Use queue instead of stack discipline for device list. When processing
dev_list with list_for_each* devices will be prosessed in order they
were added (Usually BR/EDR first and AMP later).

Also output from hciconfig looks nicer :-)
Signed-off-by: NAndrei Emeltchenko <andrei.emeltchenko@intel.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Fixes bluetooth compiling when CONFIG_BT_L2CAP is not enabled

net/built-in.o: In function `hci_dev_open':
(.text+0xdce9a): undefined reference to `enable_hs'
Reported-by: NRandy Dunlap <rdunlap@xenotime.net>
Signed-off-by: NAndrei Emeltchenko <andrei.emeltchenko@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Signed-off-by: NSzymon Janc <szymon.janc@tieto.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Signed-off-by: NAndrei Emeltchenko <andrei.emeltchenko@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

The cancel_delayed_work_sync function should not be used if we hold any
locks. Luckily all places where this is the case it is also safe to use
the non-sync version.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Now that the pending commands are within struct hci_dev we can properly
control access to them throught the hci_dev locking mechanism.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

This patch moves the pending management command list (previously global
to mgmt.c) into struct hci_dev. This makes it possible to do proper
locking when accessing it (through the existing hci_dev locks) and
thereby avoid race conditions.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

The current global pending command list in mgmt.c is racy. Possibly the
simplest way to fix it is to have per-hci dev lists instead of a global
one (all commands that need a pending struct are hci_dev specific).
This way the list can be protected using the already existing per-hci
dev lock. To enable this refactoring the first thing that needs to be
done is to ensure that the mgmt functions have access to the hci_dev
struct (instead of just the dev id).
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

BR/EDR link keys have their own commands and events (separate from SMP)
and the remove_keys command (previously remove_key) removes keys of any
kind for the specified remote address.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

The power off code doesn't need to use its own custom timer since the
delayed_work API provides the exact same functionality.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Based on the revised mgmt API set_discoverable has a timeout parameter
to specify how long the adapter will remain discoverable. A value of 0
means "indefinitively".
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

This patch adds a function to hci_core to cancel an ongoing inquiry.

According to the Bluetooth spec, the inquiry cancel command should
only be issued after the inquiry command has been issued, a command
status event has been received for the inquiry command, and before
the inquiry complete event occurs.

As HCI_INQUIRY flag is only set just after an inquiry command status
event occurs and it is cleared just after an inquiry complete event
occurs, the inquiry cancel command should be issued only if HCI_INQUIRY
flag is set.

Additionally, cancel inquiry related code from stop_discovery() were
replaced by a hci_cancel_inquiry() call.
Signed-off-by: NAndre Guedes <andre.guedes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

This patch adds a function to hci_core to carry out inquiry.

All inquiry code from start_discovery() were replaced by a
hci_do_inquiry() call.
Signed-off-by: NAndre Guedes <andre.guedes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

It's convenient to use the HCI device index the AMP controller id, but
the spec requires that an AMP controller never has id 0.
Signed-off-by: NMat Martineau <mathewm@codeaurora.org>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Acked-by: NAndrei Emeltchenko <andrei.emeltchenko@intel.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

To avoid starvation the priority is recalculated so that the starving
channels are promoted to HCI_PRIO_MAX - 1 (6).

HCI_PRIO_MAX (7) is considered special, because it requires CAP_NET_ADMIN
capability which can be used to provide more guaranties, so it is not used
when promoting.
Signed-off-by: NLuiz Augusto von Dentz <luiz.von.dentz@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

The quote is calculated based on the first buffer in the queue so if the
priority changes to something lower than the priority of the first skb
the quote needs to be recalculated.
Signed-off-by: NLuiz Augusto von Dentz <luiz.von.dentz@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

This implement priority based scheduler using skbuffer priority set via
SO_PRIORITY socket option.

It introduces hci_chan_hash (list of HCI Channel/hci_chan) per connection,
each item in this list refer to a L2CAP connection and it is used to
queue the data for transmission.
Signed-off-by: NLuiz Augusto von Dentz <luiz.von.dentz@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

I've noticed that my CSR usb dongle was not working if it was plugged in when
PC was booting. It looks like I get two HCI reset command complete events (see
hcidump logs below).
The root cause is reset called from off_timer. Timeout for this reset to
complete is set to 250ms and my bt dongle requires more time for replying with
command complete event. After that, chip seems to reply with reset command
complete event for next non-reset command.

Attached patch increase mentioned timeout to HCI_INIT_TIMEOUT, this value is
already used for timeouting hci_reset_req in hci_dev_reset().

This might also be related to BT not working after suspend that was reported
here some time ago.

Hcidump log:

2011-09-12 23:13:27.379465 < HCI Command: Reset (0x03|0x0003) plen 0
2011-09-12 23:13:27.380797 > HCI Event: Command Complete (0x0e) plen 4
    Reset (0x03|0x0003) ncmd 1
    status 0x00
2011-09-12 23:13:27.380859 < HCI Command: Read Local Supported Features (0x04|0x000
3) plen 0
2011-09-12 23:13:27.760789 > HCI Event: Command Complete (0x0e) plen 4
    Reset (0x03|0x0003) ncmd 1
    status 0x00
2011-09-12 23:13:27.760831 < HCI Command: Read Local Version Information (0x04|0x00
01) plen 0
2011-09-12 23:13:27.764780 > HCI Event: Command Complete (0x0e) plen 12
    Read Local Version Information (0x04|0x0001) ncmd 1
    status 0x00
    HCI Version: 1.1 (0x1) HCI Revision: 0x36f
    LMP Version: 1.1 (0x1) LMP Subversion: 0x36f
    Manufacturer: Cambridge Silicon Radio (10)
Signed-off-by: NSzymon Janc <szymon@janc.net.pl>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

When all items in the list have the same type there is no much of a point
to use list_for_each except if you want to use the list pointer itself.
Signed-off-by: NLuiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

hci_unregister_dev cannot fail and always returns 0. The drivers already ignore
the return value so we can safely make it return void.
Signed-off-by: NDavid Herrmann <dh.herrmann@googlemail.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

I've noticed that my CSR usb dongle was not working if it was plugged in when
PC was booting. It looks like I get two HCI reset command complete events (see
hcidump logs below).
The root cause is reset called from off_timer. Timeout for this reset to
complete is set to 250ms and my bt dongle requires more time for replying with
command complete event. After that, chip seems to reply with reset command
complete event for next non-reset command.

Attached patch increase mentioned timeout to HCI_INIT_TIMEOUT, this value is
already used for timeouting hci_reset_req in hci_dev_reset().

This might also be related to BT not working after suspend that was reported
here some time ago.

Hcidump log:

2011-09-12 23:13:27.379465 < HCI Command: Reset (0x03|0x0003) plen 0
2011-09-12 23:13:27.380797 > HCI Event: Command Complete (0x0e) plen 4
    Reset (0x03|0x0003) ncmd 1
    status 0x00
2011-09-12 23:13:27.380859 < HCI Command: Read Local Supported Features (0x04|0x000
3) plen 0
2011-09-12 23:13:27.760789 > HCI Event: Command Complete (0x0e) plen 4
    Reset (0x03|0x0003) ncmd 1
    status 0x00
2011-09-12 23:13:27.760831 < HCI Command: Read Local Version Information (0x04|0x00
01) plen 0
2011-09-12 23:13:27.764780 > HCI Event: Command Complete (0x0e) plen 12
    Read Local Version Information (0x04|0x0001) ncmd 1
    status 0x00
    HCI Version: 1.1 (0x1) HCI Revision: 0x36f
    LMP Version: 1.1 (0x1) LMP Subversion: 0x36f
    Manufacturer: Cambridge Silicon Radio (10)
Signed-off-by: NSzymon Janc <szymon@janc.net.pl>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

We need to catch errors when calling hci_add_sysfs() and return them to
the caller to avoid kernel oopses on device_add() failure.
Signed-off-by: NDavid Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

As we introduced hci_init_sysfs() we should also rename
hci_register_sysfs() and hci_unregister_sysfs() to hci_add_sysfs() and
hci_del_sysfs() like we do with hci_conn_add/del_sysfs(). It looks more
consistent now.
Signed-off-by: NDavid Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

We must not call device_del() if we didn't use device_add(). See module.c
for comments on that. Therefore, we need to call device_initialize() when
allocating the hci device and later device_add() instead of
device_register().

This also fixes a bug when hci_register_dev() failed and we call
hci_free_dev() without a valid core device. hci_free_dev() segfaults while
calling put_device() on invalid memory.

We already do this with hci_conn connections (hci_conn_init_sysfs()) so
they do not need to be fixed.
Signed-off-by: NDavid Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Add management interface events for blocking/unblocking a device.
Sender of the block device command gets cmd complete and other
mgmt sockets get the event. Event is also sent to mgmt sockets when
blocking is done with ioctl, e.g when blocking a device with
hciconfig. This makes it possible for bluetoothd to track status
of blocked devices when a third party block or unblocks a device.

Event sending is handled in mgmt_device_blocked function which gets
called from hci_blacklist_add in hci_core.c. A pending command is
added in mgmt_block_device, so that it can found when sending the
event - the event is not sent to the socket from which the pending
command came. Locks were moved out from hci_core.c to hci_sock.c
and mgmt.c, because locking is needed also for mgmt_pending_add in
mgmt.c.
Signed-off-by: NAntti Julku <antti.julku@nokia.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

The objective is to make the core to have as little as possible
information about SMP procedures and logic. Now, all the SMP
specific information is hidden from the core.
Signed-off-by: NVinicius Costa Gomes <vinicius.gomes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

This checks if there is any existing connection according to its type
before start iterating in the list and immediately stop iterating when
reaching the number of connections.
Signed-off-by: NLuiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

No command should be send before Command Complete event for HCI
reset is received. This fix regression introduced by commit
6bd32326(Bluetooth: Use proper timer for hci command timout)
for chips whose reset command takes longer to complete (e.g. CSR)
resulting in next command being send before HCI reset completed.
Signed-off-by: NSzymon Janc <szymon@janc.net.pl>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

There was a small typo here so we never actually hit the goto which
would call hci_dev_unlock_bh().
Signed-off-by: NDan Carpenter <error27@gmail.com>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

In some cases it will be useful having the key size used for
encrypting the link. For example, some profiles may restrict
some operations depending on the key length.

The key size is stored in the key that is passed to userspace
using the pin_length field in the key structure.

For now this field is only valid for LE controllers. 3.0+HS
controllers define the Read Encryption Key Size command, this
field is intended for storing the value returned by that
command.
Signed-off-by: NVinicius Costa Gomes <vinicius.gomes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

As the LTK (the new type of key being handled now) has more data
associated with it, we need to store this extra data and retrieve
the keys based on that data.

Methods for searching for a key and for adding a new LTK are
introduced here.
Signed-off-by: NVinicius Costa Gomes <vinicius.gomes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Since hdev->lock may be acquired by threads runnning in interrupt
context, all threads running in process context should disable
local bottom halve before locking hdev->lock. This can be done by
using hci_dev_lock_bh macro.

This way, we avoid potencial deadlocks like this one reported by
CONFIG_PROVE_LOCKING=y.

[  304.788780] =================================
[  304.789686] [ INFO: inconsistent lock state ]
[  304.789686] 2.6.39+ #1
[  304.789686] ---------------------------------
[  304.789686] inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage.
[  304.789686] ksoftirqd/0/3 [HC0[0]:SC1[1]:HE1:SE0] takes:
[  304.789686]  (&(&hdev->lock)->rlock){+.?...}, at: [<ffffffffa000bbfe>] hci_conn_check_pending+0x38/0x76 [bluetooth]
[  304.789686] {SOFTIRQ-ON-W} state was registered at:
[  304.789686]   [<ffffffff8105188b>] __lock_acquire+0x347/0xd52
[  304.789686]   [<ffffffff810526ac>] lock_acquire+0x8a/0xa7
[  304.789686]   [<ffffffff812b3758>] _raw_spin_lock+0x2c/0x3b
[  304.789686]   [<ffffffffa0009cf0>] hci_blacklist_del+0x1f/0x8a [bluetooth]
[  304.789686]   [<ffffffffa00139fd>] hci_sock_ioctl+0x2d9/0x314 [bluetooth]
[  304.789686]   [<ffffffff812197d8>] sock_ioctl+0x1f2/0x214
[  304.789686]   [<ffffffff810b0fd6>] do_vfs_ioctl+0x46c/0x4ad
[  304.789686]   [<ffffffff810b1059>] sys_ioctl+0x42/0x65
[  304.789686]   [<ffffffff812b4892>] system_call_fastpath+0x16/0x1b
[  304.789686] irq event stamp: 9768
[  304.789686] hardirqs last  enabled at (9768): [<ffffffff812b40d4>] restore_args+0x0/0x30
[  304.789686] hardirqs last disabled at (9767): [<ffffffff812b3f6a>] save_args+0x6a/0x70
[  304.789686] softirqs last  enabled at (9726): [<ffffffff8102fa9b>] __do_softirq+0x129/0x13f
[  304.789686] softirqs last disabled at (9739): [<ffffffff8102fb33>] run_ksoftirqd+0x82/0x133
[  304.789686]
[  304.789686] other info that might help us debug this:
[  304.789686]  Possible unsafe locking scenario:
[  304.789686]
[  304.789686]        CPU0
[  304.789686]        ----
[  304.789686]   lock(&(&hdev->lock)->rlock);
[  304.789686]   <Interrupt>
[  304.789686]     lock(&(&hdev->lock)->rlock);
[  304.789686]
[  304.789686]  *** DEADLOCK ***
[  304.789686]
[  304.789686] 1 lock held by ksoftirqd/0/3:
[  304.789686]  #0:  (hci_task_lock){++.-..}, at: [<ffffffffa0008353>] hci_rx_task+0x49/0x2f3 [bluetooth]
[  304.789686]
[  304.789686] stack backtrace:
[  304.789686] Pid: 3, comm: ksoftirqd/0 Not tainted 2.6.39+ #1
[  304.789686] Call Trace:
[  304.789686]  [<ffffffff812ae901>] print_usage_bug+0x1e7/0x1f8
[  304.789686]  [<ffffffff8100a796>] ? save_stack_trace+0x27/0x44
[  304.789686]  [<ffffffff8104fc3f>] ? print_irq_inversion_bug.part.26+0x19a/0x19a
[  304.789686]  [<ffffffff810504bb>] mark_lock+0x106/0x258
[  304.789686]  [<ffffffff812b40d4>] ? retint_restore_args+0x13/0x13
[  304.789686]  [<ffffffff81051817>] __lock_acquire+0x2d3/0xd52
[  304.789686]  [<ffffffff8102be73>] ? vprintk+0x3ab/0x3d7
[  304.789686]  [<ffffffff812ae126>] ? printk+0x3c/0x3e
[  304.789686]  [<ffffffff810526ac>] lock_acquire+0x8a/0xa7
[  304.789686]  [<ffffffffa000bbfe>] ? hci_conn_check_pending+0x38/0x76 [bluetooth]
[  304.789686]  [<ffffffff811601c6>] ? __dynamic_pr_debug+0x10c/0x11a
[  304.789686]  [<ffffffff812b3758>] _raw_spin_lock+0x2c/0x3b
[  304.789686]  [<ffffffffa000bbfe>] ? hci_conn_check_pending+0x38/0x76 [bluetooth]
[  304.789686]  [<ffffffffa000bbfe>] hci_conn_check_pending+0x38/0x76 [bluetooth]
[  304.789686]  [<ffffffffa000c561>] hci_event_packet+0x38e/0x3e12 [bluetooth]
[  304.789686]  [<ffffffff81052615>] ? lock_release+0x16c/0x179
[  304.789686]  [<ffffffff812b3b41>] ? _raw_read_unlock+0x23/0x27
[  304.789686]  [<ffffffffa0013e7f>] ? hci_send_to_sock+0x179/0x188 [bluetooth]
[  304.789686]  [<ffffffffa00083d2>] hci_rx_task+0xc8/0x2f3 [bluetooth]
[  304.789686]  [<ffffffff8102f5a9>] tasklet_action+0x87/0xe6
[  304.789686]  [<ffffffff8102fa11>] __do_softirq+0x9f/0x13f
[  304.789686]  [<ffffffff8102fb33>] run_ksoftirqd+0x82/0x133
[  304.789686]  [<ffffffff8102fab1>] ? __do_softirq+0x13f/0x13f
[  304.789686]  [<ffffffff81040f0a>] kthread+0x7f/0x87
[  304.789686]  [<ffffffff812b55c4>] kernel_thread_helper+0x4/0x10
[  304.789686]  [<ffffffff812b40d4>] ? retint_restore_args+0x13/0x13
[  304.789686]  [<ffffffff81040e8b>] ? __init_kthread_worker+0x53/0x53
[  304.789686]  [<ffffffff812b55c0>] ? gs_change+0x13/0x13
Signed-off-by: NAndre Guedes <andre.guedes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

The enable_smp parameter is no longer needed. It can be replaced by
checking lmp_host_le_capable.
Signed-off-by: NAndre Guedes <andre.guedes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>

Since we have the extended LMP features properly implemented, we
should check the LMP_HOST_LE bit to know if the host supports LE.
Signed-off-by: NAndre Guedes <andre.guedes@openbossa.org>
Signed-off-by: NGustavo F. Padovan <padovan@profusion.mobi>