If user space attempts to set the local name to the same value that's
already set we should simply return a direct command complete for this
mgmt command.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

This patch fixes the mgmt_set_local_name command to send the appropriate
HCI commands based on BR/EDR support and LE support. Local name and EIR
data should only be sent for BR/EDR capable controllers whereas an
update to the AD should only happen for LE capable controllers.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

For proper control of the AD update and the related HCI commands it's
best to run the AD update through an async request instead of a
standalone HCI command. This patch changes the hci_update_ad() function
to take a request pointer and updates its users appropriately. E.g. the
function is no longer called after the init sequence but during stage 3
of the init sequence.

The TX power is read during the init sequence, so we don't need an
explicit update whenever it is read and the AD update based on the local
name should be done through the local name mgmt handler. The only other
user is the update based on enabling advertising. This part is still
kept as there is no mgmt API to enable it.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

We shouldn't respond to the mgmt_set_local_name command until all
related HCI commands have completed. This patch fixes the issue by
running the local name HCI command and the EIR update in the same
asynchronous request, and returning the mgmt command complete through
the complete callback of the request.

The downside of this is that we must set hdev->dev_name before the local
name HCI command has completed since otherwise the generated EIR
command doesn't contain the new name. This means that we can no-longer
reliably detect when the name has really changed and when not. Luckily
this only affects scenarios where the mgmt interface is *not* used (e.g.
hciconfig) so redundant mgmt_ev_local_name_changed events in these cases
are an acceptable drawback.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

When powering off the device the hdev->flags and hdev->dev_flags need to
be cleared before calling mgmt_powered(). If this is not done the
resulting events sent to user space may contain incorrect values.

Note that the HCI_AUTO_OFF flag accessed right after this is part of the
persistent flags, so it's unchanged by the hdev->dev_flags reset.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

When hci_dev_do_close() is called we should make sure to clear all
non-persistent flags in hci->dev_flags.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

We'll need to use this mask also when powering off the HCI device
so it's better to have this in a single and visible place.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

Now that class related operations are tracked through asynchronous HCI
requests this flag is no longer needed.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

We should only return a mgmt command complete once all HCI commands to a
mgmt_set_dev_class or mgmt_add/remove_uuid command have completed. This
patch fixes the issue by having a proper async request complete callback
for these actions and responding to user space in the callback.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

The add/remove_uuid and set_dev_class mgmt commands can trigger both EIR
and class HCI commands, so testing just for a pending class command is
enough. The simplest way to monitor conflicts that should trigger "busy"
error returns is to check for any pending mgmt command that can trigger
these HCI commands. This patch adds a helper function for this
(pending_eir_or_class) and uses it instead of the old HCI_PENDING_CLASS
flag to test for busy conditions.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

We should only notify user space that the adapter has been powered on
after all HCI commands related to the action have completed. This patch
fixes the issue by instating an async request complete callback for
these HCI commands and only notifies user space in the callback.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

This patch updates sending of HCI commands related to mgmt_set_powered
(e.g. class, name and EIR data) to be sent using asynchronous requests.
This is necessary since it's the only (well, at least the cleanest) way
to keep the power on procedure synchronized and let user space know it
has completed only when all HCI commands are completed (this actual fix
is coming in a subsequent patch).
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

These commands will in a subsequent patch be performed in their own
asynchronous request, so it's more readable (not just from a resulting
code perspective but also the way the patches look like) to have them
performed in their own function.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

Replaced calls to kzalloc followed by memcpy with a single call to kmemdup.
Signed-off-by: NAlexandru Gheorghiu <gheorghiuandru@gmail.com>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

Used PTR_RET function instead of IS_ERR and PTR_ERR.
Patch found using coccinelle.
Signed-off-by: NAlexandru Gheorghiu <gheorghiuandru@gmail.com>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

The opcode in cmd_complete and cmd_status events is 16 bits, so we
should only be comparing it after having converted it to the host
endianness. There's already an opcode variable in both functions which
is in host endiannes so the right fix is to just start using it instead
of ev->opcode.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

If req->err is set, there is no point in queueing the HCI command
in HCI request command queue since it won't be sent anyway.
Signed-off-by: NAndre Guedes <andre.guedes@openbossa.org>
Acked-by: NJohan Hedberg <johan.hedberg@intel.com>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

Since no one checks the returning value of hci_req_add and HCI
request errors are now handled in hci_req_run, we can make hci_
req_add returning void.
Signed-off-by: NAndre Guedes <andre.guedes@openbossa.org>
Acked-by: NJohan Hedberg <johan.hedberg@intel.com>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

When we are building a HCI request with more than one HCI command
and one of the hci_req_add calls fail, we should have some cleanup
routine so the HCI commands already queued on HCI request can be
deleted. Otherwise, we will face some memory leaks issues.

This patch implements the HCI request error handling which is the
following: If a hci_req_add fails, we save the error code in hci_
request. Once hci_req_run is called, we verify the error field. If
it is different from zero, we delete all HCI commands already queued
and return the error code.
Signed-off-by: NAndre Guedes <andre.guedes@openbossa.org>
Acked-by: NJohan Hedberg <johan.hedberg@intel.com>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

Since hci_req_run will be returning more than one error code, we
should check its returning value in __hci_req_sync.
Signed-off-by: NAndre Guedes <andre.guedes@openbossa.org>
Acked-by: NJohan Hedberg <johan.hedberg@intel.com>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

In case the HCI request queue is empty, hci_req_run should return
ENODATA instead of EINVAL. This way, hci_req_run returns a more
meaningful error value.
Signed-off-by: NAndre Guedes <andre.guedes@openbossa.org>
Acked-by: NJohan Hedberg <johan.hedberg@intel.com>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

If hci_req_run returns error, we erroneously leave the current
process in TASK_INTERRUPTABLE state. If we leave the process in
TASK_INTERRUPTABLE and it is preempted, this process will never
be scheduled again.

This patch fixes this issue by moving the preparation for scheduling
(add to waitqueue and set process state) to just after the hci_req_run
call.
Signed-off-by: NAndre Guedes <andre.guedes@openbossa.org>
Acked-by: NJohan Hedberg <johan.hedberg@intel.com>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

With the removal of hci_req_complete() several HCI event handlers have
essentially become empty and can be removed. The only potential benefit
of these could have been logging, but the hci_event, hci_cmd_complete
and hci_cmd_status already provide a log for events which they do not
have an explicit handler for.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

This variable is no longer needed (due to async HCI request support and
the conversion of hci_req_sync to use it), so it can be safely removed.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

This patch converts the hci_req_sync() procedure to internaly use the
asynchronous HCI requests.

The hci_req_sync mechanism relies on hci_req_complete() calls from
hci_event.c into hci_core.c whenever a HCI command completes. This is
very similar to what asynchronous requests do and makes the conversion
fairly straight forward by converting hci_req_complete into a request
complete callback. By this change hci_req_complete (renamed to
hci_req_sync_complete) becomes private to hci_core.c and all calls to it
can be removed from hci_event.c.

The commands in each hci_req_sync procedure are collected into their own
request by passing the hci_request pointer to the request callback
(instead of the hci_dev pointer). The one slight exception is the HCI
init request which has the special handling of HCI driver specific
initialization commands. These commands are run in their own request
prior to the "main" init request.

One other extra change that this patch must contain is the handling of
spontaneous HCI reset complete events that some controllers exhibit.
These were previously handled in the hci_req_complete function but the
right place for them now becomes the hci_req_cmd_complete function.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

This patch introduces functions to process the HCI request state when
receiving HCI Command Status or Command Complete events. Some HCI
commands, like Inquiry do not result in a Command complete event so
special handling is needed for them. Inquiry is a particularly important
one since it is the only forseeable "non-cmd_complete" command that will
make good use of the request functionality, and its completion is either
indicated by an Inquiry Complete event of a successful Command Complete
for HCI_Inquiry_Cancel.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

To have a consistent content for hdev->cmd_q all entries need to follow
the semantics of asynchronous HCI requests. This means that even single
commands need to be dressed as requests by having a request start
indicator. This patch adds these indicators to the two places needing
it (hci_send_cmd and hci_sock_sendmsg).
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

This function is analogous to hci_send_cmd() but instead of directly
queuing the command to hdev->cmd_q it adds it to the local queue of the
asynchronous HCI request being build (inside struct hci_request).

This is the main function used for building asynchronous requests and
there should be one or more calls to it between calls to hci_req_init
and hci_req_run.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

This patch moves out the skb creation from hci_send_cmd() into its own
prepare_cmd() function. This is essential so the same prepare_cmd()
function can be easily reused for skb creation for asynchronous HCI
requests.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

This patch adds the initial definitions and functions for asynchronous
HCI requests. Asynchronous requests are essentially a group of HCI
commands together with an optional completion callback. The request is
tracked through the already existing command queue by having the
necessary context information as part of the control buffer of each skb.

The only information needed in the skb control buffer is a flag for
indicating that the skb is the start of a request as well as the
optional complete callback that should be used when the request is
complete (this will be found in the last skb of the request).
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

Having conditional command sending during a request has always been
problematic and caused hacks like the hdev->init_last_cmd variable. This
patch removes these conditionals and instead splits the init sequence
into three stages, each with its own __hci_req_sync() call.

This also paves the way to the upcoming asynchronous request support
swhich will also benefit by having a simpler implementation if it
doesn't need to cater for requests that change on the fly.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

If a request callback doesn't send any commands __hci_req_sync() should
fail imediately instead of waiting for the inevitable timeout to occur.
This is particularly important once we start creating requests with
conditional command sending which can potentially result in no commands
being sent at all.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

We'll be introducing an async version of hci_request. To make things
clear it makes sense to rename the existing API to have a _sync suffix.
Signed-off-by: NJohan Hedberg <johan.hedberg@intel.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

rfcomm_session_close() sets the RFCOMM session state to BT_CLOSED.
However, in multiple places immediately before the function is
called, the RFCOMM session is set to BT_CLOSED. Therefore,
remove these unnecessary state settings.
Signed-off-by: NDean Jenkins <Dean_Jenkins@mentor.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

In rfcomm_session_del() remove the redundant call to
rfcomm_send_disc() because it is not possible for the
session to be in BT_CONNECTED state during deletion
of the session.
Signed-off-by: NDean Jenkins <Dean_Jenkins@mentor.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

Previous commits have improved the handling of the RFCOMM session
timer and the RFCOMM session pointers such that freed RFCOMM
session structures should no longer be erroneously accessed. The
RFCOMM session refcnt now has no purpose and will be deleted by
this commit.

Note that the RFCOMM session is now deleted as soon as the
RFCOMM control channel link is no longer required. This makes the
lifetime of the RFCOMM session deterministic and absolute.
Previously with the refcnt, there was uncertainty about when
the session structure would be deleted because the relative
refcnt prevented the session structure from being deleted at will.

It was noted that the refcnt could malfunction under very heavy
real-time processor loading in embedded SMP environments. This
could cause premature RFCOMM session deletion or double session
deletion that could result in kernel crashes. Removal of the
refcnt prevents this issue.

There are 4 connection / disconnection RFCOMM session scenarios:
host initiated control link ---> host disconnected control link
host initiated ctrl link ---> remote device disconnected ctrl link
remote device initiated ctrl link ---> host disconnected ctrl link
remote device initiated ctrl link ---> remote device disc'ed ctrl link

The control channel connection procedures are independent of the
disconnection procedures. Strangely, the RFCOMM session refcnt was
applying special treatment so erroneously combining connection and
disconnection events. This commit fixes this issue by removing
some session code that used the "initiator" member of the session
structure that was intended for use with the data channels.
Signed-off-by: NDean Jenkins <Dean_Jenkins@mentor.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

Unfortunately, the design retains local copies of the s RFCOMM
session pointer in various code blocks and this invites the erroneous
access to a freed RFCOMM session structure.

Therefore, return the RFCOMM session pointer back up the call stack
to avoid accessing a freed RFCOMM session structure. When the RFCOMM
session is deleted, NULL is passed up the call stack.

If active DLCs exist when the rfcomm session is terminating,
avoid a memory leak of rfcomm_dlc structures by ensuring that
rfcomm_session_close() is used instead of rfcomm_session_del().
Signed-off-by: NDean Jenkins <Dean_Jenkins@mentor.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

A race condition exists between near simultaneous asynchronous
DLC data channel disconnection requests from the host and remote device.
This causes the socket layer to request a socket shutdown at the same
time the rfcomm core is processing the disconnect request from the remote
device.

The socket layer retains a copy of a struct rfcomm_dlc d pointer.
The d pointer refers to a copy of a struct rfcomm_session.
When the socket layer thread performs a socket shutdown, the thread
may wait on a rfcomm lock in rfcomm_dlc_close(). This means that
whilst the thread waits, the rfcomm_session and/or rfcomm_dlc structures
pointed to by d maybe freed due to rfcomm core handling. Consequently,
when the rfcomm lock becomes available and the thread runs, a
malfunction could occur as a freed rfcomm_session structure and/or a
freed rfcomm_dlc structure will be erroneously accessed.

Therefore, after the rfcomm lock is acquired, check that the struct
rfcomm_session is still valid by searching the rfcomm session list.
If the session is valid then validate the d pointer by searching the
rfcomm session list of active DLCs for the rfcomm_dlc structure
pointed by d.
Signed-off-by: NDean Jenkins <Dean_Jenkins@mentor.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

Use del_timer_sync() instead of del_timer() as this ensures
that rfcomm_session_timeout() is not running on a different
CPU when rfcomm_session_put() is called. This avoids a race
condition on SMP systems because potentially
rfcomm_session_timeout() could reuse the freed RFCOMM session
structure caused by the execution of rfcomm_session_put().

Note that this modification makes the reason for the RFCOMM
session refcnt mechanism redundant.
Signed-off-by: NDean Jenkins <Dean_Jenkins@mentor.com>
Acked-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>

There is no reason a caller ever wants to check the return type of this
call. _Iff_ a user successfully called bt_sock_register(), they're allowed
to call bt_sock_unregister().
All other calls in the kernel (device_del, device_unregister, kfree(), ..)
that are logically equivalent return void. Lets not make callers think
they have to check the return type of this call and instead simply return
void.

We guarantee that after bt_sock_unregister() is called, the socket type
_is_ unregistered. If that is not what the caller wants, they're using the
wrong function, anyway.
Signed-off-by: NDavid Herrmann <dh.herrmann@gmail.com>
Signed-off-by: NGustavo Padovan <gustavo.padovan@collabora.co.uk>