Commit Graph

6822 Commits

Author SHA1 Message Date
Jose Antonio Santos Cadenas
0edbc77755 Remove extra spaces from tools dir 2010-05-07 01:48:03 +03:00
Jose Antonio Santos Cadenas
6968c52091 Remove extra spaces in test dir 2010-05-07 01:47:59 +03:00
Jose Antonio Santos Cadenas
8c83ffe7c3 Remove extra spaces from sdpd-request.c 2010-05-07 01:47:56 +03:00
Jose Antonio Santos Cadenas
078185a230 Remove extra spaces at end of line in lib dir 2010-05-07 01:47:54 +03:00
Jose Antonio Santos Cadenas
588f659f3f Remove extra spaces at end of line in compat dir 2010-05-07 01:47:50 +03:00
Johan Hedberg
800e8fa257 Fix parent path introspection data invalidation for multiple levels
In the case that parent path data needs to be invalidated we shouldn't
stop at the immediate parent if it doesn't have our own handler
registered but should continue upwards in the tree until we reach root
or our own handler.
2010-04-30 01:23:45 +03:00
Johan Hedberg
9bbcbff039 Fix memory leak in g_dbus_register_interface 2010-04-30 00:27:54 +03:00
Marcel Holtmann
9e379872b0 Release 4.64 2010-04-29 00:21:20 +02:00
Marcel Holtmann
16f0eeda60 Add missing author credits 2010-04-29 00:20:20 +02:00
Marcel Holtmann
cf0913d056 Update library version 2010-04-29 00:10:50 +02:00
Johan Hedberg
d22b2062cf Fix scope bdaddr variable to not be larger than necessary 2010-04-28 22:58:56 +03:00
Elvis Pfutzenreuter
25c753748b Update service class on adapter-by-adapter basis
When a service is added (or removed) for a specific adapter, only the
service class of that specific adapter shall be updated.
2010-04-28 22:52:42 +03:00
José Antonio Santos Cadenas
1d1154156d Fix list parsing bug in sdp_set_supp_features
When the data is a string or a sequence, it is not ok to dereference
data->val because it is already a pointer. Additionally a lengths
variable is added because the strings are not terminated with '\0' and
otherwise it would not be possible to know their length.
2010-04-28 22:47:59 +03:00
Gustavo F. Padovan
5261de27f3 Set l2cap mode always
As ERTM can be the default mode with SOCK_STREAM we have to set
opts.mode always.
2010-04-28 18:47:35 +03:00
Gustavo F. Padovan
449367e8da Add options to test MaxTx and TxWindow 2010-04-28 18:47:35 +03:00
Gustavo F. Padovan
d006c7abe4 Add txWindow and MaxTt to l2cap_options 2010-04-28 18:47:35 +03:00
Gustavo F. Padovan
a43d90eeb9 Add duplex test case to l2test
Use a SOCK_NONBLOCK to send and receive at the same time.
2010-04-28 18:47:35 +03:00
Gustavo F. Padovan
c5c6abdf4c l2test: add option to use SOCK_STREAM 2010-04-28 18:47:35 +03:00
Marcel Holtmann
bd3ebad37b Make some UUID parameters const 2010-04-28 10:40:48 +02:00
Marcel Holtmann
d108edd5cd Remove another bunch of pointless casting attempts 2010-04-28 10:36:55 +02:00
Santiago Carot-Nemesio
114cb3f810 Add memory checks whenever memory is allocated in SDP
This patch checks memory allocated with malloc and removes uneccessary
(**void) castings.
2010-04-28 10:37:24 +03:00
Santiago Carot Nemesio
f58df2db7a Use NULL instead of 0 for functions returning a pointer 2010-04-27 17:56:40 +03:00
Santiago Carot Nemesio
400772d67b Remove some blanks spaces in sdp source code. 2010-04-27 17:56:14 +03:00
Luiz Augusto Von Dentz
48a2c31cdb Fix Device.SetProperty(Alias,... always emit PropertyChanged
Similar to 6da6e4e3de but for Alias
Property.
2010-04-22 11:37:44 +03:00
Luiz Augusto Von Dentz
6da6e4e3de Fix Device.SetProperty("Trusted",... always emit PropertyChanged
Currently any change done to Trusted property generates a PropertyChanged
since no verification is done prior to apply the new value which may
generate an unnecessary write to the storage and extra D-Bus traffic.

To fix this trusted is now loaded from storage and cached in memory when
the device is created, which latter is used to make sure the value has
really changed avoiding storage reads to check current value.
2010-04-22 11:35:45 +03:00
Marcel Holtmann
fa35778e2f Add some more company identifiers 2010-04-14 04:35:49 -07:00
Francisco Alecrim
86190ea09c List with UUIDs per-adapter should be sorted 2010-04-12 01:43:34 +03:00
Johan Hedberg
78c8da0521 Set HID socket security back to LOW after the BtIO default change
At least mice require a lower security level, however for now this patch
changes the level back to low for all HID sockets.
2010-04-12 01:35:19 +03:00
Daniel Orstadius
32c43b9556 Return error properly on CreateDevice
Adds a check in function search_cb for whether the initiating DBus
call was CreateDevice or CreatePairedDevice. Previously both calls
triggered a positive reply regardless of the error code, but in case
of CreateDevice any error should be propagated in the DBus reply.

Modifies the store_profiles function to return immediately for a
temporary device.
2010-04-09 18:11:00 +03:00
Daniel Orstadius
be35d00671 Handle connection ending events on higher layers first
When sockets are notified simultaneously of a disconnection or
connection failure, the sockets on the higher levels should be allowed
to handle the disconnect event before cleaning up the connection at
sockets on the lower level (HCI).

The issue behind creating the patch was that on a failed connection
attempt hcid_dbus_conn_complete could get called before the SDP
callbacks, which lead to the latter not being called.

Lowers the priority of the HCI socket to trigger it after higher level
sockets such as L2CAP and RFCOMM.
2010-04-09 18:06:24 +03:00
Johan Hedberg
4d3a817a4c Fix minor indentation issues 2010-04-07 20:31:31 +03:00
Francisco Alecrim
a05689dde7 Correct UUIDs list to EIR
Correct create_ext_inquiry_response to response only UUIDs per-adapter.
2010-04-07 20:31:27 +03:00
Francisco Alecrim
9266dfeac2 Report local services(UUIDs) through D-Bus
* Include UUIDs field to method GetProperties(org.bluez.Adapter).
Applications can get local services(UUIDs) available through D-Bus.
* UUIDs per-adapter stored at btd_adapter to prevent some searches not
necessary regarding it requires information from access_db and service_db.
* Emit Adapter.PropertyChanged signal when UUIDs change.
2010-04-07 20:31:03 +03:00
Luiz Augusto Von Dentz
b5eb420ebf Fix maemo6 API changes since maemo5
This should cover the changes in D-Bus API of csnet and simpb since
maemo5.
2010-04-06 12:29:53 +03:00
Luiz Augusto Von Dentz
e3300f188a Introduce maemo6 telephony driver
Copy from maemo5 driver to accomodate API changes done in maemo6.
2010-04-06 12:29:27 +03:00
Johan Hedberg
2499fb1f17 Fix device object reference counting with SDP browse requests
The req->device pointer should be properly reference counted in order to
avoid invalid memory access.
2010-04-03 20:00:17 +03:00
Johan Hedberg
903804a0f1 Make use of common fail label in pnat_probe 2010-04-03 16:45:44 +03:00
Johan Hedberg
76208a5d26 Fix GError memory leak 2010-04-03 16:37:26 +03:00
Johan Hedberg
0a658cd945 Remove bogus if-check when DUN client disconnects 2010-04-03 16:32:36 +03:00
Johan Hedberg
23f75004d8 Fix invalid memory access in headset_get_nrec when hs->slc is NULL
headset_get_nrec might get called when there's no SLC and it should
return the default NREC value in that case which is TRUE.
2010-04-02 22:41:53 +03:00
Johan Hedberg
895eea1347 Add initial phonet integration with the help of pnatd/phonet-at
This patch implements initial phonet support to the pnat bluetoothd
plugin by utilizing the pnatd/phonet-at phonet AT proxy/wrapper tool.
2010-04-01 15:58:23 +03:00
Johan Hedberg
d8f4711922 Add possibility to explicitly specify source/dest channel in BtIO
This is needed e.g. to create RFCOMM TTY's from incoming connections
since in that case the remote (clients) RFCOMM channel needs to be known
for the RFCOMMCREATEDEV ioctl.
2010-03-31 14:40:51 +03:00
Johan Hedberg
aee26b30bb Make BtIO default security level MEDIUM
MEDIUM makes more sense than the kernel default LOW which maps to "no
bonding" with SSP (something that's useful only for very special cases
such as OPP).
2010-03-30 13:36:46 +03:00
Forrest Zhao
c5321032b0 Return error when modem_obj_path is NULL
This could prevent crash in case modem_obj_path is NULL.
2010-03-29 11:11:13 +03:00
Marcel Holtmann
8dca8bf737 Release 4.63 2010-03-25 18:34:44 -07:00
Forrest Zhao
ffcc20892f Set voice_call.watch in correct way
It was set wrongly to device_watch.
2010-03-25 12:05:46 +02:00
Luiz Augusto Von Dentz
ab816e6549 Add immediate parameter to avdtp_close
When TRUE if there is an outstanding command in the sending queue it will
cause abort which cancel any pending command.
2010-03-24 23:32:24 +02:00
Luiz Augusto Von Dentz
f8d790861c Send abort on while cancelling a2dp setup
This should guarantee that any outstanding command in the queue is
cancelled, accourding to the avdtp spec:

"6.16 Abort
Because AVDTP is typically used on unreliable channels, signaling messages
can be lost due to L2CAP Flush Timeouts. To react to inconsistencies
between the INT and ACP states, the Abort Command may be used."

It also says abort can be send regardless of the state even in case of
packet fragmantation:

"Note that it is permitted to breakdown a packet sequence (command or
response message) to send an Abort command message in the same direction.
The remaining packets consecutive to an aborted message shall be flushed by
the sender."
2010-03-24 23:32:24 +02:00
Luiz Augusto Von Dentz
abe8bb2d3b Make abort requests to have a shorter timeout
Regular request takes 4 secs to timeout, but this may be too long to wait
for abort command so it is now reduced to 2 secs.
2010-03-24 23:32:24 +02:00
Luiz Augusto Von Dentz
43830baea4 Fix avdtp_abort not canceling pending requests 2010-03-24 23:32:24 +02:00