summaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)AuthorFilesLines
2022-02-28Media control enablement and bug fixesricefish_18.0.2ricefish_18.0.1ricefish_18.0.0ricefish/18.0.2ricefish/18.0.1ricefish/18.0.0quillback_17.1.2quillback_17.1.1quillback_17.1.0quillback_17.0.2quillback_17.0.1quillback_17.0.0quillback/17.1.2quillback/17.1.1quillback/17.1.0quillback/17.0.2quillback/17.0.1quillback/17.0.0needlefish_13.93.0needlefish/13.93.0marlin_12.93.0marlin/12.93.0lamprey_12.1.6lamprey_12.1.5lamprey_12.1.4lamprey_12.1.3lamprey_12.1.2lamprey/12.1.6lamprey/12.1.5lamprey/12.1.4lamprey/12.1.3lamprey/12.1.2devtool-patcheddevtool-base18.0.218.0.118.0.017.1.217.1.117.1.017.0.217.0.117.0.013.93.012.93.012.1.612.1.512.1.412.1.312.1.2Scott Murray4-10/+149
Changes: - Properly split out media control and transport callbacks, working on an implementation in a client pointed out some issues that needed to be fixed. - Add new exported API functions bluez_get_media_control_properties and bluez_get_media_player_properties to simplify the ability of clients to determine the current state. This is more required now for the usecase of standalone clients starting up separately as opposed to having a binding tracking the state for them. - Added logic in bluez_devices_signal_callback to be paranoid about the input to the callback, as it became clear from debugging that having two different libraries use g_dbus_connection_signal_subscribe in the same process results in the separately registered callbacks all getting called with the superset of the filtered signal requests. At best this leads to confusion with respect to logging/debugging, at worst it opens the door to tripping over unexpected inputs in the callbacks. - Fixed the device path being passed to bluez_autoconnect, this seems to re-enable the workaround wrt devices disconnecting after the initial connect. Further investigation is required into what happens here, but the behavior now seems to match the pre-refactored code in the binding. - Added a couple of DEBUG statements to allow vetting the arguments passed to g_dbus_connection_call in a couple of places. - Fix install of the test program if building it is enabled with the meson option. Bug-AGL: SPEC-4231 Signed-off-by: Scott Murray <scott.murray@konsulko.com> Change-Id: Iba5bb6e4bde3a9f4172d11b9eefdddf8c5da2203
2022-01-14Avoid init delay without BT adapter presentmarlin_12.92.0marlin/12.92.0lamprey_12.1.1lamprey/12.1.112.92.012.1.1Scott Murray1-3/+3
One of the changes in my refactoring out of the binding code resulted in the asynchronous code path for BT adapter detection accidentally becoming synchronous, resulting in a significant delay in bluez_init calls when no adapter is present. Fix this by tweaking the signalling of init status from the handler thread to bluez_init to better match what was being done in the old codebase. Bug-AGL: SPEC-4216 Signed-off-by: Scott Murray <scott.murray@konsulko.com> Change-Id: Ib58201eb9549ebcb3bc835809f7c4df47a87f45c
2021-12-20Initial check-inmarlin_12.91.0marlin/12.91.012.91.0Scott Murray19-0/+3672
GLib based interface library for BlueZ, factored out of the agl-service-bluetooth binding code. See README.md for build and usage notes and the mapping of the new source files to those in the binding if that is for some reason required. Currently untested functionality: * autoconnect on startup * mediaplayer controls Bug-AGL: SPEC-4182 Signed-off-by: Scott Murray <scott.murray@konsulko.com> Change-Id: I7e70962ebabb138f81b9ba69af82f4ca0152dc31
2021-12-18Initial empty repositoryJan-Simon Moeller0-0/+0