Don't remove effects until the session they are in goes away or all
AudioEffects have been explicitly released. This allows the control
panel process to die without stopping the effects.
Change-Id: I4496e5df080230ca1af149dec95c1309ab8ea888
the first time a surface was connected, the values returned
by query NATIVE_WINDOW_DEFAULT_{WIDTH|HEIGHT} and
NATIVE_WINDOW_TRANSFORM_HINT were wrong until a call
to queueBuffer was performed.
Bug: 5137366, 5121607
Change-Id: I7ac6b5b0daa876638f6bed7c20f286a6e6d984f6
Update to API for a client to notify what type of information
changed (e.g. playstate, metadata) to optimize small updates
without the remote control display having to do expensive
queries such as album art.
When the remote control display retrieves the flags about
what information changed, the flag gets cleared.
Change-Id: I7d3d8d3eecd1da44695d84905ed9e7b70fe38b86
The media scanner was crashing while tring to insert genre data in the internal
media database, which does not support genres.
This resulted in some or all of the built in ringtones being unavailable
Bug: 5124718
Change-Id: Ib306348dd9808e59477b327b4adcb413957d8d24
Signed-off-by: Mike Lockwood <lockwood@android.com>
The way the old way worked relied on the process retrieving the
current remote control client to be in the exact same process as
the AudioService. This removes that dependency.
Change-Id: I1ba8bf32b61ec1e979ef7eee9661ba801aa19690
- Atribute network activity to uid calling the mediaplayer
- Enables logging of chromium network stack in logcat
Change-Id: I2d28c8392248a056b3cee305dd4d4475ebba4337
next segment. This can happen if for whatever reason we playback video slower than
it is served.
Also removed some unnecessary verbosity from ChromiumHTTPDataSource.
Change-Id: I6e870879310c9efe4d50d7dc0883c08405442d79
related-to-bug: 5120425
Ducati binary starting to support frame-rate on i/p port rather than o/p port,
stagefright can start giving frame-rate frame rate on i/p port and 0 on o/p
port.
Change-Id: I70830747e43e6bfcac73f0fb01de0ab0235a7b28
Signed-Off-By: Sarthak Aggarwal<sarthak@ti.com>
Signed-off-by: Iliyan Malchev <malchev@google.com>
Extend the media button event registration AudioManager API to
enable applications to register as a client of "remote controls"
and let them provide information meant to be displayed
by the remotes.
AudioService sends a AudioManager.REMOTE_CONTROL_CLIENT_CHANGED
intent to let remote controls know when / from whom they can
retrieve the information to display.
Only application that own audio focus, are the currently
registered media button event receiver, and have registered
a remote control client, are eligible to appear on the
remote control.
To address in future CLs:
- change how a remote control client forces a refresh
- rename methods called under lock to ___Locked()
- make API public
Change-Id: Icca30ab05dac2605ee9246f8acb27a03dcea077a
Do not take NOTIFICATIONS_USE_RING_VOLUME setting into account now
that it has been deprecated. STREAM_NOTIFICATION and STREAM_RING
always use the same volume setting.
Change-Id: Ibf6e5f86133d3781fd9e66455be64582318d93e6
the original connect/disconnect hooks are deprecated
and replace by api_connect/api_disconnect. the original
hooks are no no-ops.
api_connect/api_disconnect is now only called from the
android framework.
Bug: 5057915
Change-Id: I8ca64cd1acd6cabf915bf54689ec2e5f6dfa495a
No specific order is specified in the C++ standard, but the order of
the calls to Parcel read commands matters. Move any calls with multiple
reads to local variables.
Fix for bug 5104979.
Change-Id: I709aa040e990d2659e7a3a089f7a42ae812de9ff
Also properly signal a "hard" discontinuity, i.e. a possible format change
when a discontinuity is signalled explicitly as part of the playlist.
Change-Id: Ic347d3d11d39b0411c3726a7c723bcf13092b8bc
related-to-bug: 5103155, 5103013
Fixes problems introduced with bulk insert support
Bug: 5092877
Change-Id: If3c0c9054d5effe0a1d7a75e85635b41ba1591f5
Signed-off-by: Mike Lockwood <lockwood@android.com>
Add a destroyOMXPlugin() to complemenet createOMXPlugin(). Since the latter is
an opaque call into a library, it is not safe to assume that the object thus
returned may be destroyed by calling delete. This patch will call
destroyOMXPlugin() in libstagefrighthw.so, if the symbol is defined there.
Otherwise, it will default to a delete as before.
Also, prefer C linkage for the symbol names. Using mangled C++ symbols is not
very safe and makes it hard to grep for those symbols through the code. We
attempt to locate createOMXPlugin as a C symbol then try it as a mangled C++
one. We only use C linkage for destroyOMXPlugin since this symbol is new.
Change-Id: Ied23c910486856a0f18056df049c871234797c5c
Signed-off-by: Iliyan Malchev <malchev@google.com>
