Bug 969322 - make :-moz-window-inactive state follow GDK_WINDOW_STATE_FOCUSED when drawing in titlebar r=karlt,stransky This patch fixes activation of the :-moz-window-inactive pseudoclass in client side decoration during drag operations. It depends on and complements the fixes provided in support of Bug 1491808. Obsoletes D50445 Differential Revision: https://phabricator.services.mozilla.com/D51307

Bug 1586293 - simplify thread passing after merging PVideoDecoderManager into PRemoteDecoderManager. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D49021

Bug 1591810 - Restore deleted wpt metadata after wpt-sync. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D50897

Bug 1587244 - Allow NaN in IIR filter output. r=karlt It's perfectly possible to create unstable filter, that can easily go to infinity, so this assert is not valid. Differential Revision: https://phabricator.services.mozilla.com/D49608

Bug 1499597 - Don't clamp the Q of a BiquadFilterNode when in low-pass and high-pass mode. r=karlt Bug 1265395 implemented the new filter equations from https://github.com/webaudio/web-audio-api/commit/a6842f2f733911a8ac6b330a405eac19878adc15, but missed removing the clamping. Differential Revision: https://phabricator.services.mozilla.com/D49600

Bug 1587622 - Use MOZ_ASSERT_UNREACHABLE() instead of double-conversion's UNREACHABLE() in DOM media code. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D48880

Bug 1585671 - Add a WPT to test that the delayTime of a DelayNode in a cycle is clamped to 128 samples. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D47931

Bug 1585671 - Add a WPT to test the minimum duration of cycles with a DelayNode in a feedback loop. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D47930

Bug 1585671 - Add a WPT to test that cycles without DelayNodes are muted. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D47929

Bug 1454998 - Rename streams to tracks. r=padenot,karlt,smaug This renames the following (in alphabetical order, non-exhaustive): AudioCaptureStream -> AudioCaptureTrack AudioNodeStream -> AudioNodeTrack AudioNodeExternalInputStream -> AudioNodeExternalInputTrack DirectMediaStreamTrackListener -> DirectMediaTrackListener MediaStream -> MediaTrack - Note that there's also dom::MediaTrack. Namespaces differentiate them. MediaStreamGraph -> MediaTrackGraph MediaStreamTrackListener -> MediaTrackListener MSG -> MTG (in comments) ProcessedMediaStream -> ProcessedMediaTrack SharedDummyStream -> SharedDummyTrack SourceMediaStream -> SourceMediaTrack StreamTime -> TrackTime TrackUnionStream -> ForwardedInputTrack - Because this no longer takes a union of anything, but only a single track as input. Other minor classes, members and comments have been updated to reflect these name changes. Differential Revision: https://phabricator.services.mozilla.com/D46146

Bug 1454998 - Remove AudioNodeStream::SampleRate(). r=karlt AudioNodeStream is a subclass of MediaStream, which now exposes a public const mSampleRate member. Differential Revision: https://phabricator.services.mozilla.com/D47688

Bug 1454998 - Remove the notion of streams from MediaStreamGraph. r=padenot,karlt This change mainly removes the `mTracks` member from MediaStream and moves all associated members up a level, so that a MediaStream in practice represents a single track. Classes will be renamed in a future patch to reflect this. Other changes include: The new `mEnded` member of MediaStream changes meaning to only become true when all data in the stream has been processed. It stems from StreamTracks::Track::mEnded which used to become true as soon as the last bit of data had been added to a track, and there could still be data in the track that would get processed in future iterations. We are moving towards not having any future data in tracks, which is why this change is ok to make -- keeping the old behavior will soon not make sense. TrackUnionStream is changed to no longer take a list of streams as input and forward the union of their tracks to itself. Instead it's limited to having one track as input at a time. The autofinishing functionality that TrackUnionStream had before has been transformed into an autoending functionality to allow it to defer ending until its been told that it's ok to end through the control API. This lets a single TrackUnionStream span the lifetime of multiple inputs, which will be useful for making DecodedStream spec compliant with HTMLMediaElement::CaptureStream(), and for implementing the currently discussed MediaRecorder::ReplaceTrack(), to name a few potential use cases. AudioNodeStreams used to only have a track (and thus an AudioSegment) if the EXTERNAL_OUTPUT flag was enabled on them. With all MediaStreams now representing a track, AudioNodeStreams inherently have an AudioSegment as a member. It is however only used with data if the EXTERNAL_OUTPUT flag is enabled. Differential Revision: https://phabricator.services.mozilla.com/D45821

Bug 1454998 - Remove calls to SetStreamOrderDirty() after affecting MediaStreams' finished state. r=karlt The stream order does not depend on streams' finished state, so these calls are unnecessary. Differential Revision: https://phabricator.services.mozilla.com/D47687

Bug 1454998 - Remove track blocking from MediaInputPort. r=karlt This functionality is not used since we moved to only having a single track per MediaStream (bug 1493613). Differential Revision: https://phabricator.services.mozilla.com/D47686

Bug 1454998 - Move TrackInfo::mTrackId away from TrackID as they're independent. r=karlt TrackID has been used to denote tracks that are part of MediaStreams in the MediaStreamGraph. The TrackInfo::mTrackId IDs are no longer used in the MediaStream APIs and as such it's not necessary that they have the type TrackID. This patch changes TrackInfo::mTrackId to be of type uint32_t instead, as that is the type of Mp4parseTrackInfo::track_id, which is assigned to TrackInfo::mTrackId as the sole remaining non-static user. Other users set a static trackId of `1` or `2`, and as such remain supported by this change. Differential Revision: https://phabricator.services.mozilla.com/D46763

Bug 1454998 - Rename streams to tracks. r=padenot,karlt,smaug This renames the following (in alphabetical order, non-exhaustive): AudioCaptureStream -> AudioCaptureTrack AudioNodeStream -> AudioNodeTrack AudioNodeExternalInputStream -> AudioNodeExternalInputTrack DirectMediaStreamTrackListener -> DirectMediaTrackListener MediaStream -> MediaTrack - Note that there's also dom::MediaTrack. Namespaces differentiate them. MediaStreamGraph -> MediaTrackGraph MediaStreamTrackListener -> MediaTrackListener MSG -> MTG (in comments) ProcessedMediaStream -> ProcessedMediaTrack SharedDummyStream -> SharedDummyTrack SourceMediaStream -> SourceMediaTrack StreamTime -> TrackTime TrackUnionStream -> ForwardedInputTrack - Because this no longer takes a union of anything, but only a single track as input. Other minor classes, members and comments have been updated to reflect these name changes. Differential Revision: https://phabricator.services.mozilla.com/D46146

Bug 1454998 - Remove AudioNodeStream::SampleRate(). r=karlt AudioNodeStream is a subclass of MediaStream, which now exposes a public const mSampleRate member. Differential Revision: https://phabricator.services.mozilla.com/D47688

Bug 1454998 - Remove the notion of streams from MediaStreamGraph. r=padenot,karlt This change mainly removes the `mTracks` member from MediaStream and moves all associated members up a level, so that a MediaStream in practice represents a single track. Classes will be renamed in a future patch to reflect this. Other changes include: The new `mEnded` member of MediaStream changes meaning to only become true when all data in the stream has been processed. It stems from StreamTracks::Track::mEnded which used to become true as soon as the last bit of data had been added to a track, and there could still be data in the track that would get processed in future iterations. We are moving towards not having any future data in tracks, which is why this change is ok to make -- keeping the old behavior will soon not make sense. TrackUnionStream is changed to no longer take a list of streams as input and forward the union of their tracks to itself. Instead it's limited to having one track as input at a time. The autofinishing functionality that TrackUnionStream had before has been transformed into an autoending functionality to allow it to defer ending until its been told that it's ok to end through the control API. This lets a single TrackUnionStream span the lifetime of multiple inputs, which will be useful for making DecodedStream spec compliant with HTMLMediaElement::CaptureStream(), and for implementing the currently discussed MediaRecorder::ReplaceTrack(), to name a few potential use cases. AudioNodeStreams used to only have a track (and thus an AudioSegment) if the EXTERNAL_OUTPUT flag was enabled on them. With all MediaStreams now representing a track, AudioNodeStreams inherently have an AudioSegment as a member. It is however only used with data if the EXTERNAL_OUTPUT flag is enabled. Differential Revision: https://phabricator.services.mozilla.com/D45821

Bug 1454998 - Remove calls to SetStreamOrderDirty() after affecting MediaStreams' finished state. r=karlt The stream order does not depend on streams' finished state, so these calls are unnecessary. Differential Revision: https://phabricator.services.mozilla.com/D47687

Bug 1454998 - Remove track blocking from MediaInputPort. r=karlt This functionality is not used since we moved to only having a single track per MediaStream (bug 1493613). Differential Revision: https://phabricator.services.mozilla.com/D47686

Bug 1454998 - Move TrackInfo::mTrackId away from TrackID as they're independent. r=karlt TrackID has been used to denote tracks that are part of MediaStreams in the MediaStreamGraph. The TrackInfo::mTrackId IDs are no longer used in the MediaStream APIs and as such it's not necessary that they have the type TrackID. This patch changes TrackInfo::mTrackId to be of type uint32_t instead, as that is the type of Mp4parseTrackInfo::track_id, which is assigned to TrackInfo::mTrackId as the sole remaining non-static user. Other users set a static trackId of `1` or `2`, and as such remain supported by this change. Differential Revision: https://phabricator.services.mozilla.com/D46763

Bug 1583512 - Remove remaining references to doppler shift from AudioBufferSourceNode. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D46934

Bug 1580659 - part5 : stop audio capturing explicitly. r=chunmin,karlt As we start audio capturing explicitly, we should also take responsibility to stop audio capturing when we don't need it. We were hiding too many details on `AudioChannelAgent` before, which allow us hard to know who and where we handle audio capturing. Differential Revision: https://phabricator.services.mozilla.com/D45752

Bug 1580659 - part3.5 : suspend nsNPAPIPluginInstance should not affect its muted state. r=karlt The suspend state and muted state are two different things, so changing the suspended state should not affect on the muted state. As nsAPI doesn't provide suspending API, both muting and suspending are actually sharing the same implementation, which is to mute/unmute nsAPI. Although we have used two separate variables to store muted and suspended state, in order to make them independent, we still don't want to unexpectedly unmute nsAPI when resuming from suspended state. Therefore, before unmuting nsAPI, we should consider both states. Differential Revision: https://phabricator.services.mozilla.com/D46319

Bug 1580659 - part3 : call 'WindowAudioCaptureChanged()' explicitly. r=chunmin,karlt Previously `AudioChannelService` called `WindowAudioCaptureChanged()` implicitly whenever we added the agent to the service [1], which made the audio capturing callback happen before finishing registration. There are two drawbacks, (1) it's hard to be aware of that the audio capturing callback would be called before finishing `AudioChannelAgent::NotifyStartedPlaying()` [2], which causes unclear call flow. (2) If someone checks `AudioChannelAgent::IsPlayingStarted()` [3] inside audio capturing callback, then we would find that the `mIsRegToService` is false even if we have registered the agent to `AudioChannelService` because `mIsRegToService` is updated in the last line in the `AudioChannelAgent::NotifyStartedPlaying()`, but the audio capturing callback could be executed before that. [1] https://searchfox.org/mozilla-central/rev/a777ff11b6d700a698c61e5bd17e73b044304494/dom/audiochannel/AudioChannelService.cpp#723 [2] https://searchfox.org/mozilla-central/rev/a777ff11b6d700a698c61e5bd17e73b044304494/dom/audiochannel/AudioChannelAgent.cpp#144 [3] https://searchfox.org/mozilla-central/rev/a777ff11b6d700a698c61e5bd17e73b044304494/dom/audiochannel/AudioChannelAgent.cpp#285 Differential Revision: https://phabricator.services.mozilla.com/D45750

Bug 1580659 - part5 : stop audio capturing explicitly. r=chunmin,karlt As we start audio capturing explicitly, we should also take responsibility to stop audio capturing when we don't need it. We were hiding too many details on `AudioChannelAgent` before, which allow us hard to know who and where we handle audio capturing. Differential Revision: https://phabricator.services.mozilla.com/D45752

Bug 1580659 - part3.5 : suspend nsNPAPIPluginInstance should not affect its muted state. r=karlt The suspend state and muted state are two different things, so changing the suspended state should not affect on the muted state. As nsAPI doesn't provide suspending API, both muting and suspending are actually sharing the same implementation, which is to mute/unmute nsAPI. Although we have used two separate variables to store muted and suspended state, in order to make them independent, we still don't want to unexpectedly unmute nsAPI when resuming from suspended state. Therefore, before unmuting nsAPI, we should consider both states. Differential Revision: https://phabricator.services.mozilla.com/D46319

Bug 1580659 - part3 : call 'WindowAudioCaptureChanged()' explicitly. r=chunmin,karlt Previously `AudioChannelService` called `WindowAudioCaptureChanged()` implicitly whenever we added the agent to the service [1], which made the audio capturing callback happen before finishing registration. There are two drawbacks, (1) it's hard to be aware of that the audio capturing callback would be called before finishing `AudioChannelAgent::NotifyStartedPlaying()` [2], which causes unclear call flow. (2) If someone checks `AudioChannelAgent::IsPlayingStarted()` [3] inside audio capturing callback, then we would find that the `mIsRegToService` is false even if we have registered the agent to `AudioChannelService` because `mIsRegToService` is updated in the last line in the `AudioChannelAgent::NotifyStartedPlaying()`, but the audio capturing callback could be executed before that. [1] https://searchfox.org/mozilla-central/rev/a777ff11b6d700a698c61e5bd17e73b044304494/dom/audiochannel/AudioChannelService.cpp#723 [2] https://searchfox.org/mozilla-central/rev/a777ff11b6d700a698c61e5bd17e73b044304494/dom/audiochannel/AudioChannelAgent.cpp#144 [3] https://searchfox.org/mozilla-central/rev/a777ff11b6d700a698c61e5bd17e73b044304494/dom/audiochannel/AudioChannelAgent.cpp#285 Differential Revision: https://phabricator.services.mozilla.com/D45750

Bug 1581315 part 2. Fix incorrect behavior of PannerNode::SetRolloffFactor. r=karlt Our implementation modified the state, in addition to throwing the exception. And the tests not only didn't test for this, but wouldn't have reached that code anyway, due to the harness giving up as soon as anything failed. Differential Revision: https://phabricator.services.mozilla.com/D46106

Bug 1547063 - fix advance scaling for emoji. r=karlt This fixes the regression in reftests/text/color-opacity-rtl-[12].html. Differential Revision: https://phabricator.services.mozilla.com/D45956

Bug 1547063 - fix advance scaling for emoji. r=karlt This fixes the regression in reftests/text/color-opacity-rtl-[12].html. This also reverts an earlier erroneous change to WPT content-counter-004.xht that was done in https://hg.mozilla.org/mozilla-central/rev/c969a93b0ca7 to silence an unexpected-pass that accidentally arose because of matching failures in test/ref when we switched to GetFTGlyphExtents in https://hg.mozilla.org/mozilla-central/rev/003f5a79c6a7. Now that the actual scaling issue is fixed, the behavior is again consistent with Cairo metrics causing the failure to reappear like before with Cairo metrics. Differential Revision: https://phabricator.services.mozilla.com/D45956

Bug 1575271 - Keep a SharedDummyStream in MediaEncoder for access to the MediaStreamGraph. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D45195

Bug 1575271 - Add crashtest. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D45194

Bug 1577184 - Guard against a shutting down AudioContext in MediaElementAudioSourceNode. r=karlt Depends on D44486 Differential Revision: https://phabricator.services.mozilla.com/D44487

Bug 1577184 - Add crashtest. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D44486

Bug 1577534 - Give MediaStreamGraphs individual ShutdownBlockers. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D44380

Bug 1577537 - Catch AppendMessage calls that could be too late. r=karlt If all streams and all ports have been destroyed, there's no guarantee that the graph is still alive. By forbidding AppendMessage calls after this point, we can catch bugs with the offending callsite still being in the stack. Differential Revision: https://phabricator.services.mozilla.com/D44223

Bug 1576656 - Only allow setting an AudioBuffer as convolution impulse if it has the same sample-rate than the associated AudioContext. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D43477

Bug 1572878 - Compute the reverb on the main thread. r=karlt This is necessary to keep the invariant that once the setter returns, the reverb is ready to process. Differential Revision: https://phabricator.services.mozilla.com/D43476

Bug 1572878 - Add a new method on AudioNodeStream, to be able to set a Reverb object on a ConvolverNode, from the main thread, using a ControlMessage. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D42900

Bug 1577537 - Catch AppendMessage calls that could be too late. r=karlt If all streams and all ports have been destroyed, there's no guarantee that the graph is still alive. By forbidding AppendMessage calls after this point, we can catch bugs with the offending callsite still being in the stack. Differential Revision: https://phabricator.services.mozilla.com/D44223

Bug 1576938 - Make MediaInputPort::Init work on an already-disconnected port. r=karlt Depends on D44020 Differential Revision: https://phabricator.services.mozilla.com/D44021

Bug 1576938 - Allow a live MediaStreamTrack to be created with a destroyed input stream. r=karlt This is a valid use case when cloning a MediaStreamTrack, if the source track of the clone has had its input stream destroyed out-of-band and that change hasn't been asynchronously communicated to the source track yet, i.e., it's still live. Depends on D44019 Differential Revision: https://phabricator.services.mozilla.com/D44020

Bug 1576938 - Add crashtest. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D44019

Bug 1577495 - Use a single SharedDummyStream per HTMLMediaElement. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D43982

Bug 1562827 - [Linux] Merge nsWindow::ReparentNativeWidgetInternal() to nsWindow::SetParent() and use nsWindow::ReparentNativeWidget() for toplevel windows only, r=karlt - Use former ReparentNativeWidgetInternal() code in nsWindow::SetParent(). Also update mToplevelParentWindow for Wayland to hold default toplevel window. - Make nsWindow::ReparentNativeWidget() to work on toplevel windows only and use only gtk_window_set_transient_for() to reparent a toplevel window. Also update mToplevelParentWindow for Wayland. Differential Revision: https://phabricator.services.mozilla.com/D43600

Bug 1573102 - Remove MediaStreamGraphImpl strong-refs from MediaEngineWebRTCAudio. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D43668

Bug 1573102 - Remove mGraph from MediaStreamTrack::MSGListener. r=karlt To accomodate this, MediaStreamTrackListener::NotifyEnded/Removed get an additional aGraph argument. NotifyEnabledStateChanged is getting it too, for completion. Differential Revision: https://phabricator.services.mozilla.com/D43667

Bug 1573102 - Remove mGraph from HTMLMediaElement::OutputMediaStream and OutputStreamManager. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D43666

Bug 1573102 - Remove mGraph from HTMLMediaElement::MediaStreamRenderer. r=karlt Differential Revision: https://phabricator.services.mozilla.com/D43665