searching for reviewer(pehrsons)
61396b1eaa14477e196564ea0c4c0ba79439a64d: Bug 1483926 - Take into account mInputDeviceID is always nullptr on Android for now. r=pehrsons
Paul Adenot <paul@paul.cx> - Tue, 21 Aug 2018 16:32:24 +0000 - rev 830767
Push 118854 by bmo:a.beingessner@gmail.com at Wed, 22 Aug 2018 20:13:35 +0000
Bug 1483926 - Take into account mInputDeviceID is always nullptr on Android for now. r=pehrsons Differential Revision: https://phabricator.services.mozilla.com/D3899
f2dc5d8ac9ac55899c14d7a031951ccef0bc720d: Bug 1480036 - Allow scaling all MSG volume with a pref. r=pehrsons
Paul Adenot <paul@paul.cx> - Wed, 01 Aug 2018 12:10:32 +0200 - rev 830514
Push 118839 by mixedpuppy@gmail.com at Tue, 21 Aug 2018 18:59:46 +0000
Bug 1480036 - Allow scaling all MSG volume with a pref. r=pehrsons Differential Revision: https://phabricator.services.mozilla.com/D2594
352e47aeafac716617392486f88c20107dce2a00: Bug 1480036 - Stop relying on having the default device set globally. r=pehrsons
Paul Adenot <paul@paul.cx> - Wed, 01 Aug 2018 15:18:32 +0200 - rev 830513
Push 118839 by mixedpuppy@gmail.com at Tue, 21 Aug 2018 18:59:46 +0000
Bug 1480036 - Stop relying on having the default device set globally. r=pehrsons Differential Revision: https://phabricator.services.mozilla.com/D3508
3f99a091261fb67d798c7f16808d5a19efc7364b: Bug 1480036 - Allow forcing a specific audio output device from a test. r=pehrsons
Paul Adenot <paul@paul.cx> - Wed, 01 Aug 2018 14:16:30 +0200 - rev 830512
Push 118839 by mixedpuppy@gmail.com at Tue, 21 Aug 2018 18:59:46 +0000
Bug 1480036 - Allow forcing a specific audio output device from a test. r=pehrsons Differential Revision: https://phabricator.services.mozilla.com/D3507
655d8b94c8f1f78185f51a11ff0d30f929ac5016: Bug 1483926 - Fix Android device enumeration. r=pehrsons
Paul Adenot <paul@paul.cx> - Fri, 17 Aug 2018 15:34:16 +0000 - rev 830220
Push 118825 by bmo:rob@robwu.nl at Mon, 20 Aug 2018 17:24:40 +0000
Bug 1483926 - Fix Android device enumeration. r=pehrsons Not having a return here means that we end up clearing our device list. Differential Revision: https://phabricator.services.mozilla.com/D3613
67c4ed3c73d7bb770edcbdffebfb5688a3822a5b: Bug 1404977 - Tests P2: Add test to ensure multiple, concurrent gUM calls in a single window succeed. r=pehrsons
Bryce Van Dyk <bvandyk@mozilla.com> - Fri, 03 Aug 2018 10:35:59 -0400 - rev 826457
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Tests P2: Add test to ensure multiple, concurrent gUM calls in a single window succeed. r=pehrsons MozReview-Commit-ID: HGwcu6Z2mDn
c2d350fb074fb0c59eb922973ee33fc8ceeec0a5: Bug 1404977 - Tests P1: Add test to ensure multiple, concurrent gUM calls in separate iframes succeed. r=pehrsons
Bryce Van Dyk <bvandyk@mozilla.com> - Tue, 10 Jul 2018 15:09:50 +0200 - rev 826456
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Tests P1: Add test to ensure multiple, concurrent gUM calls in separate iframes succeed. r=pehrsons MozReview-Commit-ID: Iyr2A0R1rMd
d3fc55ebb24ab7117d5bbe740653af807dfe05b0: Bug 1404977 - Part 17 - Re-implement the workaround for the lack of input device enumeration on Android. r=pehrsons
Paul Adenot <paul@paul.cx> - Wed, 04 Jul 2018 18:00:57 +0200 - rev 826454
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 17 - Re-implement the workaround for the lack of input device enumeration on Android. r=pehrsons MozReview-Commit-ID: 5EiQ6a3OaIR
edc7a43bb7592c5407564151feeeb38c3684281f: Bug 1404977 - Part 16 - Unit test CubebDeviceEnumerator. r=pehrsons
Paul Adenot <paul@paul.cx> - Fri, 22 Jun 2018 11:55:31 +0200 - rev 826453
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 16 - Unit test CubebDeviceEnumerator. r=pehrsons This is done by implementing a fake cubeb backend that implements the subset of operations we need, while offering an API to be able to control what this backend is doing. Because we're reimplementing the private cubeb API, it is necessary to copy part of a cubeb internal header, and mimick exactly how the vtable mechanism to do the dynamic dispatch to the diffferent backends in cubeb works. This is not ideal but works. When the cubeb API functions are called (from deep in the Gecko process), we re-bind the call to the mock cubeb backend object and behave exactly like a normal backend (calling various callbacks and returning fake objects). Finally, we inject this mock cubeb backend to the running Gecko process (in lieu of the real one that would have been picked) by setting the global sCubebBackend variable via a private API exposed only in the test in CubebUtils.h. MozReview-Commit-ID: 8ZbJhl7pZ2t
ecbae0557eaae616491c2c87fe42591bb64d7b48: Bug 1404977 - Part 15 - Invalidate the device cache before re-enumerating devices when the cubeb backend does not support dynamic device collection invalidation. r=pehrsons
Paul Adenot <paul@paul.cx> - Fri, 22 Jun 2018 11:47:06 +0200 - rev 826452
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 15 - Invalidate the device cache before re-enumerating devices when the cubeb backend does not support dynamic device collection invalidation. r=pehrsons Also, clear the array that's been passed in before appending the new devices. MozReview-Commit-ID: BTnwzyKBrb5
3259ebfa7dea7816ce60dbd0814908e2d66031d4: Bug 1404977 - Part 14 - Add a way to set the global cubeb* singleton at runtime, from a test. r=pehrsons
Paul Adenot <paul@paul.cx> - Fri, 22 Jun 2018 11:45:31 +0200 - rev 826451
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 14 - Add a way to set the global cubeb* singleton at runtime, from a test. r=pehrsons This is for testing purposes only. Defining ENABLE_SET_CUBEB_BACKEND before including CubebUtils.h will expose the function. This is not to be set outside of test files. MozReview-Commit-ID: D0V8oLj9xo6
dc0f442f208cec6baa8d2675faff00a5adedb143: Bug 1404977 - Part 13 - Remove useless mutex, and assert why they were useless, in WebRTCAudioDataListeners. r=pehrsons
Paul Adenot <paul@paul.cx> - Thu, 31 May 2018 16:44:00 +0200 - rev 826450
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 13 - Remove useless mutex, and assert why they were useless, in WebRTCAudioDataListeners. r=pehrsons MozReview-Commit-ID: 2Mb5WZXbYgS
eba12c9748f882d9a1d84038cf9b3ae15ebe0987: Bug 1404977 - Part 12 - Make DeviceChanged() notification thread safe by using the MSG message queue. r=pehrsons
Paul Adenot <paul@paul.cx> - Thu, 31 May 2018 16:42:24 +0200 - rev 826449
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 12 - Make DeviceChanged() notification thread safe by using the MSG message queue. r=pehrsons This is slightly slower, especially if the main thread is busy, but it's cleaner and actually safe. MozReview-Commit-ID: 4C2FalxmE3L
d28f054ca88e4a441d0df0febc32227f005f97a3: Bug 1404977 - Part 11 - Make sure the default device is the first element in the list. r=pehrsons
Paul Adenot <paul@paul.cx> - Tue, 29 May 2018 15:20:47 +0200 - rev 826448
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 11 - Make sure the default device is the first element in the list. r=pehrsons MozReview-Commit-ID: LTJErFTm1wN
d09741d7dab28b401e4a1b753078f93be6a2eea3: Bug 1404977 - Part 10 - Make logging uniform, and consistently use GraphImpl() to access the graph impl. r=pehrsons
Paul Adenot <paul@paul.cx> - Tue, 17 Apr 2018 17:11:13 +0200 - rev 826447
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 10 - Make logging uniform, and consistently use GraphImpl() to access the graph impl. r=pehrsons MozReview-Commit-ID: DZ7zhQsBtvy
20a567fbf561dd88c059439f2b34e1024c58ca2c: Bug 1404977 - Part 9 - Propagate the changes to the GraphDrivers, simplifying them, and brokering all access through the MSG. r=pehrsons
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 16:01:56 +0200 - rev 826446
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 9 - Propagate the changes to the GraphDrivers, simplifying them, and brokering all access through the MSG. r=pehrsons For an AudioCallbackDriver, the number of input channels is immutable, and passed at construction, so that it's less necessary to rely on global state. MozReview-Commit-ID: F9TL1H92z3W
507d5c269d25aee79fe081b1b02a4970ddcc1fb0: Bug 1404977 - Part 8 - Tell the MSG the MediaEngineAudioSource are now independent and that we can have multiple of them, cleanup the MSG-side API for managing them. r=pehrsons
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 16:01:38 +0200 - rev 826445
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 8 - Tell the MSG the MediaEngineAudioSource are now independent and that we can have multiple of them, cleanup the MSG-side API for managing them. r=pehrsons The MSG now can feed microphone data to all its input listeners. This paves the way for multiple input device, if we feel it's needed at some point, but does not implement it. The method for adding/removing inputs are also cleaned up. MozReview-Commit-ID: 9OX4Da6Gjq2
c63f8579e0dbc3b3401897db3095b51ef4af3c0e: Bug 1404977 - Part 7 - Make each MediaEngineWebRTCMicrophoneSource independent. r=pehrsons
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 15:37:18 +0200 - rev 826444
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 7 - Make each MediaEngineWebRTCMicrophoneSource independent. r=pehrsons MozReview-Commit-ID: 1RSsT6AV3iI
700fadf31c3a79c98679f0986900c06d494e0f37: Bug 1404977 - Part 6 - Remove unused include for lock-free FIFO. r=pehrsons
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 15:33:28 +0200 - rev 826443
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 6 - Remove unused include for lock-free FIFO. r=pehrsons MozReview-Commit-ID: 8gVKlD7lMIX
16671b7c328ac8fe27fef7906aac7afef9dc7a3d: Bug 1404977 - Part 5 - Allow querying the number of input channels from a WebRTCAudioDataListener. r=pehrsons
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 15:30:58 +0200 - rev 826442
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 5 - Allow querying the number of input channels from a WebRTCAudioDataListener. r=pehrsons MozReview-Commit-ID: JkpEZOJpOTl
72fd73a9d4e3180fe0c16479b2da5f891addb312: Bug 1404977 - Part 3 - Remove global statics, introduce an audio device enumerator r=pehrsons
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 15:18:22 +0200 - rev 826441
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 3 - Remove global statics, introduce an audio device enumerator r=pehrsons MozReview-Commit-ID: 55VlpGIAOxp
3e1bbbc9047f732251966666c1d5a80265c3593f: Bug 1404977 - Part 2 - Augment AudioDeviceInfo with a cubeb device id. r=pehrsons
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 15:24:52 +0200 - rev 826440
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 2 - Augment AudioDeviceInfo with a cubeb device id. r=pehrsons MozReview-Commit-ID: GrOnxHuN2Oj
127bff24bfc8fa5d0e619ea727ab4a0c522c4497: Bug 1404977 - Part 1 - Add missing lock to the PREF_CUBEB_BACKEND branch of the pref callback in CubebUtils.cpp. r=pehrsons
Paul Adenot <paul@paul.cx> - Tue, 17 Apr 2018 16:45:33 +0200 - rev 826439
Push 118333 by bmo:snorp@snorp.net at Fri, 03 Aug 2018 20:42:37 +0000
Bug 1404977 - Part 1 - Add missing lock to the PREF_CUBEB_BACKEND branch of the pref callback in CubebUtils.cpp. r=pehrsons MozReview-Commit-ID: CBAmQKDq36z
fa8851738c196056c09ce5f5992ca6ade79e5f85: Bug 1404977 - Tests P2: Add test to ensure multiple, concurrent gUM calls in a single window succeed. r?pehrsons draft
Bryce Van Dyk <bvandyk@mozilla.com> - Fri, 03 Aug 2018 10:35:59 -0400 - rev 826330
Push 118293 by bvandyk@mozilla.com at Fri, 03 Aug 2018 15:18:25 +0000
Bug 1404977 - Tests P2: Add test to ensure multiple, concurrent gUM calls in a single window succeed. r?pehrsons MozReview-Commit-ID: HGwcu6Z2mDn
6e1f95afdf4134f45236f75f2fc6454fda751895: Bug 1404977 - Tests P1: Add test to ensure multiple, concurrent gUM calls in separate iframes succeed. r?pehrsons draft
Bryce Van Dyk <bvandyk@mozilla.com> - Tue, 10 Jul 2018 15:09:50 +0200 - rev 826329
Push 118293 by bvandyk@mozilla.com at Fri, 03 Aug 2018 15:18:25 +0000
Bug 1404977 - Tests P1: Add test to ensure multiple, concurrent gUM calls in separate iframes succeed. r?pehrsons MozReview-Commit-ID: Iyr2A0R1rMd
1f50722e36b4684fd51b463f230546f1b495cded: Bug 1404977 - Tests P2: Add test to ensure multiple, concurrent gUM calls in a single window succeed. r?pehrsons draft
Bryce Van Dyk <bvandyk@mozilla.com> - Wed, 01 Aug 2018 16:23:17 -0400 - rev 826299
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Tests P2: Add test to ensure multiple, concurrent gUM calls in a single window succeed. r?pehrsons MozReview-Commit-ID: HGwcu6Z2mDn
d6b16e15c793965cdffb3e47e451c3723bfc8477: Bug 1404977 - Tests P1: Add test to ensure multiple, concurrent gUM calls in separate iframes succeed. r?pehrsons draft
Bryce Van Dyk <bvandyk@mozilla.com> - Tue, 10 Jul 2018 15:09:50 +0200 - rev 826298
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Tests P1: Add test to ensure multiple, concurrent gUM calls in separate iframes succeed. r?pehrsons MozReview-Commit-ID: Iyr2A0R1rMd
09915b42e9179956fbf62e006a311be9bd178f78: Bug 1404977 - Part 17 - Re-implement the workaround for the lack of input device enumeration on Android. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Wed, 04 Jul 2018 18:00:57 +0200 - rev 826297
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 17 - Re-implement the workaround for the lack of input device enumeration on Android. r?pehrsons MozReview-Commit-ID: 5EiQ6a3OaIR
83f950576c00cb5b20910fc887ebb105cb9c06e7: Bug 1404977 - Part 16 - Unit test CubebDeviceEnumerator. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Fri, 22 Jun 2018 11:55:31 +0200 - rev 826296
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 16 - Unit test CubebDeviceEnumerator. r?pehrsons This is done by implementing a fake cubeb backend that implements the subset of operations we need, while offering an API to be able to control what this backend is doing. Because we're reimplementing the private cubeb API, it is necessary to copy part of a cubeb internal header, and mimick exactly how the vtable mechanism to do the dynamic dispatch to the diffferent backends in cubeb works. This is not ideal but works. When the cubeb API functions are called (from deep in the Gecko process), we re-bind the call to the mock cubeb backend object and behave exactly like a normal backend (calling various callbacks and returning fake objects). Finally, we inject this mock cubeb backend to the running Gecko process (in lieu of the real one that would have been picked) by setting the global sCubebBackend variable via a private API exposed only in the test in CubebUtils.h. MozReview-Commit-ID: 8ZbJhl7pZ2t
94fc24b763abde0543ef0b0bbb69af5e726bb39c: Bug 1404977 - Part 15 - Invalidate the device cache before re-enumerating devices when the cubeb backend does not support dynamic device collection invalidation. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Fri, 22 Jun 2018 11:47:06 +0200 - rev 826295
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 15 - Invalidate the device cache before re-enumerating devices when the cubeb backend does not support dynamic device collection invalidation. r?pehrsons Also, clear the array that's been passed in before appending the new devices. MozReview-Commit-ID: BTnwzyKBrb5
eab625a18952caae8751f807a9197acf97b1e743: Bug 1404977 - Part 14 - Add a way to set the global cubeb* singleton at runtime, from a test. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Fri, 22 Jun 2018 11:45:31 +0200 - rev 826294
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 14 - Add a way to set the global cubeb* singleton at runtime, from a test. r?pehrsons This is for testing purposes only. Defining ENABLE_SET_CUBEB_BACKEND before including CubebUtils.h will expose the function. This is not to be set outside of test files. MozReview-Commit-ID: D0V8oLj9xo6
e5eeb879172e339a61563ee3901d016dfa0202be: Bug 1404977 - Part 13 - Remove useless mutex, and assert why they were useless, in WebRTCAudioDataListeners. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Thu, 31 May 2018 16:44:00 +0200 - rev 826293
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 13 - Remove useless mutex, and assert why they were useless, in WebRTCAudioDataListeners. r?pehrsons MozReview-Commit-ID: 2Mb5WZXbYgS
b89eed97cc1cd7ae226f9313268deb1591b1720d: Bug 1404977 - Part 12 - Make DeviceChanged() notification thread safe by using the MSG message queue. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Thu, 31 May 2018 16:42:24 +0200 - rev 826292
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 12 - Make DeviceChanged() notification thread safe by using the MSG message queue. r?pehrsons This is slightly slower, especially if the main thread is busy, but it's cleaner and actually safe. MozReview-Commit-ID: 4C2FalxmE3L
f1e3cc27c30a0cc0a63e6b630d3e97f93ea3f18c: Bug 1404977 - Part 11 - Make sure the default device is the first element in the list. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Tue, 29 May 2018 15:20:47 +0200 - rev 826291
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 11 - Make sure the default device is the first element in the list. r?pehrsons MozReview-Commit-ID: LTJErFTm1wN
6d51ba46c4e93898963048b4338d3b13819d94fb: Bug 1404977 - Part 10 - Make logging uniform, and consistently use GraphImpl() to access the graph impl. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Tue, 17 Apr 2018 17:11:13 +0200 - rev 826290
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 10 - Make logging uniform, and consistently use GraphImpl() to access the graph impl. r?pehrsons MozReview-Commit-ID: DZ7zhQsBtvy
af9c5f46323a0ae32c1de3098e3b13686489befd: Bug 1404977 - Part 9 - Propagate the changes to the GraphDrivers, simplifying them, and brokering all access through the MSG. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 16:01:56 +0200 - rev 826289
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 9 - Propagate the changes to the GraphDrivers, simplifying them, and brokering all access through the MSG. r?pehrsons For an AudioCallbackDriver, the number of input channels is immutable, and passed at construction, so that it's less necessary to rely on global state. MozReview-Commit-ID: F9TL1H92z3W
bdaf37f087047f427e84a10db3eb0a35bae6e37c: Bug 1404977 - Part 8 - Tell the MSG the MediaEngineAudioSource are now independent and that we can have multiple of them, cleanup the MSG-side API for managing them. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 16:01:38 +0200 - rev 826288
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 8 - Tell the MSG the MediaEngineAudioSource are now independent and that we can have multiple of them, cleanup the MSG-side API for managing them. r?pehrsons The MSG now can feed microphone data to all its input listeners. This paves the way for multiple input device, if we feel it's needed at some point, but does not implement it. The method for adding/removing inputs are also cleaned up. MozReview-Commit-ID: 9OX4Da6Gjq2
27c9ff2b84662a41a5c35ccde45a6598e6b20f75: Bug 1404977 - Part 7 - Make each MediaEngineWebRTCMicrophoneSource independent. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 15:37:18 +0200 - rev 826287
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 7 - Make each MediaEngineWebRTCMicrophoneSource independent. r?pehrsons MozReview-Commit-ID: 1RSsT6AV3iI
1cb21dd94ed06b03848ce38a2dcafa0e5eb7d160: Bug 1404977 - Part 6 - Remove unused include for lock-free FIFO. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 15:33:28 +0200 - rev 826286
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 6 - Remove unused include for lock-free FIFO. r?pehrsons MozReview-Commit-ID: 8gVKlD7lMIX
e940939f2c5f7e7079976e9d062a7771cb9c3e5d: Bug 1404977 - Part 5 - Allow querying the number of input channels from a WebRTCAudioDataListener. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 15:30:58 +0200 - rev 826285
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 5 - Allow querying the number of input channels from a WebRTCAudioDataListener. r?pehrsons MozReview-Commit-ID: JkpEZOJpOTl
930c0977400520427d1fe05ff6e455897e44bba6: Bug 1404977 - Part 3 - Remove global statics, introduce an audio device enumerator r?pehrsons draft
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 15:18:22 +0200 - rev 826284
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 3 - Remove global statics, introduce an audio device enumerator r?pehrsons MozReview-Commit-ID: 55VlpGIAOxp
a7fa6a208169221f9f16a5c10326c9618142665b: Bug 1404977 - Part 2 - Augment AudioDeviceInfo with a cubeb device id. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Mon, 30 Apr 2018 15:24:52 +0200 - rev 826283
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 2 - Augment AudioDeviceInfo with a cubeb device id. r?pehrsons MozReview-Commit-ID: GrOnxHuN2Oj
526dcc927bf577e26e0e5c250443ebfd9dd54897: Bug 1404977 - Part 1 - Add missing lock to the PREF_CUBEB_BACKEND branch of the pref callback in CubebUtils.cpp. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Tue, 17 Apr 2018 16:45:33 +0200 - rev 826282
Push 118279 by bmo:apehrson@mozilla.com at Fri, 03 Aug 2018 12:32:15 +0000
Bug 1404977 - Part 1 - Add missing lock to the PREF_CUBEB_BACKEND branch of the pref callback in CubebUtils.cpp. r?pehrsons MozReview-Commit-ID: CBAmQKDq36z
975db62dd302e468665e4e658af7dd9fa27b3b20: Bug 1404977 - Tests P2: Add test to ensure multiple, concurrent gUM calls in a single window succeed. r?pehrsons draft
Bryce Van Dyk <bvandyk@mozilla.com> - Wed, 01 Aug 2018 16:23:17 -0400 - rev 826025
Push 118231 by bvandyk@mozilla.com at Thu, 02 Aug 2018 20:32:42 +0000
Bug 1404977 - Tests P2: Add test to ensure multiple, concurrent gUM calls in a single window succeed. r?pehrsons MozReview-Commit-ID: HGwcu6Z2mDn
ad2104b6d8becc65005dd89a975fac97cb0d0dfd: Bug 1404977 - Tests P1: Add test to ensure multiple, concurrent gUM calls in separate iframes succeed. r?pehrsons draft
Bryce Van Dyk <bvandyk@mozilla.com> - Tue, 10 Jul 2018 15:09:50 +0200 - rev 826024
Push 118231 by bvandyk@mozilla.com at Thu, 02 Aug 2018 20:32:42 +0000
Bug 1404977 - Tests P1: Add test to ensure multiple, concurrent gUM calls in separate iframes succeed. r?pehrsons MozReview-Commit-ID: Iyr2A0R1rMd
c6046b98dbb4e79e509012fa03440b124e55a0e5: Bug 1404977 - Part 17 - Re-implement the workaround for the lack of input device enumeration on Android. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Wed, 04 Jul 2018 18:00:57 +0200 - rev 826023
Push 118231 by bvandyk@mozilla.com at Thu, 02 Aug 2018 20:32:42 +0000
Bug 1404977 - Part 17 - Re-implement the workaround for the lack of input device enumeration on Android. r?pehrsons MozReview-Commit-ID: 5EiQ6a3OaIR
1222d9999db932f8076c1d103968da66d8e781a1: Bug 1404977 - Part 16 - Unit test CubebDeviceEnumerator. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Fri, 22 Jun 2018 11:55:31 +0200 - rev 826022
Push 118231 by bvandyk@mozilla.com at Thu, 02 Aug 2018 20:32:42 +0000
Bug 1404977 - Part 16 - Unit test CubebDeviceEnumerator. r?pehrsons This is done by implementing a fake cubeb backend that implements the subset of operations we need, while offering an API to be able to control what this backend is doing. Because we're reimplementing the private cubeb API, it is necessary to copy part of a cubeb internal header, and mimick exactly how the vtable mechanism to do the dynamic dispatch to the diffferent backends in cubeb works. This is not ideal but works. When the cubeb API functions are called (from deep in the Gecko process), we re-bind the call to the mock cubeb backend object and behave exactly like a normal backend (calling various callbacks and returning fake objects). Finally, we inject this mock cubeb backend to the running Gecko process (in lieu of the real one that would have been picked) by setting the global sCubebBackend variable via a private API exposed only in the test in CubebUtils.h. MozReview-Commit-ID: 8ZbJhl7pZ2t
eb6f24009c0caa4d5e7357990f8d44c6d4314f75: Bug 1404977 - Part 15 - Invalidate the device cache before re-enumerating devices when the cubeb backend does not support dynamic device collection invalidation. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Fri, 22 Jun 2018 11:47:06 +0200 - rev 826021
Push 118231 by bvandyk@mozilla.com at Thu, 02 Aug 2018 20:32:42 +0000
Bug 1404977 - Part 15 - Invalidate the device cache before re-enumerating devices when the cubeb backend does not support dynamic device collection invalidation. r?pehrsons Also, clear the array that's been passed in before appending the new devices. MozReview-Commit-ID: BTnwzyKBrb5
4f275cada6189f75f1fe882d7ecbaa02379d8b07: Bug 1404977 - Part 14 - Add a way to set the global cubeb* singleton at runtime, from a test. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Fri, 22 Jun 2018 11:45:31 +0200 - rev 826020
Push 118231 by bvandyk@mozilla.com at Thu, 02 Aug 2018 20:32:42 +0000
Bug 1404977 - Part 14 - Add a way to set the global cubeb* singleton at runtime, from a test. r?pehrsons This is for testing purposes only. Defining ENABLE_SET_CUBEB_BACKEND before including CubebUtils.h will expose the function. This is not to be set outside of test files. MozReview-Commit-ID: D0V8oLj9xo6
2e731964bdb7e39ae6f0dab9e086145fe7c40782: Bug 1404977 - Part 13 - Remove useless mutex, and assert why they were useless, in WebRTCAudioDataListeners. r?pehrsons draft
Paul Adenot <paul@paul.cx> - Thu, 31 May 2018 16:44:00 +0200 - rev 826019
Push 118231 by bvandyk@mozilla.com at Thu, 02 Aug 2018 20:32:42 +0000
Bug 1404977 - Part 13 - Remove useless mutex, and assert why they were useless, in WebRTCAudioDataListeners. r?pehrsons MozReview-Commit-ID: 2Mb5WZXbYgS