26e8ea5ddf5999f3bf6cc0742cb4fda5b0fd3eb2: Bug 1270572 - write tests for the umprompted gUM request; r=florian draft
Munro Mengjue Chiang <mchiang@mozilla.com> - Thu, 15 Dec 2016 15:41:58 +0800 - rev 450837
Push 38960 by mchiang@mozilla.com at Mon, 19 Dec 2016 03:19:27 +0000
Bug 1270572 - write tests for the umprompted gUM request; r=florian MozReview-Commit-ID: LwXICrIVAfd
af0367bfbe920f4aa6c23212117214d5907e90dd: Bug 1270572 - allow un-prompted gUM access if the page has a live track connected to the same device; r=florian, gcp draft
Munro Mengjue Chiang <mchiang@mozilla.com> - Thu, 01 Dec 2016 15:12:38 +0800 - rev 450836
Push 38960 by mchiang@mozilla.com at Mon, 19 Dec 2016 03:19:27 +0000
Bug 1270572 - allow un-prompted gUM access if the page has a live track connected to the same device; r=florian, gcp MozReview-Commit-ID: AcLD2TZ3t6S
183a6786863eef7f820858cc08bc99532e3d0912: Bug 1316482 - remove collapsible white spaces as CSS Text Module Level 3 specifies. draft
Jeremy Chen <jeremychen@mozilla.com> - Mon, 19 Dec 2016 11:15:35 +0800 - rev 450835
Push 38959 by jichen@mozilla.com at Mon, 19 Dec 2016 03:16:30 +0000
Bug 1316482 - remove collapsible white spaces as CSS Text Module Level 3 specifies. According to the specification, if white space characters are considered collapsible, they should be removed before applying segment break transformation rules during the text transform. In this patch, we move the white space characters processing into a sub-loop. So, we can observe the context of the white space characters and apply the collapsing and transformation rules accordingly. MozReview-Commit-ID: 1JStjFk5TBs
95b5e90d6d4142c88bb78ea43f20a68402842c27: Bug 1324251 - Unconditionally parse windows drive letters in file paths; r?valentin draft
Manish Goregaokar <manishearth@gmail.com> - Sun, 18 Dec 2016 18:38:01 -0800 - rev 450834
Push 38958 by bmo:manishearth@gmail.com at Mon, 19 Dec 2016 02:38:18 +0000
Bug 1324251 - Unconditionally parse windows drive letters in file paths; r?valentin MozReview-Commit-ID: 6nBPHYHNprF
913cd310b0d1c1f1303e67b73c5d189bc7da8b06: Bug 1295921 - PD: Keep last keyframe in MFR. r?jwwang, jya draft
Dan Glastonbury <dglastonbury@mozilla.com> - Mon, 19 Dec 2016 11:47:27 +1000 - rev 450833
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - PD: Keep last keyframe in MFR. r?jwwang, jya To implement resuming video playback when drawImage() is invoked with a suspended video element, the MFR is extended to keep the last demuxed keyframe. This is done to avoid invoking track demuxing which can cause a dead lock via the media cache which ends up blocking on main thread, which is blocked. A new seek type, PrevFrame, is used to invoke the MFR to decode the stored keyframe and flush the decoder to force out the image. Because the decoder has been flushed, the MDSM starts a second seek to the current time stamp to recover the video stream. MozReview-Commit-ID: HiPxfVnG5j5
c727af5916c4847841afd4b11c5ea646b7a7538a: Bug 1295921 - PC: Implement blocking image get for SurfaceFromElement. r?jwwang draft
Dan Glastonbury <dglastonbury@mozilla.com> - Tue, 27 Sep 2016 15:30:59 +1000 - rev 450832
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - PC: Implement blocking image get for SurfaceFromElement. r?jwwang Connect all the pieces: * HTMLMediaElement::GetCurrentImage() waits on next rendered video frame before locking ImageContainer. * MediaDecoder::WaitOnNextRenderedVideoFrame(): * Sets suspend taint flag and returns if ImageContainer has a frame. * Otherwise, sets up monitor protected flag, sets suspend taint, and dispatches change tasks to MDSM TaskQueue before waiting on done flag. * MDSM resumes decoding, and * Signals done flag when after VideoSink::RenderVideoFrames adds new frame to ImageContainer, or * If resuming seeking fails, signals done flag to unblock main thread. MozReview-Commit-ID: HlGf5UM70EJ
f04eae06ddd9467003bb538780ed260f3e15b52b: Bug 1295921 - PB: Implement dispatching of pending main thread MediaDecoder tasks. r?jwwang draft
Dan Glastonbury <dglastonbury@mozilla.com> - Tue, 27 Sep 2016 15:26:41 +1000 - rev 450831
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - PB: Implement dispatching of pending main thread MediaDecoder tasks. r?jwwang Add support to flush pending tasks in the MainThread TailDispatcher, ensuring they are delivered in the correct order. This is used in the case that a video element with a suspended decoder is used in nsLayoutUtils::SurfaceFromElement. The 'has suspend taint' variable change needs to be dispatched to the MDSM TaskQueue before blocking the main thread to wait for the next frame. Once the variable change has been successfully dispatched, the MDSM with start decoding video frames again on it's thread pool. MozReview-Commit-ID: JrZepBGmjJt
d06b67edd6b9a8e62fabab93e06bfc8224e116cb: Bug 1295921 - PA: Return blank decoder to being green. r?jwwang draft
Dan Glastonbury <dglastonbury@mozilla.com> - Wed, 21 Sep 2016 14:06:51 +1000 - rev 450830
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - PA: Return blank decoder to being green. r?jwwang MozReview-Commit-ID: 44QZeHKlt3o
968275f698f2b6cdb08c696288e36a1104e6e978: Bug 1295921 - P9: NullDecoderModule returns null VideoData. r?jwwang draft
Dan Glastonbury <dglastonbury@mozilla.com> - Wed, 21 Sep 2016 15:36:48 +1000 - rev 450829
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - P9: NullDecoderModule returns null VideoData. r?jwwang Return a VideoData with dimension of zero size and no image data. MozReview-Commit-ID: 59J4leiqSvP
30ec63c26d6f174a8a97ee01e09f0835f76c3ae3: Bug 1295921 - P8: Extract BlankMediaDataDecoder so it can be shared. r?jwwang draft
Dan Glastonbury <dglastonbury@mozilla.com> - Wed, 21 Sep 2016 14:27:58 +1000 - rev 450828
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - P8: Extract BlankMediaDataDecoder so it can be shared. r?jwwang Change name to 'Dummy' to signify it's base for decoders that don't decode. MozReview-Commit-ID: 8RY8eKpWJE3
7751fe83806471fb98f124018daab78055fdc551: Bug 1295921 - P7: Test drawImage gets a non-white image from suspended video. r?jwwang draft
Dan Glastonbury <dglastonbury@mozilla.com> - Wed, 14 Sep 2016 17:04:12 +1000 - rev 450827
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - P7: Test drawImage gets a non-white image from suspended video. r?jwwang Test that drawImage call using element source with a video, that has video decode suspended, will get image and a white frame from the blank decoder. MozReview-Commit-ID: LM0jLPWwxU7
7f516a0e3426d3214d71ae4697d4b456016a8504: Bug 1295921 - P6: Move resume from suspend into new function. r?jwwang draft
Dan Glastonbury <dglastonbury@mozilla.com> - Wed, 28 Sep 2016 10:04:57 +1000 - rev 450826
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - P6: Move resume from suspend into new function. r?jwwang Extract resume from suspend code from visibility changed handler and put into a new function to allow this code be called when the suspend taint changes. MozReview-Commit-ID: 5qXCQrnNy0l
5a2984a53741f19df84b4215c46dd54101b4be3b: Bug 1295921 - P5: Test video suspend canceling. r?jwwang draft
Dan Glastonbury <dglastonbury@mozilla.com> - Thu, 29 Sep 2016 14:07:23 +1000 - rev 450825
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - P5: Test video suspend canceling. r?jwwang MozReview-Commit-ID: B1zLWS9HeSG
35c365c257bde40bbd556bd70ec1981b292ae5ee: Bug 1295921 - P4: Clean up suspend timer canceling. r?jwwang draft
Dan Glastonbury <dglastonbury@mozilla.com> - Wed, 31 Aug 2016 13:12:56 +1000 - rev 450824
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - P4: Clean up suspend timer canceling. r?jwwang MozReview-Commit-ID: 9yObB8msjtY
de02b8d96f9d6a1525fa7e0c81613e26e6445750: Bug 1295921 - P3: Test element becomes tainted by DrawImage. r?jwwang, smaug draft
Dan Glastonbury <dglastonbury@mozilla.com> - Mon, 29 Aug 2016 16:29:04 +1000 - rev 450823
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - P3: Test element becomes tainted by DrawImage. r?jwwang, smaug MozReview-Commit-ID: AHFbNxRtJqG
f7de938ddd61070049bdceee1adb10faf41efd91: Bug 1295921 - P2: Mark element tainted when DrawImage is used. r?jwwang, mattwoodrow draft
Dan Glastonbury <dglastonbury@mozilla.com> - Mon, 29 Aug 2016 16:25:22 +1000 - rev 450822
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - P2: Mark element tainted when DrawImage is used. r?jwwang, mattwoodrow Mark video element as tainted (stored on the decoder owned by video element) when the video is used as source to drawImage() on canvas. MozReview-Commit-ID: DkDgXflTN49
016e870f2a6adc2847d0594827c02ccc844e527e: Bug 1295921 - P1: Track decoder tainting. r?jwwang draft
Dan Glastonbury <dglastonbury@mozilla.com> - Mon, 29 Aug 2016 16:19:50 +1000 - rev 450821
Push 38957 by bmo:dglastonbury@mozilla.com at Mon, 19 Dec 2016 02:15:56 +0000
Bug 1295921 - P1: Track decoder tainting. r?jwwang Some uses of media elements should 'taint' the element so that the video doesn't participate in video decode suspending. Add the infrastructure to track the taint status on MediaDecoder and mirror the status to MediaDecoderStateMachine. MozReview-Commit-ID: 1nrNqg0KavT
b340d50cab713452b999112294a2e45ccb9b19fe: Converting from YUVA to RGBA on the software side, in the case of alpha. r?jya draft
karo <kkoorts@mozilla.com> - Fri, 16 Dec 2016 11:44:24 +1300 - rev 450820
Push 38956 by bmo:kkoorts@mozilla.com at Mon, 19 Dec 2016 02:01:50 +0000
Converting from YUVA to RGBA on the software side, in the case of alpha. r?jya MozReview-Commit-ID: JZDdbM2gt0g
85aa12c025526b15dba42ed7517d88125f5119b0: Refactored YCbCrBuffer to YCbCrABuffer. r?jya draft
karo <kkoorts@mozilla.com> - Fri, 16 Dec 2016 09:25:54 +1300 - rev 450819
Push 38956 by bmo:kkoorts@mozilla.com at Mon, 19 Dec 2016 02:01:50 +0000
Refactored YCbCrBuffer to YCbCrABuffer. r?jya MozReview-Commit-ID: LcZi4jLjB1T
efbf72a1e1d8cf00671512d15c7d6b5c52a0fdb3: Bug 1311910 - Add telemetry to measure update error and update timeout rate for V2 and V4. r=francois draft
DimiL <dlee@mozilla.com> - Mon, 19 Dec 2016 09:43:02 +0800 - rev 450818
Push 38955 by dlee@mozilla.com at Mon, 19 Dec 2016 01:44:31 +0000
Bug 1311910 - Add telemetry to measure update error and update timeout rate for V2 and V4. r=francois MozReview-Commit-ID: JL4aZrUOGH7
(0) -300000 -100000 -30000 -10000 -3000 -1000 -300 -100 -50 -20 +20 +50 +100 +300 +1000 +3000 +10000 +30000 +100000 +300000 tip