summaryrefslogtreecommitdiffstats
path: root/dom
Commit message (Collapse)AuthorAgeLines
...
* Revert "Increase purple limit triggers for CC."wolfbeast2019-02-07-2/+2
| | | | This reverts commit babedf6c696f88734e59b63d0c6614962cc57519.
* Get RefPtr to transaction before using it. (DiD)wolfbeast2019-02-07-1/+3
| | | | | Avoid potential use-after-free by getting a RefPtr to a transaction before calling content code that could cause its de-allocation.
* Increase slice time for longer running CCs.Andrew McCreight2019-01-29-3/+14
| | | | | | | | | If a CC takes too long (around 50 slices) or gets interrupted by a GC, we have to finish it synchronously, which can cause a big pause. This patch tries to avoid that by eagerly increasing the slice budget the longer a CC goes on. It linearly increases the slice time from 5ms to 40ms as we approach the halfway point of a CC (1 second), matching GC pauses, and then leaves it at 40ms.
* Increase purple limit triggers for CC.wolfbeast2019-01-29-2/+2
| | | | | Making these much larger to allow more purple buffer buildup and prevent overzealous cycle collection on purple buffer pressure.
* Merge branch 'master' into cycle_collector-workwolfbeast2019-01-29-2/+12
|\
| * Properly camelCase dom.intersectionObserver.enabled pref.wolfbeast2019-01-19-2/+2
| |
| * Actually unlink targets from registered intersection observers.wolfbeast2019-01-19-0/+10
| | | | | | | | | | | | | | When a node is released/removed, and it has an intersection observer attached to it, that observer should be unlinked at the time of release. This resolves #935.
* | Remove NS_IMPL_CYCLE_COLLECTION_TRAVERSE_SCRIPT_OBJECTSwolfbeast2019-01-18-96/+1
| |
* | Consolidate tracing and traversing.wolfbeast2019-01-18-1/+1
|/
* Rewrite IntersectionObserver list handling to be more robust.wolfbeast2019-01-18-8/+40
| | | | Tag #935.
* Add isIntersecting property to IntersectionObserverEntry.wolfbeast2019-01-17-0/+11
| | | | | | Per updated spec. This resolves the issue raised in #249.
* Remove unused telemetry timer from HTMLMediaElement.hadeshkp2019-01-14-7/+0
|
* Telemetry: Remove stubs and related codeadeshkp2019-01-12-1160/+25
|
* Align Element.ScrollIntoView() with the spec.wolfbeast2019-01-08-25/+56
| | | | | | | This also removes the (unused) shadow alias from nsIDOMHTMLElement which used the different calling convention. This resolves #927
* Remove proprietary constructor on AudioContext.wolfbeast2019-01-04-33/+1
| | | | | | | | | This is a B2G leftover, was proprietary, and is causing issues because `AudioContext` can now have a parameter that is a property bag, per spec (we do not do anything with the property bag now; the audio back-end will use automatic values for everything). This resolves #924.
* Avoid redefined macro warning in MediaFormatReadertrav902018-12-31-0/+4
|
* [FFmpeg] Don't define AV_CODEC_ID_VPx with libavcodec 58trav902018-12-28-3/+6
|
* [FFmpeg] Use new decode API with recent FFmpeg version.trav902018-12-28-7/+66
| | | | In libavcodec 58 and later, the old avcodec_decode_video2 is broken and only return the first visible frame found after a VP9 super-frame.
* Fix crashiness of IntersectionObservers.wolfbeast2018-12-24-21/+25
| | | | Mozilla hashtables -still- suck.
* [intersection-observer] Calculate areas using int64_t.wolfbeast2018-12-22-4/+6
| | | | Tag #249
* Fix singed/unsigned type confusion for intersection threshold.wolfbeast2018-12-22-3/+3
| | | | Tag #249
* Use content area as the intersection rectangle ...wolfbeast2018-12-22-7/+11
| | | | | ... for custom root with overflow clip. Tag #249
* Map intersectionRect to the coordinate space of the target document.wolfbeast2018-12-22-7/+8
| | | | | | Spec says: "Map intersectionRect to the coordinate space of the viewport of the Document containing the target." Tag #249
* Use targetFrame->GetRectRelativeToSelf() as the initial intersection rect.wolfbeast2018-12-22-1/+1
| | | | Tag #249.
* Intersection ratio should be 1.0 for zero-area intersections.wolfbeast2018-12-22-1/+7
| | | | Tag #249
* Revise lifetime management of IntersectionObservers.wolfbeast2018-12-22-26/+46
| | | | Tag #249
* Merge pull request #911 from trav90/code-cleanupMoonchild2018-12-16-1782/+0
|\ | | | | Remove AndroidMediaDecoder
| * Remove AndroidMediaDecoder and friendstrav902018-12-16-1782/+0
| | | | | | | | They are no longer supported and don't work with newer Android versions anyway.
* | Issue #910 part 3. Throw SyntaxError from Location::SetProtocol on URI parse ↵wolfbeast2018-12-16-1/+4
| | | | | | | | failures.
* | Issue #910 part 2. Strip ':' and anything following it from the string ↵wolfbeast2018-12-16-1/+8
| | | | | | | | passed to the location.protocol setter.
* | Issue #910 part 1. Don't navigate when location.protocol is set to anything ↵wolfbeast2018-12-16-0/+17
|/ | | | other than http or https.
* Make HTMLOptionsCollection::mSelect into a strong reference.wolfbeast2018-12-14-59/+8
| | | | The cycle collector makes weak references like this obsolete.
* Do not report resource-timing subdocument loads triggered by that subdocument.wolfbeast2018-12-14-0/+3
|
* Clear CanvasShutdownObserver::mCanvas when the canvas goes away.wolfbeast2018-12-13-2/+11
| | | | | | | | | This is fallout from Bug 1167235 - Use a fast method of double buffering for canvas. It is possible for the CanvasRenderingContext2D to be destroyed while we're in the middle of the call to nsObserverService::NotifyObservers() for shutdown. This leaves the shutdown observer with a dangling pointer to the canvas, so this patch explicitly clears the pointer when the context goes away.
* Root parameter dictionaries in AesTask::Init().wolfbeast2018-12-13-3/+3
|
* Clear weak pointers in VTT shutdown observers.wolfbeast2018-12-11-3/+14
|
* Merge pull request #902 from trav90/media-workMoonchild2018-12-09-823/+21
|\ | | | | Remove the old MP3Parser and reorganize remaining MP3 code
| * Add missing #includetrav902018-12-08-0/+1
| |
| * Move the MP3 code to it's own directorytrav902018-12-08-4/+18
| |
| * Remove separate MP3 name spacetrav902018-12-08-7/+2
| | | | | | | | It is no longer needed.
| * Remove the old MP3FrameParsertrav902018-12-08-812/+0
| | | | | | | | All former users of the old MP3 parsing code are gone, so we can now just remove the parser itself as well.
* | Remove files and comments related to MP4Readertrav902018-12-08-222/+1
| | | | | | | | The MP4Reader no longer exists in the tree.
* | Remove Rust files from media gteststrav902018-12-08-22/+0
|/
* Merge pull request #897 from trav90/telemetry-cleanupMoonchild2018-12-03-25/+0
|\ | | | | Remove MediaTelemetryConstants
| * Remove MediaTelemetryConstantstrav902018-12-02-25/+0
| |
* | Remove DirectShowReader Part 4: Teststrav902018-12-01-4/+3
| |
* | Remove DirectShowReader Part 3: Directoriestrav902018-12-01-2750/+0
| |
* | Remove DirectShowReader Part 2: Build systemtrav902018-12-01-8/+0
| |
* | Remove DirectShowReader Part 1: Conditional codetrav902018-12-01-41/+3
|/
* Remove VR Hardware input support (gamepad type)wolfbeast2018-11-27-27/+1
| | | | This resolves #881