Bug 1366620 - Use `unicode-bidi: embed` to fix reversed hebrew placeholders. r?gandalf MozReview-Commit-ID: 1tTkps8a96N

Bug 1217238 - Reduce time precision when javascript.options.privacy.resistFingerprinting is on. This patch is adapted from Tor bug 1517. To offer some protection against timing attacks by JS content pages, in this patch we round the various time-exposing APIs (such as Date and Event.timeStamps) to the nearest 100 ms when the pref "privacy.resistFingerprinting" is on. MozReview-Commit-ID: eGucM9nGTn

Bug 1217238 - Regression tests for reducing precision of time exposed by Javascript. This patch is adapted from Tor bug 1517. Test that the following javascript codes will return timeStamps that are rounded to 100ms. performance.now() new Date().getTime() new Event("").timeStamp new File([], "").lastModified new File([], "").lastModifiedDate.getTime() audioContext.currentTime * 1000 canvasStream.currentTime * 1000 video.currentTime * 1000 audio.currentTime * 1000 The first five codes are also tested in simple workers and nested workers, created before and after the pref is on, to ensure that the pref has correctly propagated. MozReview-Commit-ID: CuoxmGRrBnm

Bug 1367679. P5 - add a gtest to test chaining between different promise types. MozReview-Commit-ID: 8l3clGN3pWd

Bug 1367679. P4 - specialize the type of mCompletionPromise according to whether chaining is supported. MozReview-Commit-ID: 8dPXiGl5njE

Bug 1367679. P3 - pass mCompletionPromise to InvokeCallbackMethod(). This reduces duplicate code by handling promise chaining in one place. MozReview-Commit-ID: 474T5hvf9oM

Bug 1367679. P2 - overload InvokeCallbackMethod() according to whether promise-chaining is supported. This patch fix InvokeCallbackMethod() which should return null if promise-chaining is not supported. Before this patch, it could return non-null if one of the resolve/reject callbacks returns a MozPromise while the other not. MozReview-Commit-ID: 7YKNvRKEHQx

Bug 1367679. P1 - refactor InvokeCallbackMethod() to deal with one concern at a time. InvokeMethod() handles optional arguments. InvokeCallbackMethod() handles optional return value. MozReview-Commit-ID: AyT6TEKRqbs

Backed out 3 changesets (bug 1368674) for Android test failures. a=merge Backed out changeset ff3c813fcdea (bug 1368674) Backed out changeset 0d9bb636b9a9 (bug 1368674) Backed out changeset 1d0202706572 (bug 1368674) MozReview-Commit-ID: CrCFYIEDH4o

Backed out 4 changesets (bug 1365419) for android mochitest failures in test_bug583533.html a=backout Backed out changeset d82aa6aec2c1 (bug 1365419) Backed out changeset 9aafdd1ca0bd (bug 1365419) Backed out changeset 261eecf8497f (bug 1365419) Backed out changeset 1d9b50fadc05 (bug 1365419) MozReview-Commit-ID: HkhXf1Dkkbk

Bug 1367930: adjust ssrc's and encondings as simulcast answerer. r?bwc MozReview-Commit-ID: EPdIWF5nn7u

Bug 1366941 - Checking LayersIPCChannel type when using recycled textureClient; r?nical MozReview-Commit-ID: AJUJIC2tkHC

Bug 1367293 - Drop ServoComputedValuesWithParent entirely. r?birtles MozReview-Commit-ID: DGWNHiXoOlH

Bug 1367293 - Don't get parent style for GetComputedKeyframeValuesFor. r?birtles MozReview-Commit-ID: 7WjsO7P2QGz

Bug 1367293 - Don't get parent style for ComputeAnimationValue. r?birtles MozReview-Commit-ID: 4TskjmoKchD

Bug 1367293 - Get parent style from the target element. r?birtles We can get the parent style from element tree in servo side. The parent style has been already restyled. E.g.; When we get target element style with nsComputedDOMStyle::GetStyleContext, the function flushes styles so that the parent style has also restyled. When we call Servo_GetComputedKeyframeValues, it's called from a SequentialTask, that means all elements have been already restyled. Unfortunately we can't assert that the parent style is not stale to check the parent element has no dirty descendant bit since Servo_GetComputedKeyframeValues is called in a SequantialTask that is processed before post traversal, that means elements still have the dirty bit to update nsStyleContext in the post traversal. MozReview-Commit-ID: AC2ZRyzk3eu

Bug 1367293 - Early return from ValueFromStringHelper() if the target element is not associated with any documents. r?birtles MozReview-Commit-ID: 1mcCb5txtB4

Bug 1367293 - Explicitly cast nsStyleContext* to nullptr. r?birtles In subsequent patches, the servo version of KeyframeUtils::ApplySpacing() changes fifth argument to const ServoComputedValues*. If we still use nullptr for the fifth argument in caller side, compilers can't determine whether nsStyleContext* or const ServoComputedValues* should be used. MozReview-Commit-ID: 1bE2cA7gRdi

Merge m-c to autoland, a=merge MozReview-Commit-ID: AiJ1z1wSNbR

Bug 1272371 - Update multiplexed input stream seek behaviour. r=froydnj This patch is originally from :baku and updates the seek behaviour of nsMulitplexInputStream. This fixes an issue with absolute seeking where if a seek is performed and the correct offset reached, the members of the multiplex stream were not updated to reflect this. MozReview-Commit-ID: 7lEvuXxzYkI