Two improvements.
- More testing of markers by introducing and using a ProfilerMarkerPayload
subclass that keeps track of how many instances have been created, streamed,
and destroyed. This change would have caught the defect that was in the first
landing of part 3 of this bug.
- More testing of JSON output, to check for various substrings that are
expected to be present.
FSFindFolder is deprecated as of 10.9 and apparently continues to
work...except for the case of finding the temporary directory when the
user's home directory lives on an NFS mount (!). Using
NSTemporaryDirectory does The Right Thing here, so let's use that instead.
Remove now useless self reference in OpenDatabaseOp::SendResults(). Add comments for how to call
FactoryOp::Run() safely and also add comments to all the call sites. This prevents a self destruction
while Run() (or a method that is called by Run()) is still being executed. Also fix
TransactionDatabaseOperationBase which also calls Run() directly in NoteContinueReceived().
--HG--
extra : rebase_source : dcc5f6e7e2974e8e0730687dd0dedac9d3e60060
Make PromptService and Prompt.jsm not depend as much on BrowserApp and
the main Fennec window, so it can be used from a GeckoView-based custom
tabs Activity. This approach is used because it involves a lot less work
than trying to implement a PromptDelegate for Fennec.
MozReview-Commit-ID: AEH0HgAOx2S
This means they will be copied to $PROFILE/extensions, which the sandbox allows
access to; if they are installed as temporary addons, loading frame scripts in
the content process tries to read from wherever they happen to be on disk. This
breaks running tests with a packaged build once we have full read-restrictions
for the content process sandbox.
MozReview-Commit-ID: 7ZiiM9FMXfG
--HG--
extra : rebase_source : d2cf3a2d06df9099dc6056fae351200eaa1d0ca9
Without this change, browser_update.js "resets" a preference that it never
changed to a different value, which leaks through to future tests. This was
introduced in a8fcca075fde, and appears to be a simple mistake since that change
removes a setup/teardown pref change pair, but the prefs it changes are two
different ones!
This leaked pref change leads to test failures when special powers and mochitest
are installed as non-temporary addons.
MozReview-Commit-ID: 2jx3fB1iZMx
--HG--
extra : rebase_source : 35394dda16814d80116854bd40c00c95f30d34e2
The current DOM tests for about:debugging implicitly assume that there is a
temporary addon installed before they start. This is currently always true
because mochitest and specialpowers are installed as temporary addons, but the
goal of this bug is to install them as non-temporary addons.
The specific cause of the breakage is the assumption that uninstalling an addon
removes an <li> from the addons <ul>, but when you go from one addon to zero the
entire <ul> is removed.
MozReview-Commit-ID: VJIcy17uQc
--HG--
extra : rebase_source : a398f20e9befb7e8ff307368557ae7b9c0c75fba
If we set a transform in push_stacking_context, it changes the internal
WebRender behaviour to make that stacking context a reference frame, and
things inside it are positioned differently. This is true even if the
transform is an identity transform.
In most cases we are hitting this and sending an identity transform
through, when in fact we want to be sending a None value to WebRender so
that it doesn't create reference frames. This is a partial fix, a proper
fix will be done in bug 1345577 by separating the CSS transform from the
other transforms that FrameLayerBuilder invents.
MozReview-Commit-ID: ElSs3hFMD2D
In addition to updating webrender and webrender_traits, this patch:
- bumps the euclid dependency in webrender_bindings to match webrender
- updates the Cargo.lock files and re-vendors third-party rust packages
- updates the push_yuv_image callers due to an API change in WR cset a4b9e25.
The encrypted data is nonsensical as far as the parsing of NALs is concerned.
MozReview-Commit-ID: Hm1fJf6h2S7
--HG--
extra : rebase_source : b32f56d748ad33afbd227379c1639a5aa449fb84