Sharing DBus connections can interfere with other modules (e.g., watching
connections on multiple threads concurrently might introduce race conditions).
This patch changes |nsDBusHandlerApp| to use it's private connection to
the session bus to avoid interference.
Currently if we have transition-property: 'all' and trigger a transition
on the 'color' property we will end up generating a transition on
-webkit-text-fill-color even if that property is disabled.
However, when we later call StyleAnimationValue::ToValue() in
nsTransitionManager::UpdateTransitions() to see if there are any transitions we
need to cancel, the comparison for currentValue != anim->ToValue() will pass
(since, as of the first patch in this patch series, ToValue() returns a null
value) so we end up cancelling the transition as soon as we create it).
Nevertheless, we will still trigger the warning introduced in the first patch
in this series when we call ToValue().
This patch stops us from creating transitions in the first place (and hence
triggering the warning). It also removes the code that suppresses transition
events for transitions on disabled properties since we should no longer be
generating such transitions in the first place (unless the pref is switched
while the transition is in motion which is probably not worth worrying about).
Note that we only test if the property is enabled for all content. This is
consistent with what we do throughout animation code including the existing
code in nsTransitionManager which iterates through shorthand sub-properties
using CSSPROPS_FOR_SHORTHAND_SUBPROPERTIES with the flag
nsCSSProps::eEnabledForAllContent.
The test case in this patch doesn't actually fail without this change, all it
does it trigger the warning in StyleAnimationValue::ToValue() introduced
in the first patch in this series. It's still a useful regression test however,
particularly if we later upgrade the warning in StyleAnimationValue::ToValue()
to a fatal assertion.
MozReview-Commit-ID: H9swDKLyiOf
I have verified that without the fix in the first patch in this series this
test fails, but passes with the fix applied.
MozReview-Commit-ID: JmncnapbVLa
Written purely with sed, over .h and .cpp files in layout/.
(While this wasn't explicitly reviewed, I'm considering it as r=dholbert
based on the request in comment 47 in the bug.)
MozReview-Commit-ID: 6Q0F4ViOyjJ
Written purely with sed, over .h and .cpp files in layout/.
(While this wasn't explicitly reviewed, I'm considering it as r=dholbert
based on the request in comment 47 in the bug.)
MozReview-Commit-ID: 6D6A0XqArES
Written purely with sed, over .h and .cpp files in layout/.
(While this wasn't explicitly reviewed, I'm considering it as r=dholbert
based on the request in comment 47 in the bug.)
MozReview-Commit-ID: 5gQYEEY6jJ1
This moves the declaration of a method that was incorrectly inserted between:
// BOX LAYOUT METHODS
and:
// END OF BOX LAYOUT METHODS
MozReview-Commit-ID: AbboPVCM4Gx
Written purely with sed, over .h and .cpp files in layout/, except for
the removal of one comment change in nsSliderFrame.cpp.
MozReview-Commit-ID: 2jpzaby0S9r
This is a manual subset of changes written with sed, over .h and .cpp
files in layout/.
This also renames a static method on nsSprocketLayout.
Note that nsFlexContainerFrame and nsRangeFrame also have IsHorizontal
methods that are not renamed here, but this can be found to be
relatively safe because none of the IsHorizontal methods are virtual.
MozReview-Commit-ID: Jsdy7I4Q7mX
This is a manual subset of changes written with sed, over .h and .cpp
files in layout/.
It also includes manual changes in accessible/.
MozReview-Commit-ID: D1nwxJmUzc9
This is a manual subset of changes written with sed, over .h and .cpp
files in layout/, with additional manual indentation fixes.
MozReview-Commit-ID: weRWGmQLjh
This is a manual subset of changes written with sed, over .h and .cpp
files in layout/. It's a subset because there is also a Selection
method called IsCollapsed, which is not changed here.
MozReview-Commit-ID: 9JgnPv0Hkff
