Here we call StartRebuildAllStyleData from BeginProcessingRestyles (much
like patch 9 and EndProcessingRestyles). But we will later also call it
from the code that handles a root element font size change when we have
'rem' units. That's because it's fine to *start* the rebuild process in
the middle of processing the queue of pending restyles. (We have to end
after the whole process is done, though, in order to avoid wanting to
destroy the old rule tree while we still have style contexts referencing
it.)
We only call StartRebuildAllStyleData in this case when we're processing
our primary restyle queue (mPendingRestyles), not the animation restyles
(to be removed in bug 960465) or the animation-only restyles, since a
rebuild-all should be processed (in terms of animation phases, or in
terms of having an animation-only update before it) like a normal
restyle. (This isn't true for the 'rem' unit restyle, which could
happen during any sort of update.)
This moves the code that finishes the rebuild-all process into
EndProcessingRestyles(), which is part of the main restyling codepath.
Patch 7 ensures that we'll always get to EndProcessingRestyles in this
case, when we're going through the normal ProcessPendingRestyles()
codepath rather than the special DoRebuildAllStyleData() codepath (which
will be removed later in this patch series).
This fixes one of the omissions in the rebuild-all codepaths (where it
incorrectly differs from the regular ProcessPendingRestyles codepath).
Note that the explicit FlushOverflowChangedTracker() is no longer needed
because that's part of EndProcessingRestyles.
(This will all get refactored more substantially in the following
patches.)
This adds a member variable that is currently only used within a single
function, but that function will be split apart so that different parts
of it can be called from different places within ProcessPendingRestyles.
This is the variable that says we *need to* rebuild style data. Since
the next patch will introduce a variable that says we're *currently*
rebuilding all style data, renaming this one makes things clearer.
Part of this refactoring involves the ability to start the rebuild-all
process within the processing of restyles. This means we can't pass
parameters directly from RebuildAllStyleData into DoRebuildAllStyleData.
So this continues storing the hints as member variables a little bit
deeper into the process.
(I tried to move in a different direction in this patch queue, and store
these hints in mPendingRestyles, for the root element. But that broke
layout/style/test/test_counter_style.html and
layout/style/test/test_font_loading_api.html, and I didn't want to
figure out why. It would be somewhat better in the long run, since
currently these hints will get processed if we do a rebuild-all on a
RestyleTracker other than mPendingRestyles, which can happen if we have
'rem' units and have a root element font size change in the
animation-only update or in mPendingAnimationRestyles.)
The patches in this series refactor the process of rebuilding all style
data (RestyleManager::RebuildAllStyleData and
RestyleManager::DoRebuildAllStyleData) so that the process of rebuilding
all style data uses the existing restyle processing loops in
ProcessPendingRestyles. (Rebuilding all style data is what we do when
we need to throw away the rule tree because something has invalidated
the cached data in it.) This removes (increasing, especially with bug
960465 coming) code duplicated between the two codepaths, fixes some
omissions from the separate rebuild-all codepath, and (more immediately)
allows fixing lifetime issues of ReframingStyleContexts objects in bug
1110277 so that we can have a single ReframingStyleContexts for all of
the restyle processing in each restyle processing operation. In other
words, the goal is to change the rebuild-all process from a separate
codepath to a few variables that modify the way ProcessPendingRestyles
works (and make it do the extra work).
This is just a small first step in that process, which moves one piece
of code from a chunk of duplicated and to-be-removed code into a chunk
of code that will be preserved.
This patch is not needed to fix the bug, but it seems like it's probably
desirable. It's not needed for this bug because
MaybeReframeForBeforePseudo and MaybeReframeForAfterPseudo are already
called (by ElementRestyler::RestyleChildren) on only the first and last
continuation or ib-split sibling with the same style. So this patch
should only actually change anything for cases like a block-in-inline
split whose initial inline part is inside of a ::first-line (where
different parts of the block-in-inline split chain have different style).
Since the symptom of this bug is (once patch 6 is in the tree) only
causing extra reframes, it can only be tested using the new API (from
bug 1115691) for observing reframes. I confirmed that the test for this
bug fails without the patch and passes with the patch (as noted by the
removal of its todo annotation).
For some kinds of changes we need to update the layer tree even though there is
no change to style. For example, if an animation is paused via the Web
Animations API, we need to remove the animation from the layer even though the
style will not change.
This patch detects such changes by making ElementRestyler check for an
out-of-date animation generation on layers. This is complicated by the fact that
we currently maintain *two* animation generation numbers: one for the set of
animations and one for the set of transitions, but we only have *one* animation
generation number on each layer. This is a known issue (bug 847286).
As a result, until bug 847286 is fixed, we need to be careful to compare against
the greater of the two numbers.
Until we get rid of animation phases in bug 960465, we need to ensure
we're producing style data for the correct animation phase. This makes
this optimization slightly less beneficial until then.
Bug 931668 added the eRestyle_ForceDescendants case for all cases that
go through RebuildAllStyleData, but there is another case that calls
DoRebuildAllStyleData directly, and we should cover that too.
This is needed because when we're rebuilding the rule tree, we need to
create new style contexts for all frames.
I don't know of any bugs caused by this, but it seems worth fixing.
This is just moving one bit of data from the pres context without any
logic change. But given the other refactoring, it seems to make more
sense here now.
The MOZ_DEBUG_RESTYLE_STRUCTS environment variable can be set to a comma-
separated list of style struct names. When restyle logging is enabled,
this will cause the style context tree -- showing cached style struct
pointers for those structs specified -- to be logged before each
individual restyle is processed. It will also show the struct pointer
values involved when swapping structs between style contexts.
For example, set MOZ_DEBUG_RESTYLE_STRUCTS=Font,UserInterface to show
the cached nsStyleFont and nsStyleUserInterface pointers on the style
contexts involved in the restyle process.
Set the MOZ_DEBUG_RESTYLE environment variable and every restyle will have
detailed logging printed to stderr. By default, restyles for animations are
not logged; you can include them by also setting MOZ_DEBUG_RESTYLE_ANIMATIONS.
If you wish to limit restyle logging to a particular change, you can call
nsPresContext::StartRestyleLogging() and nsPresContext::StopRestyleLogging()
at appropriate points. (You might want to add a couple of helper methods
temporarily on nsIDocument and then expose them to your page with Web IDL
to make them easier to call.) You do not need to have set MOZ_DEBUG_RESTYLE
for this to work.
Set the MOZ_DEBUG_RESTYLE environment variable and every restyle will have
detailed logging printed to stderr. By default, restyles for animations are
not logged; you can include them by also setting MOZ_DEBUG_RESTYLE_ANIMATIONS.
If you wish to limit restyle logging to a particular change, you can call
nsPresContext::StartRestyleLogging() and nsPresContext::StopRestyleLogging()
at appropriate points. (You might want to add a couple of helper methods
temporarily on nsIDocument and then expose them to your page with Web IDL
to make them easier to call.) You do not need to have set MOZ_DEBUG_RESTYLE
for this to work.
This is needed because patch 1 and patch 3 did not add a mechanism that
allows restyling pseudo-elements, which would be substantially more work
and very little use (since the only case in which they have style
attributes is for our internal use on the ::-moz-color-swatch
pseudo-element).
RestyleUndisplayedChildren does not need the same fix because it's only
used on elements, and never on pseudo-elements or anonymous boxes.
We don't do struct swapping if either the old or new style contexts are
shared, since we can't guarantee that such a change will be valid for
all frames that use the shared style context.
--HG--
extra : rebase_source : 10f7c90422de23b119287ba478d36433423440da
This matches patch 2, and also fixes an incorrect use of eRestyle_Self
on the parents of pseudo-elements in order to restyle those
pseudo-elements, where it would not previously have been effective.
This should all be temporary, since this code can go away with bug
960465, when animation phases are removed.
This changes the coalescing behavior during the animation-only style
flush by doing tree-based coalescing between the style updates required
by animations and those required by transitions, rather than doing
animations and transitions separately.
Note that both the old and the new code update all
animating/transitioning styles rather than only the throttled styles,
though we should fix that eventually as noted in the FIXME comment in
the code (but only for the existing caller, and not for the new one to
be introduced in bug 960465).
Note that this depends for its correctness on the previous patches to
make the restyling process exact.
The test changes are because the effects of bug 1031688 are changed by
the change in coalescing. In the old code, we updated transition styles
before animation styles; in the new code we do a single pass over the
tree, which for the relevant test means updating animations on the
parent before transitions on the child, which changes the effects of the
bug.
This will be necessary when we use the restyle tracker for the
animation-only style flush, because animation-only style flushes need to
update *only* the animation style data and no other style data. Thus
using the RestyleTracker for animation-only style flushes requires that
we do this.
This changes the coalescing behavior during the animation-only style
flush by doing tree-based coalescing between the style updates required
by animations and those required by transitions, rather than doing
animations and transitions separately.
Note that both the old and the new code update all
animating/transitioning styles rather than only the throttled styles,
though we should fix that eventually as noted in the FIXME comment in
the code (but only for the existing caller, and not for the new one to
be introduced in bug 960465).
Note that this depends for its correctness on the previous patches to
make the restyling process exact.
The test changes are because the effects of bug 1031688 are changed by
the change in coalescing. In the old code, we updated transition styles
before animation styles; in the new code we do a single pass over the
tree, which for the relevant test means updating animations on the
parent before transitions on the child, which changes the effects of the
bug.
This will be necessary when we use the restyle tracker for the
animation-only style flush, because animation-only style flushes need to
update *only* the animation style data and no other style data. Thus
using the RestyleTracker for animation-only style flushes requires that
we do this.
This means ResolveStyleWithReplacement will replace only the rules in
the nsRestyleHint rather than always replacing both the animations rule
and the transitions rule. (This will become even more useful later when
we add the ability to replace the style attribute rule.)
This is the start of a refactoring of the existing code in
CommonAnimationManager, nsTransitionManager, and nsAnimationManager,
which will be removed in later patches (after patches in between cause
this code to be used instead of the current code).
But it also handles interesting cases more correctly, such as cases
where style context parents are unusual.
FIXME: It would be nice to add tests for these cases, but they're mostly
pretty obscure (e.g., a transition of 'transform' that's affected by
inheriting 'transform' from ::first-line), so I'm not treating it as a
priority right now.
FIXME: Adjust for eRestyle_Force if bug 931668 lands first.
Later patches will add such types that are like eRestyle_Self but that
only replace the style attribute, animation rule, or transition rule.
Later patches will combine the coalescing of restyling between the
miniflush performed for animations and for transitions into a single
RestyleTracker, which will coalesce the work better. This patch changes
the API exposed for doing that so that the coalescing patch will contain
only the internals.
This affects the correctness of transitions that take over from a
running animation. (In the current code this can happen on a single
element; once the cascading changes in bug 960465 are complete it can
only happen via inheritance.)
This version of the patch changes to do the test using opacity rather
than transform, since testing using transforms encountered issues
related to bug 1031688: the presence of phantom transitions due to the
interaction of the computed value rules for transforms distinguishing
between values that the interpolation rules consider identical. (These
problems only appear after patch 24 in this bug changes the coalescing
order between a parent with animations and a child with transitions so
that the parent is handled before the child, instead of transitions
being handled before animations.)
The final two added tests fail without the patch and pass with the patch.
(With the opacity version, the third to last test also fails without the
patch, probably due to the value not having yet been sent to the layer.
This was not an issue pre-patch with the original test using transform,
though. I haven't tested that this happened with patch 0, though, so it
might have been the issue patch 0 fixes.)
- RestyleManager::RecomputePosition now calls
StickyScrollContainer::PositionContinuation with the first continuation
or ib-split-sibling of the passed frame, as PositionContinuations will
update all continuations at once, and expects the first continuation
to be passed in.
This is needed to avoid hitting the assertion "Why did this not get
handled while processing mRestyleRoots?" in CollectRestyles, because we
have a restyle posted for a frame that has the NEEDS_FRAME bit flag set.
- OverflowChangedTracker::AddFrame now accepts an enumerated type parameter to
indicate if the overflow areas of children have changed (CHILDREN_CHANGED),
the overflow areas of the children have changed and the parent have changed
(CHILDREN_AND_PARENT_CHANGED), or if only the transform has changed
(TRANSFORM_CHANGED).
- OverflowChangedTracker::Flush no longer falls back to calling
nsIFrame::UpdateOverflow when a frame lacks a PreTransformOverflowAreas
property.
- Added an additional change hint, nsChangeHint_ChildrenOnlyTransform, which
results in TRANSFORM_CHANGED being passed in to
OverflowChangedTracker::AddFrame.
- In nsIFrame::FinishAndStoreOverflow, the passed in overflow is now stored as
the InitialTransformProperty for elements that are IsTransformed().
- Partially corrected Bug 926155, by only calling
OverflowChangedTracker::AddFrame on parents of the sticky element during
StickyScrollContainer::UpdatePositions, using CHILDREN_CHANGED.
- OverflowChangedTracker::AddFrame now accepts an enumerated type parameter to
indicate if the overflow areas of children have changed (CHILDREN_CHANGED) or
if the transform has changed (TRANSFORM_CHANGED).
- OverflowChangedTracker::Flush no longer falls back to calling
nsIFrame::UpdateOverflow when a frame lacks a PreTransformOverflowAreas
property.
- Added an additional change hint, nsChangeHint_ChildrenOnlyTransform, which
results in TRANSFORM_CHANGED being passed in to
OverflowChangedTracker::AddFrame.
- In nsIFrame::FinishAndStoreOverflow, the passed in overflow is now stored as
the InitialTransformProperty for elements that are IsTransformed().
- Partially corrected Bug 926155, by only calling
OverflowChangedTracker::AddFrame on parents of the sticky element during
StickyScrollContainer::UpdatePositions, using CHILDREN_CHANGED.
This also changes the functionality a little bit to track independent
per-property mutation counts and independent "content active" status.
--HG--
extra : rebase_source : e69b8e7a95d36720bd38d74f0789ede603e58a09
This patch does the following:
* Move nsIFrame::IntrinsicSize to mozilla::IntrinsicSize so that it can
be forward-declared.
* Move a number of templated inline nsLayoutUtils methods to nsIFrame.
* Use mozilla::layout::FrameChildListID instead of the
nsIFrame::ChildListID typedef in nsLayoutUtils.h.
* Move nsReflowFrameRunnable to its only user, nsProgressMeterFrame.cpp.
* Make a number of functions requiring nsIFrame.h out-of-line.
* Remove the nsIFrame.h #include from nsLayoutUtils.h and add it to the
places which require it implicitly.
This depends on bug 898333 in order to avoid causing:
TEST-UNEXPECTED-FAIL | chrome://mochitests/content/chrome/dom/tests/mochitest/chrome/test_focused_link_scroll.xul | Assertion count 1 is greater than expected range 0-0 assertions.
due to the assertion:
###!!! ASSERTION: Shouldn't be trying to restyle non-elements directly: '!aContent || aContent->IsElement()', file ../../../layout/base/nsStyleChangeList.cpp, line 62
The assertion count change in layout/generic/crashtests/571995.xhtml is
expected because it changes us from having 7 of:
###!!! ASSERTION: Shouldn't be trying to restyle non-elements directly: '!aContent || aContent->IsElement()', file ../../../layout/base/nsStyleChangeList.cpp, line 62
with the stack:
mozilla::ElementRestyler::CaptureChange(nsStyleContext*, nsStyleContext*, nsChangeHint) [layout/base/nsChangeHint.h:191]
mozilla::ElementRestyler::RestyleSelf(nsRestyleHint) [layout/base/RestyleManager.cpp:2304]
to only having one. This is expected since this patch changes
RestyleSelf to only call CaptureChange for the first continuation or
block-in-inline sibling.
(This is part of the patch stack making change hints apply across all
continuations and block-in-inline siblings. In this case, however, the
change hint only needs to apply once, globally.)
This patch fundamentally makes three changes:
(1) Change the way RestyleSelf is used so that it is called in a loop,
over all of the same-style continuations and block-in-inline
siblings.
(I had a note here reminding myself:
TODO: Don't set hints for descendants!
but I no longer remember what it meant.)
(2) Change the traversal of children to traverse all of the children of
the items traversed in (1).
(3) When traversing children, skip children that are continuations,
since we already reached them in (1) and (2).
A later patch will change the RestyleSelf loop to copy for the later
continuations rather than repeating the work. This will mean reverting
many of the RestyleSelf changes contained here.
Some of the pieces marked temporary will be removed in bug 828312 patch
11, and the rest should hopefully be removed in bug 918064.
Reftest inline-3.html fails because handling the sticky element's margin correctly will take more work; similarly, border/padding on containing-block continuations won't be handled quite right. Reftest column-contain-1a fails because some of the anonymous blocks inside an nsColumnSetFrame have 0 height.
This replaces newContext with mFrame->StyleContext(), which is a valid
replacement since all of the replacements are inside a test that we
don't have a framechange hint. If we have a framechange hint, then
mFrame still has its old style context.
This avoids doing work that is unnecessary; if we're going to reframe
anyway, the frame destruction will clear the undisplayed content list,
and we'll then rebuild it when we recreate the frame (if needed).
It is also a prerequisite for patch 4, which replaces some uses of
newContext with mFrame->StyleContext(). These are the same as long as
we don't have a framechange hint, since when we do have a framechange
hint we leave the old style context on the frame to avoid violating
invariants. This patch ensures that all such uses replaced in patch 4
are actually inside a check that we don't have a framechange hint.
(Most of them already were.)