This patch also makes FrameLayerBuilder responsible for setting the visible region
of a fixed-pos layer to the displayport intersected with the draw-region of
the layer. This will let us remove some displaylist code in later patches.
We add special handling to GetAnimatedGeometryRootFor for nsDisplayScrollLayers,
to match what nsDisplayList::IsFixedItem does.
Restyling margin-top and margin-bottom will typically require reflow and cause
sibling elements to be moved and invalidated, so it's not a big win to layerize
such nodes. However, layerizing them should be relatively harmless.
Restyling margin-top and margin-bottom will typically require reflow and cause
sibling elements to be moved and invalidated, so it's not a big win to layerize
such nodes. However, layerizing them should be relatively harmless.
The main change here is to have nsCSSFrameConstructor able to construct
a scrollframe for nsFieldSetFrame's inner frame.
To make this work properly (as much like normal scrolled elements, and
to be consistent with Chrome), we need another major change, which is to move
the padding from the nsFieldSetFrame to its inner frame. We do this by copying
the padding to the inner frame and ignoring the padding on the outer frame.
To get this right for percentage padding and intrinsic widths is a little
tricky. For that, we need nsLayoutUtils::IntrinsicForContainer to be able to
ignore the padding on a frame so we don't add it twice.
Overriding nsFieldSetFrame::GetScrollTargetFrame makes setting scrollTop/
scrollLeft on a <fieldset> work as expected.
* * *
Bug 261037 - A fieldset overflow:auto reftest.
The main change here is to have nsCSSFrameConstructor able to construct
a scrollframe for nsFieldSetFrame's inner frame.
To make this work properly (as much like normal scrolled elements, and
to be consistent with Chrome), we need another major change, which is to move
the padding from the nsFieldSetFrame to its inner frame. We do this by copying
the padding to the inner frame and ignoring the padding on the outer frame.
To get this right for percentage padding and intrinsic widths is a little
tricky. For that, we need nsLayoutUtils::IntrinsicForContainer to be able to
ignore the padding on a frame so we don't add it twice.
Overriding nsFieldSetFrame::GetScrollTargetFrame makes setting scrollTop/
scrollLeft on a <fieldset> work as expected.
* * *
Bug 261037 - A fieldset overflow:auto reftest.
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 assumes that the specification for how position:sticky behaves for
block-in-inline splits matches the specification for position:relative,
in other words, matches
http://www.w3.org/TR/CSS21/visuren.html#anonymous-block-level .
It's also necessary for patch 9b since the new rule for handling of
style change hints is that a style change hint applies to all
continuations and all block-in-inline siblings ("special siblings").
The change in StickyScrollContainer::GetScrollRanges is really the fix
for bug 918994, but adjusted for the change here to use block-in-inline
siblings ("special siblings") in addition to continuations.
