Collect scroll roots during the collect_stacking_context phase instead
of during display list construction. This will be useful in order to
collect containing block scroll roots as well as to give scroll roots
sequential ids in the future. This change also pulls stacking context
children out of the StackingContext struct itself, which should reduce
very slightly the memory used by the finished display list. This also
simplifies the DisplayListBuilder because it no longer has to maintain
a stack of ScrollRootIds and StackingContextIds and can instead just
rely on the program stack.
This avoids storing it in the rowgroup (which doesn't need it) and
ensures that it is done sequentially, which will be important when
rowspan support is added to this function.
Note that this does not reduce parallelism in the common case where
all rows are in the same rowgroup.
This is a step in disassociating scrolling areas from stacking
contexts. Now scroll areas are defined by unique ids, which means that
in the future stacking context will be able to contain more than one.
Simplify the way that stacking contexts are collected. Instead of
passing the StackingContextId down the tree, pass the parent
StackingContext itself. This will allow future patches to get more
information about the parent stacking context (such as location).
Also remove the return value of collect_stacking_contexts, which was
unused.
WebRender.
This happens asynchronously, just as it does in non-WebRender mode.
This functionality is a prerequisite for doing proper display-list-based
hit testing in WebRender, since it moves the scroll offsets into Servo
(and, specifically, into the script thread, enabling iframe event
forwarding) instead of keeping them private to WebRender.
Requires servo/webrender_traits#55 and servo/webrender#277.
Partially addresses #11108.
This is the first part of #10185. More to follow. I have built this locally with both servo and geckolib without errors; let's see if it succeeds on all platforms as well.
`memmove` was showing up high in the profile when concatenating and
shorting display lists. This change drastically reduces the `memmove`
cost in exchange for some minor additional allocation cost.
Flatten display list structure
Instead of producing a tree of stacking contexts, display list
generation now produces a flat list of display items and a tree of
stacking contexts. This will eventually allow display list construction
to produce and modify WebRender vertex buffers directly, removing the
overhead of display list conversion. This change also moves
layerization of the display list to the paint thread, since it isn't
currently useful for WebRender.
To accomplish this, display list generation now takes three passes of
the flow tree:
1. Calculation of absolute positions.
2. Collection of a tree of stacking contexts.
3. Creation of a list of display items.
After collection of display items, they are sorted based upon the index
of their parent stacking contexts and their position in CSS 2.1
Appendeix E stacking order.
This is a big change, but it actually simplifies display list generation.
<!-- Reviewable:start -->
[<img src="https://reviewable.io/review_button.svg" height="40" alt="Review on Reviewable"/>](https://reviewable.io/reviews/servo/servo/9756)
<!-- Reviewable:end -->
Instead of producing a tree of stacking contexts, display list
generation now produces a flat list of display items and a tree of
stacking contexts. This will eventually allow display list construction
to produce and modify WebRender vertex buffers directly, removing the
overhead of display list conversion. This change also moves
layerization of the display list to the paint thread, since it isn't
currently useful for WebRender.
To accomplish this, display list generation now takes three passes of
the flow tree:
1. Calculation of absolute positions.
2. Collection of a tree of stacking contexts.
3. Creation of a list of display items.
After collection of display items, they are sorted based upon the index
of their parent stacking contexts and their position in CSS 2.1
Appendeix E stacking order.
This is a big change, but it actually simplifies display list generation.
This fixes the border-end calculation for table rows whose borders are
collapsed with rows in different rowgroups. The border collapsing code now
uses an iterator that yields all the rows as a flat sequence, regardless of
how they are grouped in rowgroups. It gets rid of
`TableRowGroupFlow::preliminary_collapsed_borders` which was never correct.
(It was read but never written.)
This may fix#8120 but I'm not 100% certain. (I haven't managed to reproduce
the intermittent failure locally, and my reduced test case still fails but in
a different way.)
Use the PrintTree utility to improve the readability of flow tree
dumps. Blocks and fragments are now split over two dump levels, because
otherwise they are impenetrable. Also start printing the restyle damage of
fragments.
The failing `float-applies-to-*` CSS 2.1 tests never really should have
been passing in the first place; they depend on floats inside
fixed-layout tables working properly, which they don't.
Closes#6078.
Closes#6709.
Closes#6858.
absolutely-positioned elements.
This also implements a little bit of the infrastructure needed to
support for fragmentation via support for multiple positioned fragments
in one flow.
Improves Google.
Table columns should be layed out according to the 'direction' property of the
table flow, regardless of the 'direction' property of any table-row,
table-rowgroup, etc. flows.
This fixes a number of the `direction-applies-to-*` tests in the CSS2.1 test
suite.
This also simplifies `propagate_column_inline_sizes_to_child` by separating
the code used for table cells from the code for non-cell flows.
r? @pcwalton
Known issues:
* Collapsed borders do not correctly affect the border-box of the table
itself.
* The content widths of all cells in a column and the content height of
all cells in a row is the same in this patch, but not in Gecko and
WebKit.
* Corners are not painted well. The spec does not say what to do here.
* Column spans are not handled well. The spec does not say what to do
here either.
This fixes a bug in finding the top left corner of an RTL block in physical
coordinates. (The old code used the `start` point of the `position` rect,
which is not always the top left.)
It also fixes the setting of `position.start.i` in certain mixed LTR/RTL
cases.
There is still a bug related to `position.size` for RTL blocks with margins.
See the FIXME comments for details.
