Commit graph

187 commits

Author SHA1 Message Date
Martin Robinson
05fb2ef6ee Merge DisplayListEntry into DisplayItem
We don't really need two levels of abstraction for every element in the
DisplayList. This simplifies the complexity of the data structure in
preparation for providing documentation and properly handling scrolling
roots.
2016-04-22 10:28:27 -07:00
Peter
b0ad3ddf5b Updated euclid to version 0.6.5
And updated existing usages of Matrix4 to use Matrix4D<T>
2016-04-17 18:39:49 +01:00
Michael Howell
6c9efbf383 Take transform:translate into account when computing clipping regions.
Note that this only works for translation; a more general fix would
require major changes to how display lists work.

Closes #10431?
2016-04-12 14:39:13 -07:00
Patrick Walton
3518472f7c gfx: When dumping the display list, mention whether each layer scrolls
its overflow area.
2016-04-06 17:58:59 -07:00
Patrick Walton
22a1c112d5 gfx: Simplify complex clipping regions as we construct them.
This allows WebRender to correctly render complex clipping regions that
can be reduced to single rounded rectangles. WebRender still can't
render rounded rectangles with arbitrary intersections yet, but this
allows it to handle many more cases.

Closes servo/webrender#241.
2016-03-31 19:36:40 -07:00
Per Lundberg
4cb4cc93e4 Renamed TComputedValues to ComputedValues
This is a followup to #10210, and a continuation of #10185.
2016-03-29 23:30:13 +03:00
Per Lundberg
c2ad084713 ComputedValues is now ServoComputedValues
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.
2016-03-27 10:55:43 +03:00
Bobby Holley
5c749127cc Introduce traits for style structs and computed values. 2016-03-24 11:49:46 -07:00
Patrick Walton
0006c1923a gfx: Allow images to be shipped to the WebRender thread without shipping
over the data as well.

WebRender doesn't need the data, as it acquires it separately.

About a 50%-100% improvement in display list building time on
browser.html.
2016-03-22 14:49:47 -07:00
Anthony Ramine
46c505e47b Bump serde to 0.7 2016-03-17 13:52:39 +01:00
Patrick Walton
28a71c4820 Make border radii clip their contents.
Needed for browser.html rounded corners.
2016-03-14 18:32:55 -07:00
Martin Robinson
474033c39f Simplify the display list slightly
Now that WebRender uses an Iframe display item, we do not need the Noop
item for the non-WebRender path. We can simply reuse the Iframe display
item. Also remove the layer_id member from the LayeredItem struct, as
it is unused.
2016-03-10 14:33:22 -08:00
Jack Moffitt
f7aa6a32e9 Unrevert the fix for hit testing through shadows
This was originally #9428 but got accidentally reverted during rebase in

Fixes #9865.
2016-03-03 17:04:55 -07:00
Patrick Walton
dd5c574d7f gfx: Switch offsets to an FNV hasher.
SipHash traffic was showing up high in the profile.

Unfortunately, this necessitated a manual implementation of the serde
traits.
2016-03-03 13:09:53 -08:00
Patrick Walton
940bff1f9c gfx: Stop cloning clipping regions so much.
Clipping regions can contain vectors and so can be expensive to copy.
2016-03-03 13:09:50 -08:00
Patrick Walton
4233e0f163 gfx: Box stacking contexts to minimize memmove traffic.
`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.
2016-03-03 13:09:48 -08:00
Martin Robinson
e7019f2721 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.
2016-03-01 14:50:07 -08:00
Jonathan Schuster
205336bf2e Move util::range into its own crate 2016-02-25 21:06:10 -05:00
Glenn Watson
c0531c312f Add WebRender integration to Servo.
WebRender is an experimental GPU accelerated rendering backend for Servo.

The WebRender backend can be specified by running Servo with the -w option (otherwise the default rendering backend will be used).

WebRender has many bugs, and missing features - but it is usable to browse most websites - please report any WebRender specific rendering bugs you encounter!
2016-02-18 10:35:29 +10:00
Anthony Ramine
290694b27e Move util::cursor to style_traits 2016-02-16 00:50:01 +01:00
Anthony Ramine
966da85791 Derive HeapSizeOf for GradientDisplayItem 2016-02-13 11:13:20 +01:00
Anthony Ramine
55dcee94be Bump heapsize to 0.3 2016-02-13 11:13:19 +01:00
Anthony Ramine
cb5cd8d881 Say farewell to in-tree HeapSizeOf 2016-02-04 22:03:32 +01:00
Patrick Walton
32a68a43ec Make hit tests against box shadows always fail.
Closes #9268.
2016-01-25 16:30:04 -08:00
Brandon Fairchild
dce7c7bb69 Remove multiple unused imports in gfx
Fixes #9342.
2016-01-16 14:16:03 -05:00
rohan.prinja
1f02c4ebbb task -> thread 2016-01-10 17:58:13 +09:00
Martin Robinson
d720452698 Use DisplayListSection everywhere
DisplayListSection, StackingLevel, and BackgroundAndBorderLevel all
represent pretty much the same thing, a particular section of the
display list. Instead of maintaining three enums which do the same
thing, just use DisplayListSection everywhere. It's a superset of the
other two and this change will make it easier to flatten the DisplayList
in the future for WebRender.
2016-01-08 10:19:03 -08:00
Patrick Walton
3103b03262 gfx: Eagerly transform clips into ClippingRegion::max() if possible.
This helps WebRender look for useless clips and optimize them out.
2016-01-06 19:48:14 -08:00
Johannes Linke
6b215f38ee Fix a bunch of clippy lints 2016-01-02 23:27:15 +01:00
Bobby Holley
47059d2d26 Separate style+layout and layout-specific wrapper functionality.
This patch does a number of things, unfortunately all at once:
* Hoists a large subset of the layout wrapper functionality into the style system.
* Merges TElementAttributes into the newly-created TElement.
* Reorganizes LayoutData by style vs layout, and removes LayoutDataShared.
* Simplifies the API for borrowing style/layout data.

There's still more to do to make the style system usable standalone, but
this is a good start.
2015-12-29 11:50:03 -08:00
Brandon Fairchild
637afecec9 Move LayerKind and ScrollPolicy enums to gfx_traits
This also moves LayerId and LayerProperties to gfx_traits.

Fixes #8836.
2015-12-20 20:43:31 -05:00
Ms2ger
208c05f944 Fix warnings. 2015-12-04 10:31:19 +01:00
bors-servo
68922e0ac2 Auto merge of #8797 - cheba:rect-contains, r=Manishearth
Replaced rect_contains_point with Rect.contains()

This is a proposed in servo/servo#8791 clean up.

Fixes #8791.

<!-- Reviewable:start -->
[<img src="https://reviewable.io/review_button.png" height=40 alt="Review on Reviewable"/>](https://reviewable.io/reviews/servo/servo/8797)
<!-- Reviewable:end -->
2015-12-04 04:01:44 +05:30
Alexander Mankuta
96a8b442b8 Replaced rect_contains_point with Rect.contains() 2015-12-03 16:15:34 +02:00
Alexander Mankuta
546d2f564f Replaced ZERO_RECT with Rect::zero() 2015-12-03 15:39:30 +02:00
bors-servo
8d8ffa8265 Auto merge of #7696 - mskrzypkows:PaintContext_Units, r=mbrubeck
Draft. Change PaintContext rects to TypedRects #7023

I created draft. I'm not sure if we need any units conversion in PaintContext. There is also strange 'clear' method, we use PagePx origin and ScreenPx size is it OK?

<!-- Reviewable:start -->
[<img src="https://reviewable.io/review_button.png" height=40 alt="Review on Reviewable"/>](https://reviewable.io/reviews/servo/servo/7696)
<!-- Reviewable:end -->
2015-11-07 00:22:59 +05:30
Martin Robinson
e5b2feda3f Remove DisplayListBuildingResult
Always produce a DisplayList when processing nodes for display list
construction. StackingContexts are now added to the positioned content
section of DisplayLists. This makes the code a bit simpler and opens up
the possibility of producing a StackingContext in another section of
the DisplayList. This doesn't change behavior, but is a cleanup
prerequisite for proper inline stacking context support.
2015-11-04 12:52:14 -08:00
Maciej Skrzypkowski
2a7927d074 Change PaintContext rects to TypedRects #7023
Needs update to_nearest_pixel method in app_unit module.
Argument of to_nearest_pixel should be:
ScaleFactor<PagePx, ScreenPx, f32>
2015-11-04 08:39:15 +01:00
Martin Robinson
c1a38e240a Mix stacking contexts into the positioned content list
Sometimes positioned content needs to be layered on top of stacking
contexts. The layer synthesis code can do this, but the current design
prevents it because stacking contexts are stored in a separate struct
member. In order to preserve tree order, mix stacking contexts into the
positioned content list, by adding a new StackingContextClass
DisplayItem. Such items do not have a base DisplayItem.

In some ways this simplifies the code, because we no longer have to
have a separate code path in the StackingContextLayerCreator.

Fixes #7779.
Fixes #7983.
Fixes #8122.
Fixes #8310.
2015-11-03 17:47:39 -08:00
Martin Robinson
ac5525aeeb Integrate iframes into the display list
Instead of always promoting iframes to StackingContexts, integrate them
into the display list. This prevents stacking bugs when
non-stacking-context elements should be drawn on top of iframes.

To accomplish this, we add another step to ordering layer creation,
where LayeredItems in the DisplayList are added to layers described by
the LayerInfo structures collected at the end of the DisplayList.
Unlayered items that follow these layered items are added to
synthesized layers.

Another result of this change is that iframe layers can be positioned
directly at the location of the iframe fragment, eliminating the need
for the SubpageLayerInfo struct entirely.

Iframes are the first type of content treated this way, but this change
opens up the possibility to properly order canvas and all other layered
content that does not create a stacking context.

Fixes #7566.
Fixes #7796.
2015-10-20 07:29:06 -07:00
bors-servo
6d52bdf4ff Auto merge of #7891 - mrobinson:display-list-paint-layer, r=pcwalton
Properly size synthesized layers

Layers that are composed of several stacking contexts that need to be
rendered on top of other layered content need synthesized layers.
Previously, these layers were placed into a stacking context that was
the same size as their parent. This patch creates a new type of
PaintLayer which simply holds a display list. The layer is sized to the
bounds of the DisplayList and its children are positioned relative to
the parent stacking context.

This will also be useful in the future, when items outside of
StackingContexts are given their own layer for render order purposes.

<!-- Reviewable:start -->
[<img src="https://reviewable.io/review_button.png" height=40 alt="Review on Reviewable"/>](https://reviewable.io/reviews/servo/servo/7891)
<!-- Reviewable:end -->
2015-10-09 17:00:21 -06:00
Paul Rouget
99cccb2193 Snap to screen pixels instead of px 2015-10-08 06:44:05 +02:00
Martin Robinson
fb80e51bd1 Properly size synthesized layers
Layers that are composed of several stacking contexts that need to be
rendered on top of other layered content need synthesized layers.
Previously, these layers were placed into a stacking context that was
the same size as their parent. This patch creates a new type of
PaintLayer which simply holds a display list. The layer is sized to the
bounds of the DisplayList and its children are positioned relative to
the parent stacking context.

This will also be useful in the future, when items outside of
StackingContexts are given their own layer for render order purposes.
2015-10-07 15:12:36 -07:00
Patrick Walton
e3b988a1fe layout: Add a field in the display list for simple border radii on box
shadows.

Only supported in WebRender (with my upcoming PR) for now.
2015-10-06 14:04:50 -07:00
Martin Robinson
553f1fc192 Rework how StackingContexts are dynamically added to layers
StackingContexts are added to layers when it is necessary to maintain
their ordering on top of other layered StackingContexts. Instead of
tracking the information about a layer scattered around into different
structs, combine it all into LayerInfo. LayerInfo will be used in the
future to hold layer information for DisplayItems that are layerized
independently of StackingContexts.
2015-10-05 13:35:31 -07:00
Glenn Watson
339a3f869b Split Au type into separate crate, with minimal dependencies. 2015-10-01 07:16:11 +10:00
bors-servo
b1d9f438b7 Auto merge of #7751 - mrobinson:refactor-hit-testing, r=pcwalton
Make DisplayList hit testing a method on DisplayList

This will allow us to hit test into DisplayLists that are not directly
contained in StackingContexts, but instead are children of PaintLayers.
It also makes things slightly simpler.

<!-- Reviewable:start -->
[<img src="https://reviewable.io/review_button.png" height=40 alt="Review on Reviewable"/>](https://reviewable.io/reviews/servo/servo/7751)
<!-- Reviewable:end -->
2015-09-30 03:04:16 -06:00
Patrick Walton
c72d0c2ed0 layout: Make the compositor rather than layout determine the position of
each iframe.

The old code that attempted to do this during layout wasn't able to work
for multiple reasons: it couldn't know where the iframe was going to be
on the page (because of nested iframes), and at the time it was building
the display list for a fragment it couldn't know where that fragment was
going to be in page coordinates.

This patch rewrites that code so that both the sizes and positions of
iframes are determined by the compositor. Layout layerizes all iframes
and marks the iframe layers with the appropriate pipeline and subpage
IDs so that the compositor can place them correctly. This approach is
similar in spirit to Gecko's `RefLayer` infrastructure. The logic that
determines when it is time to take the screenshot for reftests has been
significantly revamped to deal with this change in delegation of
responsibility.

Additionally, this code removes the infrastructure that sends layout
data back to the layout task to be destroyed, since it is now all
thread-safe and can be destroyed on the script task.

The failing tests now fail because of a pre-existing bug related to
intrinsic heights and borders on inline replaced elements. They happened
to pass before because we never rendered the iframes at all, which meant
they never had a chance to draw the red border the tests expect to not
render!

Closes #7377.
2015-09-29 09:47:28 -07:00
Martin Robinson
a2cf8e583d Make DisplayList hit testing a method on DisplayList
This will allow us to hit test into DisplayLists that are not directly
contained in StackingContexts, but instead are children of PaintLayers.
It also makes things slightly simpler.
2015-09-28 18:34:07 -07:00
bors-servo
09b970af4a Auto merge of #7710 - mrobinson:small-dl-refactor, r=pcwalton
A few small bits of refactoring to prepare for more advanced DisplayList layerization

This will be useful when PaintLayers that contain DisplayLists instead
of StackingContexts need to draw themselves. This is a simple
preparatory refactoring and doesn't change any behavior.

<!-- Reviewable:start -->
[<img src="https://reviewable.io/review_button.png" height=40 alt="Review on Reviewable"/>](https://reviewable.io/reviews/servo/servo/7710)
<!-- Reviewable:end -->
2015-09-23 17:22:31 -06:00