Current strategy: everything from the top
Things to look at first are the manual changes:
- added tools/rewrite_includes.py
- removed -Idirectives from BUILD.gn
- various compile.sh simplifications
- tweak tools/embed_resources.py
- update gn/find_headers.py to write paths from the top
- update gn/gn_to_bp.py SkUserConfig.h layout
so that #include "include/config/SkUserConfig.h" always
gets the header we want.
No-Presubmit: true
Change-Id: I73a4b181654e0e38d229bc456c0d0854bae3363e
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/209706
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Hal Canary <halcanary@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: Florin Malita <fmalita@chromium.org>
Currently, shader effects are applied via layers for non-atomic draw fragments.
This is incorrect, as shaders on layer paints are ignored.
Update to always propagate the shader to leaf node paints.
Change-Id: I59f5a5b874e87827f61844969550600fa39e22a3
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/205931
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Two issues:
1) Empirically (per AE's model), only solid layers transform the effect space.
All other layer types implement their effects ignoring the layer xform.
2) Due to the effect deferral mechanism in sksg, shader effects are
sometimes applied with a different transform than intended.
To address #1, tweak the layer builder logic to attach the effects
before/after the transform node, depending on the layer type.
For #2, track the CTM active when the shader effect is attached and undo
any intervening transformations via a local matrix when the shader is
actually applied to SkPaint.
Change-Id: I29b5a4df3b223e7da9dfc36b0d7d40055d192ac0
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/205922
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Currently, gradients are PaintNode subclasses - which limits their
applicability to leaf DAG nodes.
In order to support generalized gradient/shader effects:
* introduce a new Shader base class
* refactor gradients as Shader subclasses
* introduce ShaderPaint (to support current Gradient-as-paint use)
* introduce ShaderEffect (to support future Gradient-as-effect use)
Tangentially:
* rename SkSGPaintNode.h -> SkSGPaint.h
* consolidate sksg::Color into SkSGPaint.h
Change-Id: I94ba949f4504c09cfde4a4f030d927411fdd66a2
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/205263
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Introduce sksg::BlendModeEffect and use to wrap Skottie layers based on
their "bm" property.
Depending on the presence of non-trivial layer blend modes and the nature
of the destination buffer (fully transparent vs. unknown/pre-filled), we
may now need to render animation frames into a separate layer for correct
compositing.
Track the presence of non-trivial layer blend modes such that we only incur
this extra layer overhead when needed. Also allow clients to pass a "drawing
to fully transparent buffer" hint such that we can avoid the extra layer even
when blend modes are present.
Change-Id: Iaf645878666da4349d0bef8890bbecad23a0aa9b
Reviewed-on: https://skia-review.googlesource.com/c/194840
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Introduce RenderNode::nodeAt(const SkPoint&) as the entry point for the hit-testing API.
This is backed by a onNodeAt() virtual, which gets dispatched throughout the render DAG,
and normally stops at the first leaf Draw node in encounters.
To support the implementation, introduce a GeometryNode::contains(const SkPoint&) API.
This is backed by a onContains() virtual, overridden in each concrete geometry class.
Expose nodeAt() on sksg::Scene, and add some basic unit tests.
Change-Id: I0c8abd9d1e51ecf2d8b4dd699f325cd636e21084
Reviewed-on: https://skia-review.googlesource.com/c/191296
Commit-Queue: Florin Malita <fmalita@chromium.org>
Reviewed-by: Mike Reed <reed@google.com>
Image filters cannot let descendent damage pass through, as they may
transform their inputs arbitrarily.
Introduce the notion of "damage override" SG nodes (nodes which block
descendent damage and replace it with their own), and update
ImageFilterEffect to opt in for this behavior.
Also make InvalidationController damage dispatch optional (such that we
don't accumulate inval rects unnecessarily).
TBR=
Change-Id: Ie9f7d32ecdcc77dbe6ea009aed9ce269b3649a04
Reviewed-on: https://skia-review.googlesource.com/c/191005
Reviewed-by: Florin Malita <fmalita@chromium.org>
Commit-Queue: Florin Malita <fmalita@chromium.org>
This reverts commit 1f43a4359d.
Reason for revert: Android build failures.
Original change's description:
> [skottie] Add drop shadow support
>
> Introduce the machinery required for general image filters in SkSG +
> a concrete drop shadow image filter effect.
>
> Wire it all up with Skottie to support drop-shadow layer effects.
>
> Change-Id: I98e9669852f58ba6481439a7fda4a56ec6c59b8a
> Reviewed-on: https://skia-review.googlesource.com/c/190426
> Reviewed-by: Mike Reed <reed@google.com>
> Commit-Queue: Florin Malita <fmalita@chromium.org>
TBR=fmalita@chromium.org,reed@google.com
Change-Id: I31d38ed4d4a15b77d1d1218b2677a891978332cb
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/190981
Reviewed-by: Florin Malita <fmalita@chromium.org>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Introduce the machinery required for general image filters in SkSG +
a concrete drop shadow image filter effect.
Wire it all up with Skottie to support drop-shadow layer effects.
Change-Id: I98e9669852f58ba6481439a7fda4a56ec6c59b8a
Reviewed-on: https://skia-review.googlesource.com/c/190426
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
ScopedRenderContext tries really hard to avoid unnecessary RenderContext
initializations.
This is most likely unneeded because
a) the structure is quite lightweight
and
b) all ScopedRenderContext instantiation sites imediately call setters
which require a writable context (assuming no-op args are uncommon)
Let's get rid of that over-engineered gunk.
Bug: skia:
Change-Id: Ieedc20c63c66b4d43744359e881c20639654040c
Reviewed-on: https://skia-review.googlesource.com/146761
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Use the deferred paint override mechanism, similar to opacity.
Change-Id: I78fa7f5d73ef333480ec72b0cb663819b1de2404
Reviewed-on: https://skia-review.googlesource.com/146527
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
- plumb a RenderContext RenderNode::render() argument to track pending
opacity
- defer opacity application until we can determine whether a layer is
required (group with multiple children) or the opacity can be pushed
onto the draw paint (for single/atomic draws)
Bug: skia:
Change-Id: I962ba08bad780395d5d738307bde986e9efa502b
Reviewed-on: https://skia-review.googlesource.com/146445
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
* relocate all SkSG-related files under modules/sksg/
* fix various tidbits to make non-sksg builds possible
* drop obsolete SampleSGInval.cpp
Change-Id: I54e6c5bb1a09f45030fa8d607b3eb3f7cba78957
Reviewed-on: https://skia-review.googlesource.com/130025
Commit-Queue: Florin Malita <fmalita@chromium.org>
Reviewed-by: Mike Klein <mtklein@google.com>
