There were two issues:
1. We were accidentally calling `as_any` on `handle,` rather than
`handle.0`
2. Calling `as_any` can invoke the wrong implementation, depending on
what traits are in scope. We want the method implemented on the
underlying type (`RegistryData`) to be used, but if `Downcast` is
explicitly imported, then we appear to invoke it on the trait object
`dyn BitmapHandleImpl` itself (using the fact that trait objects
themselves implement `Any`). We now explicitly call the generated
method on the trait object, which avoids this issue.
`BitmapHandle` now holds `Arc<dyn BitmapHandleImpl>`.
This allows us to move all of the per-bitmap backend data into
`BitmapHandle`, instead of holding an id to a backend-specific
hashmap.
This fixes the memory leak issue with bitmaps. Once the AVM side of a
bitmap (`Bitmap`/`BitmapData`) gets garbage-collected, the
`BitmapHandle` will get dropped, freeing all of the GPU resources
assoicated with the bitmap.
This PR implements core 'stage3D' APIs. We are now able
to render at least two demos from the Context3D docs - a simple
triangle render, and a rotating cube.
Implemented in this PR:
* Stage3D access and Context3D creation
* IndexBuffer3D and VertexBuffer3D creation, uploading, and usage
* Program3D uploading and usage (via `naga-agal`)
* Context3D: configureBackBuffer, clear, drawTriangles, and present
Not yet implemented:
* Any 'dispose()' methods
* Depth and stencil buffers
* Context3D texture apis
* Scissor rectangle
General implementation strategy:
A new `Object` variant is added for each of the Stage3D objects
(VertexBuffer3D, Program3D, etc). This stores a handle to the
parent `Context3D`, and (depending on the object) a handle
to the underlying native resource, via `Rc<dyn
SomeRenderBackendTrait>`).
Calling methods on Context3D does not usually result in an immediate
call to a `wgpu` method. Instead, we queue up commands in our
`Context3D` instance, and execute them all on a call to `present`.
This avoids some nasty wgpu lifetime issues, and is very similar
to the approah we use for normal rendering.
The actual rendering happens on a `Texture`, with dimensions
determined by `createBackBuffer`. During 'Stage' rendering,
we render all of these Stage3D textures *behind* the normal
stage (but in front of the overall stage background color).
We only called `get_bitmap_pixels` when creating a `BitmapData`
for an SWF-provided `Bitmap`. We now store the initial pixels
in `Character::Bitmap`, and use them to initialize a `BitmapData`
when needed.
This lets us simplify the wgpu backend, which no longer needs
to store a `Bitmap` object. In addition to saving space for
`BitmapData` objects that lack an SWF `Bitmap`, this will make
it easier to move data from `bitmap_registry` into `BitmapHandle`
itself.
Currently, we rely on ShapeTessellator being able to get a BitmapHandle
without a RenderBackend. With the upcoming BitmapData refactor,
we will always need a RenderBackend to get a BitmapHandle, which creates
borrow-checker issues in ShapeTessellator (which is stored in a
RenderBackend).
To solve this, we split BitmapSource.bitmap into two methods -
BitmapSource.bitmap and BitmapSource.bitmap_handle. ShapeTessellator
continues to use BitmapSource.bitmap, and uses the u16 bitmap id
instead of a BitmapHandle. The BitmapSource.bitmap_handle method
is used inside each render backend to convert the id to a BitmapHandle,
avoiding borrow-checker issues.
This PR fixes a numbe of interconnected bugs:
* We weren't consistently uploading a dirty BitmapData to the render
backend before drawing to/from it.
* BitmapData.draw should *not* add a fill color - it should draw over
the current contents of the BitmapData
* After drawing to a non-transparent BitmapData, we need to manually
set the opacity back to 255 for each pixel (the drawing process
takes transparency into account, but the opacity information is
thrown away at the end).
When rendering offscreen, we want the resulting image to use
premultiplied alpha, since the image will be stored in a texture.
However, when capturing an image in the exporter or test framework,
we want to use straight alpha, so that the resulting image can
be saved as a PNG.
Previously, we incorrectly used straight alpha everywhere, resulting
in incorrect output when using BitmapData.draw with transparency.
* avm2: Implement `BitmapData.draw` for `wgpu` backend
This method requires us to have the ability to render directly to a
texture. Fortunately, the `wgpu` backend already supports this in
the form of `TextureTarget`. However, the rendering code required
some refactoring in order to avoid creating duplicate `wgpu` resources.
The current implementation blocks on copying the pixels back
from the GPU to the CPU, so that we can immediately set them in
the Ruffle `BitmapData`. This is likely very inefficient, but will
work for a first implementation.
In the future, we could explore allowing the CPU image data and GPU
texture to be out of sync, and only synchronized when explicitly
necessary (e.g. on `getPixel` or `setPixel` calls).
* Rename `with_offscreen_backend` to `render_offscreen` and use Bitmap
* Don't panic when backend doesn't implement `render_offscreen`
Reverts #7267
The image tests for the upcoming 'DisplayObject.stageRect' support
differ between Linux and Windows, so we need this support again.
To avoid the Linux filename churn that we previously encountered,
we now only include the platform and graphics backend in the filename
(e.g. `expected-linux-Vulkan`). This may result in some unexpected
'mismatched image' test failures if GHA updates to a version of Lavapipe
that changes rendering output, but this should be relatively easy to
notice.
Previously, the viewport height and width were stored in
both `Stage` and the `RenderBackend`. Any changes to the viewport
dimensions (e.g. due to window resizing) needed to be updated in both
places to keep our handling of the viewport consistent.
This PR adds a new `ViewportDimensions` type, which holds the
width, height, and scale factor. It is stored inside the
`RenderBackend` impl, and is retrieved using the newly added
method `RenderBackend.get_viewport_dimensions`. After a `Player`
has been constructed, any code that needes access to the viewport
dimensions will ultimate go through this method.
Unfortunately, `Stage` needs to use the viewport dimensions
in `build_matrices`. Therefore, any code modifying the viewport
dimensions should go through `player.set_viewport_dimensions`,
which ensures that the stage matrices are rebuilt after the render
backend is updated.
The captured WGPU texture uses premultiplied alpha.
This image gets saved as a PNG, so it should use straight alpha.
Note that all of our current image tests have 'alpha = 1.0' for all
of the pixels, so this currently has no effect.
Each render backend keeps track of a stack of BlenModes,
which are pushed and popped by 'core' as we render objects
in the displaay tree. For now, I've just implemented BlendMode.ADD,
which maps directly onto blend mode supported by each backend.
All other blend modes (besides 'NORMAL') will produce a warning
when we try to render using them. This may produce a very large amount
of log output, but it's simpler than emitting each warning only once,
and will help to point developers in the right direction when they
get otherwise inexplicable rendering issues (due to a blend mode
not being implemented).
The wgpu implementation is by far the most complicated, as we need
to construct a `RenderPipeline` for each possible
`(BlendMode, MaskState)`. I haven't been able to find any documentation
about the maximum supported number of (simultaneous) WebGPU render
pipelines - if this becomes an issue, we may need to register them
on-demand when a particular blend mode is requested.
Though https://github.com/rust-lang/rust-clippy/pull/8355 has been
merged, it seems to still report false-positives on nightly channel.
For now just fix the instances reported by stable clippy, and keep
`needless_borrow` allowed.
Adjust `common_tess` to add an additional `mask_index_count` to
draws. This is used to not render strokes when drawing a shape as
a mask stencil.
Fixes#7027.
wgpu 0.13 added `Surface::get_supported_formats`, so we can check
if a non-sRGB surface format is supported on the current platform.
Because Flash colors and blending are in sRGB space, we want a
linear surface when possible to prevent the colors from being
fiddled with by the GPU.
Based on the work in #6717, plus additional adaptions mentioned in
https://github.com/gfx-rs/wgpu/blob/master/CHANGELOG.md#wgpu-013-2022-06-30,
and more not-mentioned but required changes.
Also bump `wasm-bindgen` to `0.2.81` (along with its helper crates), as
required by the new `wgpu` version.
Note that I don't fully understand some of the required changes, notably:
* `wgpu::PresentMode::Mailbox` no longer works on my machine (Windows 11) -
The `wgpu` documentation says that `wgpu::PresentMode::Fifo` is the
only guaranteed to be supported, so I switched over to it instead.
* `self.staging_belt.recall()` doesn't return a `Future` anymore -
I assume it became synchronous so I simply removed the `executor`
from there.
Currently, all three render backends hold on texture-related
resources indefinitely (`register_bitmap` pushes to a `Vec`,
and never removes anything). As a result, the resources used
by the render backend (which may include GPU memory) will grow
over time, even if the corresponding `BitmapData` has been deallocated.
This commit adds a new `unregister_bitmap` method, which is called from
`BitmapData.dispose`. All render backs are changed to now use an
`FnvHashMap<BitmapHandle, _>` instead of a `Vec`, allowing us to
remove individual entries.
Currently, we only call `unregister_bitmap in response to
`BitmapData.dispose` - when `BitmapData` is freed by the
garbage collector, `unregister_bitmap` is *not* called.
This will be addressed in a future PR.
Since all `RenderBackend::register_bitmap_*` implementations are
identical now, move them to the default implementation of `RenderBackend`.
Also, turn `RenderBackend::register_bitmap_raw` into `RenderBackend::register_bitmap`,
which accepts a single `Bitmap` parameter.
Guarantee bitmap data after decoding has the expected length for
the given width+height+format. This eases the burden from backends
to worry about this. Pad or truncate the data if it doesn't have
the expected size.
Always use a non-sRGB texture as the render target so that color
and alpha blending are in sRGB space, matching Flash behavior.
If the surface format requires sRGB, render to an intermediate
linear buffer and copy to the surface as a final render step.
naga is currently lenient about some things which are invalid by
the WGSL spec. The Tint WGSL compiler disagreed, so these shaders
failed to compile in Chrome.
This interface handles returning a bitmap given an ID and is used
by the render backend to get the bitmap used for a bitmap fill.
This will allow for bitmap fills in the drawing API, as these will
manage their own list of bitmaps.
This also removes the MovieLibrary dependency from render backends
and will allow for better decoupling in the future.
Matrices in an SWF file store their scale/skew components in
in 16.16 format (fbits).
Split `ruffle_core::Matrix` and `swf::Matrix`. `swf::Matrix` now
stores its data as `Fixed16` instead of immediately converting to
`f32`.
Use the proper types for ColorTransform:
* Fixed8 (8.8) format for multiplicative component
* i16 format for additive component
This matches the behavior of Flash (for example, alpha only changes
in units of 1/256).
* #[derive(Collect)] should be before #[collect]
* Replace redunant `&buf[..]` with `buf`
* Changes most cases of UPPERCase to UpperCase
* Allow upper_case_acronym on most SWF types, as they are from
SWF spec/more annoying to change.
Some recent changes with lyon's API could cause Ruffle to draw extra
strokes when it sees two Move commands in a row -- an extra zero-length
stroke would appear. Let's merge adjacent Move commands in the
tessellator to avoid this issue.
(TODO: Add an example to visual_tests).
Switch to debug_assert! instead of assert! in mask checks to avoid
panicking (#1347). This will render incorrectly and still needs
to be fixed, but we should fail gracefully and avoid killing the
entire player.
Pass the movie library to `register_shape` methods so that bitmap
charcter IDs can be resolved immediately on the proper SWF.
This fixes#2037, which cause incorrect bitmaps to be used when
multiple movies were loaded.
On desktop, don't render while the window is minimized.
In the wgpu backend, don't panic if swap chain fails to grab a
texture. Instead recreate the swap chain and bail on the current
frame.
Fixes panic when minimizing with the wgpu vulkan backend (#2065).
Switch to a single render pass for the whole frame, as opposed to
a render pass per draw call. This should be a large improvement to
performance. This currently requires some unsafe to work around a
self-reference between RenderPass and CommandEncoder in Frame;
this could eventually be cleaned up by changing RenderBackend
to return a Frame object instead of using begin_frame/end_frame
pairs.
Also switch to using push constants for the transform/color
uniforms.
Add --timedemo for benchmarking, which will run the given SWF as
quickly as possible for 5000 frames or the end of the root
timeline, whichever comes first. The total duration will be output
upon completion.
Change the usage of the stencil buffer to avoid running out of
stencil bits when too many nested masks are active.
This also cleans things up on wgpu which requires us to make
pipeline states in advice; now we only need a few stencil states
for masking as opposed to hundreds.
lyon would emit a small line segment if a stroke path had
consective MoveTo commands. For the lyon backends, filter out
consecutive MoveTos and only emit the final MoveTo command.
This fixes a small dot appearing in text fields using underlines.
Nathan Adams