d153290fd6
nits: Fix a whole bunch of typos all over the place
2024-01-17 23:59:19 +01:00
b5f28f6caa
wgpu: Add support for Stage3D anisotropic filtering
...
Instead of binding every supported sampler combination
and selecting the correct index in our AGAL builder,
we now determine the correct sampler on the wgpu side,
and create one sampler binding per texture slot.
2024-01-16 18:08:28 -05:00
3c115238fd
fix(deps): update rust dependencies
2024-01-16 01:13:38 +01:00
e0a22b56ce
fix(deps): update wasm-bindgen
2024-01-15 17:24:02 -05:00
57f6963133
render: Implement PixelBender Operation::Select (renamed from Loop) ( #14503 )
2024-01-11 01:04:50 +00:00
a328deab11
fix(deps): update rust dependencies
2024-01-08 02:59:04 +01:00
47db84473a
avm2: Improve handling of Stage3D profile
...
We now validate the passed in profile, and return the selected profile
from 'Context3D.profile'. We don't yet alter the available
registers/textures based on the profile.
2024-01-07 22:34:33 +01:00
217585daa8
chore: Delegate `[lints]` in `Cargo.toml` of all packages to the workspace
2024-01-05 11:28:19 +01:00
91d0520a43
fix(deps): update rust dependencies
2024-01-01 02:03:54 +01:00
badac4176c
chore: Stop allowing previously buggy lint `clippy::extra_unused_type_parameters`
...
The corresponding Clippy issue was fixed long ago.
2023-12-25 19:20:53 -05:00
216a8f3b14
fix(deps): update rust dependencies
2023-12-25 10:09:07 +01:00
837143bb12
naga-agal: Fix handling of sampler overrides
...
After some testing, and looking at OpenFL, I believe I've
determined the correct behavior for AGAL sampling:
Each time a Context3D.setProgram or Context3D.setSamplerStateAt
call is made, the sampler config for the used texture slot(s)
is updated with the new wrapping/filter behavior. For setProgram,
this comes from all of the 'tex' opcodes used within the program.
However, when the 'ignoresampler' flag is set in a 'tex' opcode,
the setProgram call does *not* override the existing sampler config.
As a result, that program will sample with the behavior determined
by the most recent setSamplerStateAt or setProgram call involving
the used texture slot(s).
Previously, we were always overriding the opcode sampler config
with the values from Context3D.setSamplerStateAt. However, I didn't
realize that the order of the calls matter, so none of my tests ended
up observing the effect of 'ignoresampler'.
We now need to process AGAL bytecode twice - a quick initial
parse to determine the sampler configs (which need to be updated
when we call 'setProgram'), and a second time when to build the
Naga module (which needs to wait until we have the vertex attributes
available, which can be changed by ActionScript after setting
the program).
2023-12-15 22:05:30 -07:00
bf9cf92af8
render/wgpu: Update wgpu to 0.18.0, naga to 0.14.1, naga_oil to 0.11.0, egui to 0.24.1
2023-12-14 01:36:42 +01:00
3a033f0f28
fix(deps): update rust dependencies
2023-12-11 15:51:12 +01:00
44c4080ae4
wgpu: Fix overly strict assertion for setVertexBufferAt
...
The offsets for bound attributes can overlap, which caused
us to compute the total size incorrectly. Fix this check
to take overlapping into account.
2023-12-08 23:53:39 +01:00
40b92e69f7
fix(deps): update rust dependencies
2023-12-05 21:08:16 +01:00
a0c33382bf
fix(deps): update wasm-bindgen
2023-12-04 12:08:22 +01:00
7289a8bd95
naga-agal: Zero-extend input vectors for cross-product
...
This matches the behavior of Flash Player
2023-11-30 01:15:21 +01:00
78c34c1be9
naga-agal: Fix using matrix opcodes with indirect source register
...
In indirect mode, We need to increment the indirect_offset value
to load subsequent matrix rows.
2023-11-30 00:59:47 +01:00
ca25f82900
chore: Bump wasm-bindgen to 0.2.89 (0.2.88 got yanked)
2023-11-28 21:43:41 -05:00
6da223a875
fix(deps): update rust dependencies
2023-11-27 09:44:47 +01:00
2a83274c7f
fix(deps): update rust dependencies
2023-11-19 21:09:42 -05:00
f2b02acf44
chore: Appease clippy-beta
2023-11-13 20:26:50 +01:00
32c65e6460
fix(deps): update rust dependencies
2023-11-13 09:10:26 +01:00
f03b6307a8
core: Don't invalidate when setting filters to the same value
2023-11-10 22:24:45 +01:00
7110be920c
Suport uploading ATFTextureData::CompressedRawAlpha
...
Fortunately, this is very simple - we can just take
the DXT5 data and upload it directly to the wgpu texture.
2023-11-10 02:00:40 +01:00
249454885b
render: In `ShapeTessellator`, set `mask_index_count` to `None` in `flush_draw` before aborting
2023-11-09 20:22:25 +01:00
acadb2b2d3
fix(deps): update wasm-bindgen
2023-11-07 12:18:50 +01:00
1377a5b07b
fix(deps): update rust dependencies
2023-11-01 00:45:43 +01:00
46320c6238
wgpu: Round down Context3D sample count to nearest power of 2 ( #13762 )
2023-10-29 00:18:52 +00:00
b5097445e6
render: Add support for Context3DTextureFormat.COMPRESSED_ALPHA
...
This is our first non-rgba texture format (it uses Bc3RgbaUnorm).
ATF files store these textures in a very convoluted way - fortunately,
the 'dds2atf' tool is open-source, which allowed me to figure out
how to decode the texture back to a DXT5/DXT1 texture.
2023-10-23 11:18:27 -04:00
cd01bfe7d6
fix(deps): update rust dependencies
2023-10-23 02:55:11 +02:00
b427940431
fix(deps): update rust dependencies
2023-10-16 09:51:22 +02:00
9506a66803
nit: Clean up Cargo.toml files slightly (remove unnecessary `version =` maps)
2023-10-15 22:39:18 +02:00
18796a67fe
fix(deps): update rust dependencies
2023-10-10 02:15:59 +02:00
2fd6cd4147
naga-agal: Stub anisotropic filters in opcodes
...
We had already stubbed these when set from ActionScript, but were
were panicking when they were selected directly in the AGAL bytecode.
2023-10-05 15:38:21 +02:00
c93020e729
wgpu: Correctly set format when rendering to Stage3D texture
...
This fixes a panic when trying to render a texture with a type
other than `TextureFormat::Rgba8Unorm`
2023-10-04 22:47:17 -04:00
6394b29962
nagal-agal: Fix detection of writing a scalar to a destination
...
We weren't covering all cases, which resulted in an attempt to
AccessIndex on a scalar, or write a scalar directly to a vector
local.
2023-10-05 02:12:28 +02:00
428e89af73
fix(deps): update rust dependencies
2023-10-04 19:19:51 -04:00
d32b19e350
render: Implement Context3DTextureFormat::RgbaHalfFloat
2023-10-03 18:05:46 -04:00
d76306d5ae
render: Fix incorrectly reading U24 in ATF parsing
...
We should be reading a big-endian U24.
2023-09-30 20:29:49 -04:00
d2f0787c2c
naga-agal: Fix Opcode::Rcp implementation
...
MathFunction::Inverse doesn't work on vectors.
2023-09-30 19:59:21 -04:00
2dcd783261
render: Disable wgpu unused shader output check when possible
...
Since wgpu hasn't yet released a version with this feature, I manually
backported it to the 0.17 branch.
This doesn't work on Windows (HLSL), but works on all other platforms.
2023-09-30 17:59:36 -04:00
742ee4052d
naga-pixelbender: Implement Opcode::LogicalAnd
2023-09-30 15:58:06 -04:00
ec7a8ac645
avm2: Add support for Stage3D bytearray/compressed textures ( #13180 )
2023-09-28 03:03:30 +00:00
7e4e4d4c5b
fix(deps): update rust dependencies
2023-09-25 11:14:39 +02:00
01b1e36ef5
wgpu: Rename DepthBuffer to StencilBuffer and stop providing depth_ops ( #13302 )
2023-09-24 20:19:14 +00:00
37e2c06034
fix(deps): update rust dependencies
2023-09-18 00:00:45 +02:00
1c0ae0b6d8
render: Bump wgpu to 0.17 and naga to 0.13
2023-09-07 18:49:59 -04:00
dabc2d56b9
fix(deps): update rust dependencies
2023-09-04 10:25:24 +02:00
ccf42c3614
naga-pixelbender: Stub out Opcode::Loop
...
Some experimentation with Pixel Bender Studio shows
that Opcode::Loop has a 23-byte payload. I haven't tried to
figure out how to interpet the payload yet, but we can now
skip over the opcode instead of bailing out entirely.
2023-08-30 17:36:31 -04:00
520045de46
wgpu: Disable Context3D.setRenderToTexture MSAA on WebGL
...
This is not supported under webgl, and results in a panic
"Tex storage 2D multisample is not supported" if we try to
do it.
2023-08-28 13:33:32 +02:00
4e75c3890a
fix(deps): update rust dependencies
2023-08-28 07:39:35 +02:00
19ad492eaa
avm2: Implement Graphics.cubicCurveTo and CUBIC_CURVE_TO
2023-08-27 20:17:43 +02:00
eb2afb19c4
render: Add DrawCommand::CubicCurveTo
2023-08-27 20:17:43 +02:00
076977cc75
render: Rename DrawCommand::CurveTo to DrawCommand::QuadraticCurveTo
2023-08-27 20:17:43 +02:00
158beaffa5
wgpu: Bail out early when trying to set empty program constants
...
There's nothing to do in this case, and we want to avoid trying
to construct a `NonZeroU64` for the size.
2023-08-27 13:53:40 -04:00
96505a7bde
video,chore: Bump h263-rs git refs (and update Cargo.lock fully while there)
2023-08-26 22:18:14 +02:00
2f4c7d2fe3
fix(deps): update rust dependencies
2023-08-21 08:59:07 +02:00
1f4bfcf10d
fix(deps): update rust dependencies
2023-08-18 23:32:15 +02:00
93f08d7738
chore(deps): lock file maintenance rust dependencies
2023-08-14 23:58:45 +02:00
1a4a808c23
render: Make Filter::calculate_dest_rect work in Twips
2023-08-12 13:00:27 +02:00
da090f0bd7
render: Move filter size calculation from wgpu to render/swf
2023-08-12 13:00:27 +02:00
a4b179e1e0
chore(deps): lock file maintenance rust dependencies
2023-08-08 23:19:16 +02:00
11a472d40b
render: fix jpeg cmyk to rgb ( #12592 )
...
Co-authored-by: tompro <tomas.prochazka@apertia.cz>
2023-08-06 20:47:07 +03:00
8031f9dd3a
webgl: Apply color transform on gradient records
...
As described in #10517 , Flash Player applies color transforms on
each gradient record, and then blends between those transformed colors.
2023-08-05 10:28:40 +03:00
6a975cdc5b
render/wgpu: Add WgpuRenderBackend<SwapChainTarget>::recreate_surface().
2023-08-04 21:04:29 +02:00
ab27c8b4ee
render: Calculate correct bounds for shapes containing curves
2023-08-02 18:43:33 +02:00
477ff400a7
chore(deps): lock file maintenance rust dependencies
2023-08-02 13:36:13 +02:00
006393c581
chore: Appease nightly clippy
2023-08-01 22:44:00 +03:00
6d6d665332
wgpu: Use correct (shifted) zero value in DisplacementMapFilter shader
2023-08-01 12:54:56 +02:00
583caa3389
avm2: Implement DisplayObject.blendShader ( #12238 )
2023-07-26 23:25:26 +00:00
1c1a4aee18
wgpu: Don't use blur pass scaling (it was observed to not be needed)
2023-07-26 13:03:11 +02:00
3e63a88c34
chore(deps): lock file maintenance rust dependencies
2023-07-25 08:23:28 +03:00
5f6d8d72ea
render: Skip unreasonably large bitmaps
...
Some SWFs report unreasonable bitmap dimensions, and trying to reserve
enough capacity for them always fails. To avoid panics, skip those bitmaps.
Fixes #1191
Fixes #2759
Fixes #10701
2023-07-24 09:37:37 +03:00
0cf6a68d05
wgpu: Make the blur filter support fractional sizes fairly accurately
2023-07-24 04:44:22 +02:00
cd9efb2dfa
wgpu: Ignore zero width/height Stage3D scissor rect
...
This matches Flash's behavior, and prevents wgpu from panicking.
2023-07-23 17:05:27 -04:00
b8562bdb91
naga-pixelbender: Implement Opcode::Fract
2023-07-23 16:34:38 -04:00
5a5206c8df
wgpu: Correctly pass PixelBender int params to shader
...
The shader expects an array of i32s, but we were writing
bitcasted f32s to the array.
2023-07-23 13:22:40 -04:00
77c0dac64b
webgl: Implement pixel snapping
2023-07-22 00:34:17 +02:00
4410d666ed
avm2: Implement Bitmap.pixelSnapping
2023-07-22 00:34:17 +02:00
289f73c85f
core: Add pixel snapping (default auto) to Bitmaps, and force it for cacheAsBitmap
2023-07-22 00:34:17 +02:00
3607a5fb23
naga-pixelbender: Implement Opcode::LessThanEqual
2023-07-20 20:09:07 -04:00
b201e19cc7
wgpu: Don't create an extra fresh texture for applying filters to a CAB
2023-07-20 23:19:20 +02:00
400ff32723
wgpu: Implement a naive DisplacementMap filter
2023-07-20 12:41:30 +02:00
715abd8985
avm2: Use invisible mask descendants for mouse picking and rendering
...
The 'DisplayObject.visible' flag seems to be ignored for mask objects
(and their descendants) - mouse picking still takes them into account,
and they get rendered as part of the mask.
This commit changes display object rendering and AVM2 mouse picking
to ignore DisplayObject.visibility when dealing with mask objects/
descendants.
2023-07-19 23:18:28 +02:00
b1ba144166
wgpu: Implement Bevel filter
2023-07-19 22:59:33 +02:00
5597535471
wgpu: Store wgpu::Instance in Descriptors
2023-07-19 21:36:30 +02:00
030c3c50e8
wgpu: Make `request_adapter_and_device` take in wgpu::Instance by ref
2023-07-19 21:36:30 +02:00
a35ea3d9f2
wgpu: Sum up all center blur pixels before multiplying
2023-07-17 23:33:30 +02:00
18e343c6ad
wgpu: Shift blur uvs over to the left of the kernel
2023-07-17 23:33:30 +02:00
8ecd9d0181
wgpu: Precompute left_weight in blur shader
2023-07-17 23:33:30 +02:00
13eda1361a
wgpu: Use push constants for blur buffer, when supported
2023-07-17 23:33:30 +02:00
0076d65865
wgpu: Speed up blur filter by about 50%
2023-07-17 23:33:30 +02:00
268403faad
chore(deps): lock file maintenance rust dependencies
2023-07-17 09:55:56 +03:00
cd2d01bf6f
wgpu: Fix crash when not using push constants
2023-07-16 00:31:50 +02:00
ee7412d733
wgpu: Fix off-by-one in blur size calculation
2023-07-15 01:49:57 +02:00
3ea67668c0
render: make Context3D renderers fully GC-agnostic
...
The 'gc_arena' dependency was only used to manipulate the `GcCell`s
containing the vertex and fragment shaders; replacing these by a
reference to a plain old `Cell` means tha the Context3D traits and
types do not need to interact with GC'd object anymore.
As a knock-on effect, we can also remove the `Activation` parameter
from most of the `Context3DObject` methods.
2023-07-14 16:06:36 -06:00
765ceb70a0
chore: Bump h263-rs git ref to latest master
2023-07-14 14:44:23 +02:00
5c13d323b3
wgpu: Implement the undocumented ability to disable compositeSource for glow/shadow
2023-07-12 18:51:11 +02:00
c9a241f12b
wgpu: When sampling outside of blur texture, use 0
2023-07-12 18:51:11 +02:00
6f98b3d596
wgpu: Saturate alpha in glow shader
2023-07-12 18:51:11 +02:00
8cb6e5bdf0
wgpu: Implement drop shadow filter
2023-07-12 18:51:11 +02:00
aab03290be
wgpu: Add extra set of UVs to glow filter, to support separate blur rects
2023-07-12 18:51:11 +02:00
aba2394732
wgpu: Respect alpha for glows
2023-07-12 18:51:11 +02:00
1aa2ac28c2
wgpu: Slight opimitsation to blur filter by not needing to calculate full_width per fragment
2023-07-12 18:51:11 +02:00
30e7a1eac6
core: Account for filter size changes when drawing cacheAsBitmaps
2023-07-12 18:51:11 +02:00
b30c522f9c
wgpu: Implement glow filter
2023-07-12 18:51:11 +02:00
f0607d6a1f
chore(deps): lock file maintenance rust dependencies
2023-07-11 22:42:51 +02:00
f5b4fbce77
Upgrade to new `gc-arena` API ( #11182 )
...
* core: add temporary, ruffle-internal copy of `gc-arena` crate
This will allow bumping the upstream `gc-arena` version while
reexporting our own version of the old `GcCell` API, so that
Ruffle's code can be gradually migrated.
Once the migration is done, this crate should be removed.
* core: bump `gc-arena` to kyren/gc-arena#56
Add back the removed `GcCell` to our internal facade crate
* core: bump `gc-arena` to current master
This bump renames `Gc::allocate` to `Gc::new`
* core: rename `GcCell::allocate` to `GcCell::new`, to match `Gc`
* core: bump gc-arena to (slighly after) v0.3.1
Add typedefs for old `*Context` names in the gc-arena facade crate
* core: replace uses of `CollectionContext<'_>` by `&Collection`
* core: Add `gc()` convenience method for `*Context` and `Activation` types
This allows shortening most instances of `[activation.]context.gc_context`
to `activation.gc()` or `context.gc()` (but not all instances, because of
borrowck) Note that this doesn't actually do these shortenings to avoid
major code churn.
2023-07-09 17:04:25 -04:00
9d3e96e3a4
render: Implement more Pixelbender opcodes, and fix matrices ( #11896 )
2023-07-09 16:54:26 +00:00
b6fae8e214
wgpu: Cache Program3D bind group layout
...
The bind group layout only depends on the texture registers
(and 2D/cubemap type) accessed by the fragment shader, not on
the runtime texture bound with Context3D. This means that we can
build and cache it when we compile the AGAL program to a Naga
module.
Since the bind group layout is used for the overall pipeline, I've
refactored the shader caching code into `ShaderPairAgal`, which
holds both the vertex and fragment shader bytecode, and compiles
both in the `compile` function.
2023-07-09 12:39:10 -04:00
e9f4a92073
wgpu: Only log 'unsupported' warning once per filter type
...
This eliminates the console spam produced when an object with
an unsupported filter was onscreen.
2023-07-07 22:38:35 +02:00
90dbcfca5d
render: Fix writing to varying register in vertex shader
2023-07-06 11:18:39 -04:00
852367b09b
wgpu: Use a vec4f for pixelbender ZEROED_OUT_OF_RANGE_MODE_INDEX
...
This could really be a single f32 (or even a boolean), but wgpu
on web requires a minimum size of 16 bytes for uniforms.
2023-07-05 21:23:22 +02:00
3006356410
render: Suppress clippy::arc_with_non_send_sync for now
2023-07-04 20:50:46 +02:00
f22bef99b4
core: Fix some Clippy lints on the latest nightly
2023-07-04 20:50:46 +02:00
146d565c02
render: Add Filter::scale() and use it to scale filter values (currently just blur_)
2023-07-03 20:13:40 +02:00
4ac13de2f3
wgpu: Switch PASS_SCALES to be accumulative scales, for better calculation of size upfront
2023-07-03 20:13:40 +02:00
13b1078142
chore(deps): lock file maintenance rust dependencies
2023-07-03 08:28:13 +03:00
26101ae979
render: Add 'renderdoc' feature/module for manual frame capture
...
This feature is disabled by default. When enabled, you can use
`ruffle_render::renderdoc::begin_frame_capture` and
`ruffle_render::renderdoc::end_frame_capture` to manually trigger
a RenderDoc frame capture (if Ruffle wasn't launched by RenderDoc,
this logs an error).
This is very useful when debugging Stage3D/PixelBender bugs, as you
can produce a capture containing only the relevant graphics calls.
2023-07-02 13:59:19 -04:00
6e1f00edf2
render: Implement more PixelBender features ( #11800 )
2023-07-01 13:50:43 -04:00
a5915a9bdb
swf: Make Color Copy
2023-06-29 20:32:36 +02:00
97ad60e66a
wgpu: Pad out the BlurUniforms
2023-06-28 22:42:31 +02:00
7e96a2fdad
wgpu: Blur size is whole kernel width, minus one and half it for the shader
2023-06-28 22:42:31 +02:00
b90bebc65d
wgpu: Blur filter is fallible, if nothing happened then just blit
2023-06-28 22:42:31 +02:00
c0e932422f
wgpu: Change blur to 2-pass separable box blur, multiplied by quality
2023-06-28 22:42:31 +02:00
447cd2a5f6
wgpu: Move create_filter_vertices to FilterSource::vertices
2023-06-28 22:42:31 +02:00
fd0d414d02
wgpu: Add FilterSource::for_entire_texture convenience method
2023-06-28 22:42:31 +02:00
1096d7d35e
wgpu: Create a FilterSource for common arg set, and removed need for #[allow(clippy::too_many_arguments)]
2023-06-28 22:42:31 +02:00
7f4d855de9
wgpu: Move Surface::apply_filter to Filters::apply, removing need for a surface at all
2023-06-28 22:42:31 +02:00
7c751d6be8
wgpu: Implement Default for TexturePool
2023-06-28 22:42:31 +02:00
4f6bac7840
wgpu: Refactor out filters into their own files
2023-06-28 22:42:31 +02:00
ad8457b54d
wgpu: Remove 'TextureWrapper.format' field
...
We can already get the foramt from the wgpu::Texture
2023-06-28 16:26:19 -04:00
54dd160897
wgpu: Remove 'width' and 'height' fields from Texture
...
This is already stored in wgpu::Texture
2023-06-28 13:00:49 -04:00
96e3d36a6e
wgpu: Don't mark the filters as needing push constants
2023-06-26 23:18:45 +02:00
e37958351e
wgpu: Removed most uniforms from filters, they're not used or needed
2023-06-26 22:07:52 +02:00
3057aa59dd
chore(deps): lock file maintenance rust dependencies
2023-06-25 21:53:00 -04:00
f62dd17284
fix(deps): update rust crate indexmap to v2
2023-06-25 20:27:33 -04:00
8c40e97ab5
wgpu: nit: Fix incorrect label in blur filter
2023-06-26 00:00:03 +02:00
fd4eaef2b6
wgpu: Don't create cacheAsBitmap for textures that are too large
2023-06-25 22:58:22 +02:00
628d8f66ad
render: Add filters to BitmapCacheEntry
2023-06-25 21:58:46 +02:00
44c870ba91
wgpu: Move run_copy_pipeline to utils
2023-06-25 21:58:46 +02:00
4dde126e64
wgpu: Take in a &wgpu::Texture not our own texture as we don't need omore than that
2023-06-25 21:58:46 +02:00
0c71b83e36
wgpu: Don't perform copy inside Surface::apply_filter
2023-06-25 21:58:46 +02:00
08dfbc96be
wgpu: Slight changes to blur filter
2023-06-24 16:55:40 +02:00
f43560ab88
render: Switch to providing cacheAsBitmap entries all at once on submit frame
2023-06-24 01:42:04 +02:00
36aee1efcf
wgpu: Specialise render_offscreen_for_cache to render directly to the texture
2023-06-24 01:42:04 +02:00
2594453831
render: Add render_offscreen_for_cache for more optimised CacheAsBitmap
2023-06-24 01:42:04 +02:00
fe742194b1
core: Implement cacheAsBitmap behaviour
2023-06-24 01:42:04 +02:00
TÖRÖK Attila