This reverts commit d5d8b5de90f40644f23d309c0f68d03e299334eb.
The flutter/tests/skp_generator test (running in flutter_tester on
Linux) was failing, along with some customer tests.
In https://g-issues.skia.org/issues/305780908 Skia is removing the
default SkFontMgr. Previous work consolidated all references to
txt/platform.h and this replaces those last references. I attempted to
mirror the existing functionality, which still responds to GN flags and
the target platform.
After this PR, Flutter should not be depending on the default fontmgr
(and the defines in flutter_defines.gni) will maintain that behavior
until the legacy functions/methods are deleted from Skia. There were a
few tests that I missed on an earlier PR which relied on the default
font (helper added in #47493). These tests were failing because they
were making some assertions related to TextBlobs, which didn't work if
the (now-empty) Typeface they loaded had no glyphs. Thus, I added a few
extra asserts to make sure these textblobs *had* glyphs which make the
failing tests less mysterious, should this issue crop up again.
I cleaned up Flutter's BUILD.gn file for Skia a bit, deleting unused
targets related to the font managers. This involved fixing an implicit
dependency from //third_party/glfw/ to `Gdi32.lib` on Windows.
## Pre-launch Checklist
- [x] I read the [Contributor Guide] and followed the process outlined
there for submitting PRs.
- [x] I read the [Tree Hygiene] wiki page, which explains my
responsibilities.
- [x] I read and followed the [Flutter Style Guide] and the [C++,
Objective-C, Java style guides].
- [x] I listed at least one issue that this PR fixes in the description
above.
- [x] I added new tests to check the change I am making or feature I am
adding, or the PR is [test-exempt]. See [testing the engine] for
instructions on writing and running engine tests.
- [ ] I updated/added relevant documentation (doc comments with `///`).
- [x] I signed the [CLA].
- [x] All existing and new tests are passing.
If you need help, consider asking for advice on the #hackers-new channel
on [Discord].
<!-- Links -->
[Contributor Guide]:
https://github.com/flutter/flutter/wiki/Tree-hygiene#overview
[Tree Hygiene]: https://github.com/flutter/flutter/wiki/Tree-hygiene
[test-exempt]:
https://github.com/flutter/flutter/wiki/Tree-hygiene#tests
[Flutter Style Guide]:
https://github.com/flutter/flutter/wiki/Style-guide-for-Flutter-repo
[C++, Objective-C, Java style guides]:
https://github.com/flutter/engine/blob/main/CONTRIBUTING.md#style
[testing the engine]:
https://github.com/flutter/flutter/wiki/Testing-the-engine
[CLA]: https://cla.developers.google.com/
[flutter/tests]: https://github.com/flutter/tests
[breaking change policy]:
https://github.com/flutter/flutter/wiki/Tree-hygiene#handling-breaking-changes
[Discord]: https://github.com/flutter/flutter/wiki/Chat
The root slice of a surface frame was requesting the construction of an RTree from its DisplayList when it was built from a layer tree, but since the layer tree already does branch culling, it would not benefit from any further culling during `DisplayList::Dispatch`. Further, if there are platform views present in the layer tree then they will need an RTree to accurately convey "pixel ownership" information to the platform, but the root slice lives below any and all platform views, so it is the only slice that doesn't need an RTree for that case.
The cost of having an RTree in that slice was the accumulation of information and lists of rects that would never prove useful.
The new `BUILD.gn` files in the Engine tree can't go under
`build/secondary` because Skia still has its own, and they'd be selected
first. So, this PR puts the new `BUILD.gn` files under `flutter/skia`.
The rendering tests currently always save the impeller failure images into a temporary directory in /tmp which is out of the way and might accumulate over time.
The images are now only saved when `--save-impeller-failures` is specified on the command line and they are now saved into a local sub-directory with multiple runs saving into new sub-directories to keep the failure images from getting confused with each other over time. The new image directories are named `./impeller_failure_images/NNNN/*.png`
Switch the impeller testing in rendertests to use TextFrame objects rather than TextBlob objects when drawing to an Impeller-managed destination.
Previously the DrawText tests would generate over 200 errors because none of the attribute variants would render anything at all. After this fix the number of failures in the DrawText calls is down to 1 which points out that wrapping a DrawTextFrame call in a nothing saveLayer somehow affects the glyph positioning (in a not very appealing manner).
Revisiting this project as it may help with [attribute] X [rendering op] matrix testing of Impeller. It's not an exhaustive combinatorial test of all combinations of rendering attributes, but it attempts each op with each attribute both default and non-default.
For Impeller it only tests that the rendering produced output (i.e wasn't a NOP due to missing implementation), that the output changed if the given attribute was supposed to change the output, and that it rendered within the computed bounds that the rest of the engine will be assuming.
These changes will not result in any automatic CI testing of Impeller as is since they are only run on the Metal backend which is not enabled by default. Running `out/host_[variant]/display_list_rendertests --enable-metal` manually on a Mac host, though, will run the tests through Impeller and there will be a lot of error output. Also, there is one known crash as reported in https://github.com/flutter/flutter/issues/135766
Due to https://github.com/flutter/flutter/issues/127500 , we can get in a state where enable-impeller is true but we're using Skia. We need to either fall back completely to Skia, make this configuration fatal, or remote the check
----------------
Conversion of SkTextBlobs to impeller::TextFrame objects is one of the most expensive operations in display list dispatching. While the rest of the engine and framework makes a reasonable attempt to cache the SkTextBlobs generated during paragraph construction, the design of the dl dispatcher means that these the Impeller backend will always reconstruct all text frames on each frame - even if the display list/picture that contained those text frames was unchanged.
Removing this overhead is one of the goals of https://github.com/flutter/engine/pull/45386 , however this patch is also fairly risky and will be difficult to land. As a more incremental solution, we can instead construct the impeller::TextFrame objects when performing paragraph painting and record them in the display list. This both moves the text frame construction to the UI thread and allows the framework/engine to cache unchanged text frames.
This also does not conflict with the dl_aiks_canvas patch directly, and is fine to land before or after it does. (though I'd argue we should land this first).
To compare the current performance levels, I ran the complex_layout_scroll perf test, since this is fairly text filled. On a Pixel 6 pro. Across several runs this is a fairly consistent ~1ms raster time improvement.
Fixes https://github.com/flutter/flutter/issues/133204
Closes https://github.com/flutter/flutter/issues/135057.
This is a fair bit more involved than previous changes, just due to the sheer number of implicit conversions.
Highlights:
- Made `public uint32_t argb` `private uint32_t argb_`, and added `argb()` instead.
- Added `ToSk(DlColor)` instead of using implicit conversions.
There were a bunch of places where I had to make a judgement call (particularly in tests) to keep the code a bit "messy", i.e. `DlColor(SK_RED)`, just to make the diff as small as possible and to prevent silly copy and paste bugs. I'd be open to filing a follow-up issue to reduce unnecessary wrapping.
Work towards https://github.com/flutter/flutter/issues/134969.
These are all self-contained, so I bundled them all together.
All fixes are generated by `clang-tidy --fix`, and manual search/replace if that wasn't sufficient.
Reverts flutter/engine#45418
Some google3 tests are hitting the CHECK I added in the DlSkCanvasDispatcher::drawTextFrame, which indicates that the SkParagraph code likely thinks impeller is enabled, whereas other code might be running with Skia.
Perhaps this could happen if its software rendering? It should be a fatal error on startup so we can track this down.
Conversion of SkTextBlobs to impeller::TextFrame objects is one of the most expensive operations in display list dispatching. While the rest of the engine and framework makes a reasonable attempt to cache the SkTextBlobs generated during paragraph construction, the design of the dl dispatcher means that these the Impeller backend will always reconstruct all text frames on each frame - even if the display list/picture that contained those text frames was unchanged.
Removing this overhead is one of the goals of https://github.com/flutter/engine/pull/45386 , however this patch is also fairly risky and will be difficult to land. As a more incremental solution, we can instead construct the impeller::TextFrame objects when performing paragraph painting and record them in the display list. This both moves the text frame construction to the UI thread and allows the framework/engine to cache unchanged text frames.
This also does not conflict with the dl_aiks_canvas patch directly, and is fine to land before or after it does. (though I'd argue we should land this first).
To compare the current performance levels, I ran the complex_layout_scroll perf test, since this is fairly text filled. On a Pixel 6 pro. Across several runs this is a fairly consistent ~1ms raster time improvement.
### Skia
```
"average_frame_build_time_millis": 1.497333333333333,
"90th_percentile_frame_build_time_millis": 2.038,
"99th_percentile_frame_build_time_millis": 17.686,
"worst_frame_build_time_millis": 23.095,
"missed_frame_build_budget_count": 3,
"average_frame_rasterizer_time_millis": 5.5078589743589745,
"stddev_frame_rasterizer_time_millis": 2.226343414420338,
"90th_percentile_frame_rasterizer_time_millis": 7.481,
"99th_percentile_frame_rasterizer_time_millis": 19.11,
"worst_frame_rasterizer_time_millis": 79.799,
"missed_frame_rasterizer_budget_count": 7,
"frame_count": 234,
"frame_rasterizer_count": 234,
"new_gen_gc_count": 10,
"old_gen_gc_count": 2,
```
### Impeller (ToT)
```
"average_frame_build_time_millis": 1.431575000000001,
"90th_percentile_frame_build_time_millis": 2.196,
"99th_percentile_frame_build_time_millis": 14.486,
"worst_frame_build_time_millis": 23.728,
"missed_frame_build_budget_count": 2,
"average_frame_rasterizer_time_millis": 6.536087499999999,
"stddev_frame_rasterizer_time_millis": 1.9902712500000004,
"90th_percentile_frame_rasterizer_time_millis": 9.705,
"99th_percentile_frame_rasterizer_time_millis": 14.727,
"worst_frame_rasterizer_time_millis": 17.838,
"missed_frame_rasterizer_budget_count": 1,
"frame_count": 240,
"frame_rasterizer_count": 240,
"new_gen_gc_count": 10,
"old_gen_gc_count": 2,
```
### Impeller (Patched)
```
"average_frame_build_time_millis": 1.4500167364016743,
"90th_percentile_frame_build_time_millis": 2.478,
"99th_percentile_frame_build_time_millis": 14.883,
"worst_frame_build_time_millis": 18.782,
"missed_frame_build_budget_count": 1,
"average_frame_rasterizer_time_millis": 5.023033333333336,
"stddev_frame_rasterizer_time_millis": 1.6445388888888894,
"90th_percentile_frame_rasterizer_time_millis": 7.814,
"99th_percentile_frame_rasterizer_time_millis": 13.497,
"worst_frame_rasterizer_time_millis": 15.008,
"missed_frame_rasterizer_budget_count": 0,
"frame_count": 239,
"frame_rasterizer_count": 240,
"new_gen_gc_count": 8,
"old_gen_gc_count": 0,
```
Fixes https://github.com/flutter/flutter/issues/133437
With this fix, the [benchmarks](https://github.com/flutter/flutter/pull/133434) for pictures with huge bounds become runnable under more conditions. Currently those benchmarks only run on platforms where we don't run into this issue. The performance on those platforms/conditions will not change much, but with this fix we can run them on Impeller and add a variant that has a platform view on the display.
Relands https://github.com/flutter/engine/pull/45131
Fixes https://github.com/flutter/flutter/issues/132416
Differences from last time:
- Some minor merge conflict fixes
- Use the RTree to get the bounds instead of recalculating the bounds
- Make the iOS platform view controller implementation use the impeller-aware slices instead of the display list ones. This has been fixed for Android and the desktop embedding, but I missed iOS. The unit tests weren't actually running before I branched for my PR, @zanderso fixed them up separately and this resulted in catching the failures on post submit last time.
Closes https://github.com/flutter/flutter/issues/132994.
I suspect this is imperfect, so please suggest any changes or additions (either using GitHub's tooling or as a review comment).
/cc @dnfield
Reverts flutter/engine#45131
This is failing the Impeller variants of the unobstructed platform views tests:
https://ci.chromium.org/ui/p/flutter/builders/prod/Mac%20Production%20Engine%20Drone/132249/overview
```
Failing tests:
-[UnobstructedPlatformViewTests testPlatformViewsMaxOverlays]
-[UnobstructedPlatformViewTests testOneOverlay]
-[UnobstructedPlatformViewTests testOneOverlayPartialIntersection]
-[UnobstructedPlatformViewTests testTwoIntersectingOverlays]
-[UnobstructedPlatformViewTests testOneOverlayAndTwoIntersectingOverlays]
** TEST FAILED **
```
We've retried it a few times so I suspect this isn't a flake.
This is a reland of https://github.com/flutter/engine/pull/44248
Fixes https://github.com/flutter/flutter/issues/132416
Changes from last time:
- The `drawPoints` benchmark was failing to render anything meaningful because of an error in `AiksLayer` that resulted in an infinitely sized bounding rectangle poisoning the bounds with `NaN` (this happens, for example, with a `drawPaint` call, which that benchmark happens to use). Added a test covering this and filed https://github.com/flutter/flutter/issues/132770 to explore ways to avoid this in the future.
- There was a bug in `DlAiksCanvas::SaveLayer` where a `nullptr` `paint` but non-`nullptr` `backdrop` was failing to actually save the layer. This resulted in incorrect rendering.
- There was a bug in `impeller::Canvas::DrawPicture` that resulted in incorrect stencil depth counting. That was fixed separately by @bdero, but was the cause of incorrect rendering in some Wonderous screens.
- I've added a simple implementation for `AiksLayer::IsReplacing`. It does not currently compare as deeply as the `DisplayListLayer` version potentially does, but it is good enough to avoid the regression noted in https://github.com/flutter/flutter/issues/132071. That regression was on a benchmark that greatly benefits from partial repaint. With the new implementation, it still gains partial repaint where it previously did not. There is more work that can be done here, filed https://github.com/flutter/flutter/issues/133361 to track that work.
I merged but did not fully integrate the `DisplayListBuilder`/`CanvasToReceiver` work that @flar has done. I have a local experiment with that, but would prefer to see this land and run through the device lab so we get some better comparison numbers for which one performs better.
Closes https://github.com/flutter/flutter/issues/132860.
- If `setDither(true)` is called, and an existing `setColorSource` is a gradient, it is ignored.
- If `setColorSource(...)` is called, and it is a gradient, and dithering was previously set, it is cleared.
I'm not sure this is fool proof.
DisplayListBuilder grew over time from a class that implemented a developer-unfriendly stateful API into one that implemented both the developer-friendly DlCanvas API as well as its original stateful API. Over time, the stateful API was buried under a "testing only" facade.
In the meantime, the optimization features that it applies to the DlCanvas calls before it records the operations in a DlOp store are useful without the recording process and so I've been wanting to break those into 2 parts for a while with the goal of removing all stateful APIs from DisplayListBuilder (see https://github.com/flutter/flutter/issues/108303).
This PR takes a major step along that direction by splitting DisplayListBuilder into essentially a convenience class that marries 2 new classes together to achieve its old functionality:
- `DlCanvasToReceiver` - a class that implements DlCanvas, optimizes common situations, and then sends commands to any object that implements `DlOpReceiver`
- `DlOpRecorder` - an implementation of DlOpReceiver that records the operations in a buffer from which to create a `DisplayList` object
- `DisplayListBuilder` now inherits from DlCanvasToReceiver to get the optimizations and provides it with an instance of `DlOpRecorder` as the receiver that it will send its results to
- Similarly, a `DlCanvasToReceiver` instance could be directed to an `impeller:DlDispatcher` to achieve a more straight-through path from the DlCanvas interface to impeller Pictures.
In the Impeller backend, we **only** support dithering of _gradients_. In addition, it will be the default (and only option).
In the [process of enabling dithering by default](https://github.com/flutter/engine/pull/44705), i.e.
```diff
class Paint {
- static bool enableDithering = false;
+ static bool enableDithering = true;
}
```
... we realized with internal Google testing this will now apply dithering on more than just gradients, i.e. images in the Skia backend. Since we won't support dithering of images in the Impeller backend, this PR gives a "hint" on whether the `colorSource` (if one is set) can be dithered by the contrived rules we've created.
Fixes https://github.com/flutter/flutter/issues/130141
The primary goal of this patch is to move dispatching of `dart:ui` `Canvas` commands to the UI thread.
Before this patch, the architecture is something like:
## UI Thread
- `dart:ui` talks to `DisplayListBuilder`, a `DlCanvas` implementation.
- `DisplayListBuilder` does some clip/bounds tracking and creates a `DisplayList` object that is held by `dart:ui`'s `Picture` objects.
- `DisplayList`s are added to `DisplayListLayer`s in `flow`.
## Raster Thread
- `flow` flattens the various operations into a single `DisplayList` via another `DisplayListBuilder`.
- A `DlOpReceiver`implementation converts that `DisplayList` into an `Aiks` `Canvas`/`Picture`.
After this patch, the architecture instead looks like:
## UI Thread
- No change for Skia.
- If Impeller, use a new `DlCanvasImplementation` that talks to `Aiks`'s `Canvas`.
- If Impeller, `dart:ui` Picture's now hold an `Aiks` `Picture`, which get shared into `AiksLayer`s in `flow`.
## Raster thread
- No change for Skia, but some light refactoring for places that assumed a `DisplayListBuilder` where they really just needed a `DlCanvas`.
- The `Aiks` `Picture`s are combined using new API on `DlCanvas` and still backed by `Aiks`.
These changes show significant improvement on raster times on Android and only very small regressions on UI times in local testing, see https://gist.github.com/dnfield/26528090194c9f5abdbac13cdcbf4f79 for old gallery transition perf numbers.
Many of the other changes in this patch are related to the following:
- Making `DlRTree` usable for Impeller.
- It would be nice to have a version of DlRTree that speaks `impeller::Rect`.
- Creating the requisite classes to support `EmbeddedViews` so that Desktop works.
This patch does not remove the `impeller::DlDispatcher`, which now would only be used in tests.
When a saveLayer is rendered with an ImageFilter that modifies the bounds of the rendered pixels, and some of the content of that saveLayer did not intersect the clip, but the filtered output of that content did intersect the clip, we might not accumulate the bounds of those rendering operations into the DisplayList bounds.
This bug was not encountered during practical testing, but showed up on some testing with the new NOP culling code.
For now, this bug only affects the accuracy of the reported bounds of the DisplayList, but that can affect raster caching and potentially the layer culling done in the LayerTree. It might also affect the accuracy of the RTree associated with the DisplayList, which would only show up in rare circumstances when the indicated operation was used in the presence of an embedded view.
