Make DisplayLists' use of DlPaint, SkPaint, and impeller::Paint agree on the default isAntiAlias by b-luk · Pull Request #193939 · flutter/flutter

b-luk · 2026-10-06T23:48:44Z

The existing code has a mismatch with the default isAntiAlias value of various initial Paint properties used throughout the code.

DisplayListBuilder uses a default DlPaint, which has a default isAntiAlias=false.
Skia's DlSkPaintDispatchHelper use a default SkPaint, which has a default isAntiAlias=false.
Impeller's DlOpReceivers use a default impeller::Paint, which has a default isAntiAlias=true.

The mismatch of impeller::Paint and DisplayListBuilder's DlPaint causes issues. DisplayListBuilder's setAntiAlias no-op exits if setting the same value as its internal DlPaint state:

flutter/engine/src/flutter/display_list/dl_builder.h

Lines 277 to 281 in d454b1b

    
           void setAntiAlias(bool aa) override { 
        
             if (current_.isAntiAlias() != aa) { 
        
               onSetAntiAlias(aa); 
        
             } 
        
           }

That works correctly if all its downstream DlOpReceivers have matching initial values for their Paint state. But this is not the case for impeller::Paint. If we get a Paint with setAntiAlias(false), DisplayListBuilder won't pass this setting to its downstream DlOpReceivers, instead incorrectly assuming that the downstream receivers already have a default false value.

All of these defaults need to agree. We can either change them all to true or all to false. This changes them all to true, which matches the default isAntiAlias=true used for the dart:ui Paint.

Adds a new kDefaultAntiAlias = true const to DlPaint, which it uses when initializing itself. Its default value is now true instead of false.
Skia's DlSkPaintDispatchHelper explicitly sets its SkPaint using DlPaint::kDefaultAntiAlias. Its default value is now true instead of false.
Impeller's impeller::Paint sets its default with DlPaint::kDefaultAntiAlias. This is the same as its existing true default, but setting it this way is more consistent and explicit about matching DlPaint.
Updates tests which relied on DlPaint being non-anti-aliased. Adds new tests to ensure the default paint in DlSkPaintDispatchHelper and the default impeller::Paint match DlPaint.

Fixes #193388

Pre-launch Checklist

If you need help, consider asking for advice on the #hackers-new channel on Discord.

If this change needs to override an active code freeze, provide a comment explaining why. The code freeze workflow can be overridden by code reviewers. See pinned issues for any active code freezes with guidance.

Note: The Flutter team is currently trialing the use of Gemini Code Assist for GitHub. Comments from the gemini-code-assist bot should not be taken as authoritative feedback from the Flutter team. If you find its comments useful you can update your code accordingly, but if you are unsure or disagree with the feedback, please feel free to wait for a Flutter team member's review for guidance on which automated comments should be addressed.

gemini-code-assist

Code Review

This pull request changes the default anti-aliasing value in DlPaint from false to true to match dart:ui Paint.isAntiAlias, updating related tests, builders, and dispatchers accordingly. Feedback suggests updating the in-class initializer for is_anti_alias_ in dl_paint.h to use kDefaultAntiAlias to maintain consistency with the constructor and avoid duplicating state.

b-luk · 2026-10-07T18:06:27Z

The Google testing shows a bunch of golden screenshot diffs. I examined a handful of them, and all the diffs I saw are barely-visible differences to the edges of some rectangles. This is caused by engine-internal draws (e.g. ShaderMask) on Skia now having AA enabled. I think the new behavior is equally correct, if not a bit more consistent, than the old behavior.

Explanation from Gemini:

Google-internal golden test results
The internal presubmit flagged ~70 golden-screenshot test targets (Linux host
tests, Skia software renderer). I looked at the diffs; all of them are the same
mechanism and none show a change to actual widget content:

Every diff is a 1‑px outline of an axis-aligned rectangle — the bounds of an
internal layer (ShaderMask is the common culprit, e.g. shimmer/skeleton
placeholders and gradient-outlined "loading" pills), not the visible shape inside it.
Pixel counts per screenshot are small (~100–1200 px out of ~250k–700k).
Cause: engine-internal draws that use a default-constructed DlPaint — e.g.
ShaderMaskLayer::Paint's DrawRect(shader_rect, dl_paint) — go from non‑AA to
AA on the Skia backend. When the layer's bounds fall on a fractional pixel, the
edge row/column now gets partial coverage instead of being either fully covered
or skipped. User-level paints are unaffected because dart:ui always sets
isAntiAlias explicitly.
Impeller already defaulted to AA=true for these draws, so Skia now matches
Impeller. Note this is a (tiny) real rendering change for apps shipping Skia —
e.g. internal apps default to Skia on Android — not just a test artifact; the
effect is the 1‑px edge behavior described above. Web is unaffected (the web
engine doesn't go through this code path).
The new result is also more consistent: in the old goldens, the top and bottom
edges of the same box could render differently depending on which side of the
pixel centre each landed (e.g. one edge left a visible dark line where the mask
skipped the row entirely).

…vior

b-luk · 2026-10-07T20:57:23Z

Discussed with flar, and decided to manually set AA=false for the ShaderMaskLayer rect mask draws to preserve Skia's existing behavior.

b-luk · 2026-10-08T00:27:26Z

Discussed with flar, and decided to manually set AA=false for the ShaderMaskLayer rect mask draws to preserve Skia's existing behavior.

With the change to make AA=false for ShaderMaskLayer, almost all of the google testing goldens are unchanged. But there is one remaining golden change: https://fusion2.corp.google.com/presubmit/995352717/OCL:995352717:BASE:995434909:1791416516403:9ab040c4;groups=PossiblyNewlyFailing,Pending/targets. Gemini tells me this is for the TextureLayer::Paint case. We could add a change to make AA=false for TextureLayer. Or we can go with the alternate all-default-false approach in #193971.

The all-default-false approach would be safer in terms of maintaining existing behavior without having to hope we find all corner cases like ShaderMaskLayer and TextureLayer. And it's a smaller change overall. The downsides are that 1. it's less "copacetic", as you say, with the Flutter default, 2. it misses out on a small optimization of omitting SetAntiAlias ops in the DL, and 3. it involves more complicated test changes to maintain our AA=true impeller golden tests.

b-luk · 2026-10-10T00:07:40Z

As you suggested, also added AA=false to TextureLayer. Let's see if all the tests pass now.

flar · 2026-10-10T20:11:45Z

Looking through the unit test failures these are due to the fact that we record the DL that the layers produce and compare them to a manually constructed DL that needs to have the AA=false added to the code that generates it.

Oddly, one of the test files generates verbose descriptions of the 2 DLs and the other just reports that they are different. I'm not sure why they do that since they are both using the same EQVerbose facility...?

b-luk · 2026-10-12T19:24:42Z

Updated the unit tests to reflect ShaderLayerMask and TextureLayer setting their internal DlPaints to have AA=false.

Also the remaining google testing failure was due to an internal DlPaint in android/surface_texture_external_texture.cc needing AA=false to preserve existing behavior. I updated that.

With this I'm expecting all tests to pass. Let's wait and see.

b-luk · 2026-10-12T19:29:39Z

Oddly, one of the test files generates verbose descriptions of the 2 DLs and the other just reports that they are different. I'm not sure why they do that since they are both using the same EQVerbose facility...?

Which specific test outputs are you looking at? Every one I looked has similar verbose descriptions on failure.

flar · 2026-10-12T21:40:45Z

Oddly, one of the test files generates verbose descriptions of the 2 DLs and the other just reports that they are different. I'm not sure why they do that since they are both using the same EQVerbose facility...?

Which specific test outputs are you looking at? Every one I looked has similar verbose descriptions on failure.

It turns out that the verbose output gets separated from the GTest output in the logs. They're all recorded there somewhere nearby.

flar

LGTM ... if we're done ... oh, looks like we are!

Roll Flutter from 4249ccbc18a3 to 426be968b5eb (44 revisions) flutter/flutter@4249ccb...426be96 2026-10-14 [email protected] Allow color delta tolerance in in-memory fuzzy image comparator (flutter/flutter#193937) 2026-10-14 [email protected] [web] Drop stale composing range after text is replaced mid-composition (flutter/flutter#194096) 2026-10-14 [email protected] Roll Skia from 2a2bbea66844 to acde3755d2e2 (1 revision) (flutter/flutter#194376) 2026-10-14 [email protected] Roll Skia from 128c983b12de to 2a2bbea66844 (1 revision) (flutter/flutter#194371) 2026-10-14 [email protected] Roll Skia from 48a294ed4955 to 128c983b12de (3 revisions) (flutter/flutter#194353) 2026-10-14 [email protected] feat(flutter_tools): extract WebDebugConfig from DebuggingOptions (flutter/flutter#194312) 2026-10-14 [email protected] Roll Skia from ea6781c5b6d1 to 48a294ed4955 (2 revisions) (flutter/flutter#194349) 2026-10-14 [email protected] Include <vector> in Android proc_table.h. (flutter/flutter#194325) 2026-10-14 [email protected] Roll Dart SDK from 5933cb486b4a to 6b50116630d0 (28 revisions) (flutter/flutter#194341) 2026-10-14 [email protected] Move GLES3 instancing API entry points to FOR_EACH_IMPELLER_GLES3_PROC (flutter/flutter#194319) 2026-10-14 [email protected] feat(flutter_tools): extract AndroidGradleConfig and move toGradleConfig to AndroidBuildInfo (flutter/flutter#194016) 2026-10-14 [email protected] Roll Skia from ae297bec6012 to ea6781c5b6d1 (8 revisions) (flutter/flutter#194331) 2026-10-14 [email protected] [flutter_tools] Migrate gradle_errors and deferred_components_prebuild_validator away from globals (flutter/flutter#193649) 2026-10-14 [email protected] [Impeller] Cache GLES sampler configuration in RenderPassStateCache (flutter/flutter#193922) 2026-10-14 [email protected] Roll pub packages (flutter/flutter#194332) 2026-10-13 [email protected] [tool] Migrate RunCommand to modular dependency injection (flutter/flutter#190791) 2026-10-13 [email protected] Include `engine/src/flutter/.ci.yaml` in quarterly `goldctl` checklist (flutter/flutter#194316) 2026-10-13 [email protected] [flutter_tools] Remove the last AMD module system support (flutter/flutter#193607) 2026-10-13 [email protected] Roll Skia from 8b3394a60654 to ae297bec6012 (5 revisions) (flutter/flutter#194310) 2026-10-13 [email protected] [web] Remove the a11y placeholder only after the gesture ends (flutter/flutter#194109) 2026-10-13 [email protected] Upgrade goldctl to the latest version (flutter/flutter#193932) 2026-10-13 [email protected] [flutter_tools] Lazily initialize AndroidStudio validation and optimize candidate selection (flutter/flutter#191968) 2026-10-13 [email protected] [flutter_tools] Clean up outdated widget preview TODOs and refactor static state (flutter/flutter#189437) 2026-10-13 [email protected] Roll Skia from e5b5db6d0d6f to 8b3394a60654 (2 revisions) (flutter/flutter#194306) 2026-10-13 [email protected] Redistribute TESTOWNERS for Android team (flutter/flutter#193861) 2026-10-13 [email protected] feat(flutter_tools): introduce Test and Drive specification objects (flutter/flutter#194017) 2026-10-13 [email protected] [Impeller] Keep the aspect ratio when shrinking images to the max texture size (flutter/flutter#194010) 2026-10-13 [email protected] [flutter tool] preserve inherited `DASH__SUPPRESS_ANALYTICS=true` for sub-tools (flutter/flutter#194261) 2026-10-13 [email protected] Roll Packages from d8d9e97 to a1ed23c (9 revisions) (flutter/flutter#194302) 2026-10-13 [email protected] Use PE/COFF AOT snapshots for Windows x64 (flutter/flutter#187594) 2026-10-13 [email protected] Fix Expansible crash when PageStorage holds a non-bool value (flutter/flutter#192195) 2026-10-13 [email protected] Roll Skia from ffeccb3ca58a to e5b5db6d0d6f (1 revision) (flutter/flutter#194297) 2026-10-13 [email protected] Explain why initState() is too early to depend on an inherited widget (flutter/flutter#192727) 2026-10-13 [email protected] Refactor TestPlatformDispatcher (flutter/flutter#192747) 2026-10-13 [email protected] Roll Skia from 3de0984208ec to ffeccb3ca58a (2 revisions) (flutter/flutter#194291) 2026-10-13 [email protected] Roll Skia from 0580c2b780f6 to 3de0984208ec (2 revisions) (flutter/flutter#194281) 2026-10-13 [email protected] Roll Skia from f641d55e07a8 to 0580c2b780f6 (3 revisions) (flutter/flutter#194275) 2026-10-13 [email protected] Roll Skia from 0e5d8d1be81b to f641d55e07a8 (1 revision) (flutter/flutter#194266) 2026-10-13 [email protected] Scaffold view metrics overrides (flutter/flutter#192764) 2026-10-13 [email protected] [web] Add SemanticsRole.listBox and SemanticsRole.option (flutter/flutter#192971) 2026-10-13 [email protected] Pin mac_unopt to macOS 15 (flutter/flutter#194254) 2026-10-13 [email protected] Roll pub packages (flutter/flutter#194257) 2026-10-12 [email protected] Roll Skia from 17f6986c226c to 0e5d8d1be81b (4 revisions) (flutter/flutter#194249) 2026-10-12 [email protected] Make DisplayLists' use of DlPaint, SkPaint, and impeller::Paint agree on the default isAntiAlias (flutter/flutter#193939) If this roll has caused a breakage, revert this CL and set the roller ...

… on the default isAntiAlias (flutter#193939) The existing code has a mismatch with the default `isAntiAlias` value of various initial `Paint` properties used throughout the code. - `DisplayListBuilder` uses a default `DlPaint`, which has a default `isAntiAlias=false`. - Skia's `DlSkPaintDispatchHelper` use a default `SkPaint`, which has a default `isAntiAlias=false`. - Impeller's `DlOpReceiver`s use a default `impeller::Paint`, which has a default `isAntiAlias=true`. The mismatch of `impeller::Paint` and `DisplayListBuilder`'s `DlPaint` causes issues. `DisplayListBuilder`'s `setAntiAlias` no-op exits if setting the same value as its internal `DlPaint` state: https://github.com/flutter/flutter/blob/d454b1b841d223f28bb725bb2928e2ac337b5611/engine/src/flutter/display_list/dl_builder.h#L277-L281 That works correctly if all its downstream `DlOpReceiver`s have matching initial values for their `Paint` state. But this is not the case for `impeller::Paint`. If we get a `Paint` with `setAntiAlias(false)`, `DisplayListBuilder` won't pass this setting to its downstream `DlOpReceiver`s, instead incorrectly assuming that the downstream receivers already have a default `false` value. All of these defaults need to agree. We can either change them all to `true` or all to `false`. This changes them all to `true`, which matches the default `isAntiAlias=true` used for the `dart:ui` `Paint`. - Adds a new `kDefaultAntiAlias = true` const to `DlPaint`, which it uses when initializing itself. Its default value is now `true` instead of `false`. - Skia's `DlSkPaintDispatchHelper` explicitly sets its `SkPaint` using `DlPaint::kDefaultAntiAlias`. Its default value is now `true` instead of `false`. - Impeller's `impeller::Paint` sets its default with `DlPaint::kDefaultAntiAlias`. This is the same as its existing `true` default, but setting it this way is more consistent and explicit about matching `DlPaint`. - Updates tests which relied on `DlPaint` being non-anti-aliased. Adds new tests to ensure the default paint in `DlSkPaintDispatchHelper` and the default `impeller::Paint` match `DlPaint`. Fixes flutter#193388 ## Pre-launch Checklist - [x] I read the [Contributor Guide] and followed the process outlined there for submitting PRs. - [x] I read the [AI contribution guidelines] and understand my responsibilities, or I am not using AI tools. - [x] I read the [Tree Hygiene] wiki page, which explains my responsibilities. - [x] I read and followed the [Flutter Style Guide], including [Features we expect every widget to implement]. - [x] I signed the [CLA]. - [x] I listed at least one issue that this PR fixes in the description above. - [x] I updated/added relevant in-code documentation (doc comments with `///`). - [x] If this PR introduces a new feature or capability, I created and linked a website documentation issue or PR in [flutter/website] (or verified none is needed). - [x] I added new tests to check the change I am making, or this PR is [test-exempt]. - [x] I followed the [breaking change policy] and added [Data Driven Fixes] where supported. - [ ] All existing and new tests are passing. If you need help, consider asking for advice on the #hackers-new channel on [Discord]. If this change needs to override an active code freeze, provide a comment explaining why. The code freeze workflow can be overridden by code reviewers. See pinned issues for any active code freezes with guidance. **Note**: The Flutter team is currently trialing the use of [Gemini Code Assist for GitHub](https://developers.google.com/gemini-code-assist/docs/review-github-code). Comments from the `gemini-code-assist` bot should not be taken as authoritative feedback from the Flutter team. If you find its comments useful you can update your code accordingly, but if you are unsure or disagree with the feedback, please feel free to wait for a Flutter team member's review for guidance on which automated comments should be addressed.  [Contributor Guide]: https://github.com/flutter/flutter/blob/main/docs/contributing/Tree-hygiene.md#overview [AI contribution guidelines]: https://github.com/flutter/flutter/blob/main/docs/contributing/Tree-hygiene.md#ai-contribution-guidelines [Tree Hygiene]: https://github.com/flutter/flutter/blob/main/docs/contributing/Tree-hygiene.md [test-exempt]: https://github.com/flutter/flutter/blob/main/docs/contributing/Tree-hygiene.md#tests [Flutter Style Guide]: https://github.com/flutter/flutter/blob/main/docs/contributing/Style-guide-for-Flutter-repo.md [Features we expect every widget to implement]: https://github.com/flutter/flutter/blob/main/docs/contributing/Style-guide-for-Flutter-repo.md#features-we-expect-every-widget-to-implement [CLA]: https://cla.developers.google.com/ [flutter/tests]: https://github.com/flutter/tests [flutter/website]: https://github.com/flutter/website [breaking change policy]: https://github.com/flutter/flutter/blob/main/docs/contributing/Tree-hygiene.md#handling-breaking-changes [Discord]: https://github.com/flutter/flutter/blob/main/docs/contributing/Chat.md [Data Driven Fixes]: https://github.com/flutter/flutter/blob/main/docs/contributing/Data-driven-Fixes.md

github-actions Bot added engine flutter/engine related. See also e: labels. e: impeller Impeller rendering backend issues and features requests labels Oct 6, 2026

b-luk marked this pull request as ready for review October 6, 2026 23:49

b-luk added the CICD Run CI/CD label Oct 6, 2026

gemini-code-assist Bot reviewed Oct 6, 2026

View reviewed changes

Comment thread engine/src/flutter/display_list/dl_paint.cc Resolved

commit

8c18b93

b-luk force-pushed the paint-antialias-default branch from b142552 to 8c18b93 Compare October 6, 2026 23:54

Merge branch 'master' into paint-antialias-default

3ecd50c

b-luk requested a review from flar October 7, 2026 18:06

Explicitly set AA=false for ShaderMaskLayer to preserve existing beha…

65578cf

…vior

b-luk mentioned this pull request Oct 7, 2026

Make impeller::Paint's default isAntiAlias false, matching DlPaint, SkPaint, and DlBuilder #193971

Closed

10 of 11 tasks

b-luk added 2 commits October 9, 2026 17:06

Explicitly set AA=false for TextureLayer as well

6d12f81

Merge branch 'master' into paint-antialias-default

0b7f4b5

b-luk added 2 commits October 12, 2026 12:11

Set DlPaint AA=false in ShaderMaskLayer and TextureLayer unittests

64c36ae

Explicitly set AA=false in Android's SurfaceTextreExternalTexture too

1386433

b-luk requested a review from a team as a code owner October 12, 2026 19:22

b-luk requested review from jesswrd and removed request for a team October 12, 2026 19:22

github-actions Bot added platform-android Android applications specifically team-android Owned by Android platform team labels Oct 12, 2026

flar approved these changes Oct 12, 2026

View reviewed changes

b-luk added the autosubmit Merge PR when tree becomes green via auto submit App label Oct 12, 2026

auto-submit Bot added this pull request to the merge queue Oct 12, 2026

Merged via the queue into flutter:master with commit 3a6417f Oct 12, 2026
23 checks passed

flutter-dashboard Bot removed the autosubmit Merge PR when tree becomes green via auto submit App label Oct 12, 2026

Navigation Menu

Repository navigation

Make DisplayLists' use of DlPaint, SkPaint, and impeller::Paint agree on the default isAntiAlias - #193939

Make DisplayLists' use of DlPaint, SkPaint, and impeller::Paint agree on the default isAntiAlias#193939
auto-submit[bot] merged 7 commits into
flutter:masterfrom
b-luk:paint-antialias-default

b-luk commented Oct 6, 2026

Uh oh!

gemini-code-assist Bot left a comment

Uh oh!

Uh oh!

b-luk commented Oct 7, 2026

Uh oh!

b-luk commented Oct 7, 2026

Uh oh!

b-luk commented Oct 8, 2026 •

edited

Loading

Uh oh!

b-luk commented Oct 10, 2026

Uh oh!

flar commented Oct 10, 2026

Uh oh!

b-luk commented Oct 12, 2026

Uh oh!

b-luk commented Oct 12, 2026

Uh oh!

flar commented Oct 12, 2026

Uh oh!

flar left a comment •

edited

Loading

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

	void setAntiAlias(bool aa) override {
	if (current_.isAntiAlias() != aa) {
	onSetAntiAlias(aa);
	}
	}

Conversation

b-luk commented Oct 6, 2026

Pre-launch Checklist

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

Uh oh!

b-luk commented Oct 7, 2026

Uh oh!

b-luk commented Oct 7, 2026

Uh oh!

b-luk commented Oct 8, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

b-luk commented Oct 10, 2026

Uh oh!

flar commented Oct 10, 2026

Uh oh!

b-luk commented Oct 12, 2026

Uh oh!

b-luk commented Oct 12, 2026

Uh oh!

flar commented Oct 12, 2026

Uh oh!

flar left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

b-luk commented Oct 8, 2026 •

edited

Loading

flar left a comment •

edited

Loading