This repository was archived by the owner on Feb 25, 2025. It is now read-only.
[Impeller] fix clear color optimization for large subpasses. #46887
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
gaaclarke
approved these changes
Oct 13, 2023
Member
gaaclarke
left a comment
gaaclarke
left a comment
There was a problem hiding this comment.
Is there any docstrings that should be updated to make this more clear that this is the correct math?
Contributor
Author
|
I'm not really sure. I sort of eyeballed this after reading through the code for a bit because it didn't seem right that we were using the root pass size to compute the clear color. |
|
Golden file changes have been found for this pull request. Click here to view and triage (e.g. because this is an intentional change). If you are still iterating on this change and are not ready to resolve the images on the Flutter Gold dashboard, consider marking this PR as a draft pull request above. You will still be able to view image results on the dashboard, commenting will be silenced, and the check will not try to resolve itself until marked ready for review. |
Contributor
Author
|
Looks like this is breaking CanRenderClippedLayers |
Contributor
Author
|
So with this change, we're now collapsing both colors into the save layer and using the CPU blends. I think the issue might just be that ColorDodge is incorrect, because If I change the advanced blend it renders as expected. |
Contributor
Author
|
ColorDodge on the CPU is computing a blended color of (1, 1, 1, 1). I think this is incorrect as it doesn't seem to match the GPU blend. |
Contributor
Author
|
I confirmed that there are otherwise no issues with the pass structures, other advanced blends show up correctly - the issue is entirely the result of Color::Blend. |
Contributor
Author
|
I'm tempted to just adjust the blend mode to a different advanced blend and file a bug for fixing color dodge. Thoughts @bdero / @gaaclarke ? |
|
Golden file changes are available for triage from new commit, Click here to view. |
Contributor
Author
|
I adjusted the test and filed a bug |
jonahwilliams
added
the
autosubmit
engine-flutter-autoroll
added a commit
to engine-flutter-autoroll/flutter
that referenced
this pull request
Oct 13, 2023
engine-flutter-autoroll
added a commit
to engine-flutter-autoroll/flutter
that referenced
this pull request
Oct 13, 2023
auto-submit bot
pushed a commit
to flutter/flutter
that referenced
this pull request
Oct 13, 2023
…ions) (#136564) Manual roll requested by [email protected] Cannot build log URL because revision "a4e62581525c" is invalid: Luci builds of "Mac mac_ios_engine" for a4e62581525c1864a3cd23a67539a3615717eb98 was STARTED 2023-10-13 [email protected] Manual roll Dart SDK from c8143a7c026f to 5844b34768ce (1 revision) (flutter/engine#46909) 2023-10-13 [email protected] Roll Skia from c9601553b0f3 to 4bc4b4d22866 (1 revision) (flutter/engine#46908) 2023-10-13 [email protected] Roll Skia from 8dc8e21a4dec to c9601553b0f3 (3 revisions) (flutter/engine#46907) 2023-10-13 [email protected] Roll Fuchsia Mac SDK from SDPa6SfaKvRRSUCij... to 4DwpR2CMJECZJ8EKz... (flutter/engine#46903) 2023-10-13 [email protected] Roll Skia from b50d7f9aa743 to 8dc8e21a4dec (4 revisions) (flutter/engine#46902) 2023-10-13 [email protected] [Impeller] fix clear color optimization for large subpasses. (flutter/engine#46887) 2023-10-13 [email protected] Oops, allow files in opted-in third_party directories. (flutter/engine#46897) 2023-10-13 [email protected] Roll Skia from e10e6765480a to b50d7f9aa743 (1 revision) (flutter/engine#46899) 2023-10-13 [email protected] Roll Fuchsia Linux SDK from pI1BVH08V0eG0M7sw... to wBb32R567S1alWbfn... (flutter/engine#46898) Also rolling transitive DEPS: fuchsia/sdk/core/linux-amd64 from pI1BVH08V0eG to wBb32R567S1a fuchsia/sdk/core/mac-amd64 from SDPa6SfaKvRR to 4DwpR2CMJECZ If this roll has caused a breakage, revert this CL and stop the roller using the controls here: https://autoroll.skia.org/r/flutter-engine-flutter-autoroll Please CC [email protected],[email protected],[email protected],[email protected] on the revert to ensure that a human is aware of the problem. To file a bug in Flutter: https://github.com/flutter/flutter/issues/new/choose To report a problem with the AutoRoller itself, please file a bug: https://issues.skia.org/issues/new?component=1389291&template=1850622 Documentation for the AutoRoller is here: https://skia.googlesource.com/buildbot/+doc/main/autoroll/README.md
harryterkelsen
pushed a commit
that referenced
this pull request
Oct 23, 2023
Fixes flutter/flutter#136507 . The cause of this bug is that we use different sizes to compute the clear color value when constructing the render target and when skipping entities: https://github.com/flutter/engine/blob/main/impeller/entity/entity_pass.cc#L632 https://github.com/flutter/engine/blob/main/impeller/entity/entity_pass.cc#L731 and https://github.com/flutter/engine/blob/main/impeller/entity/entity_pass.cc#L900C3-L900C3 Either the former should use the root pass size or the later should use the subpass size. This usually isn't an issue because if something covers the root pass size it generally always covers the subpass size. But during the page transition, the scaled subpass ends up slightly bigger than the root pass size and so these conditions are mismatched: Subpass size: (1550, 3188) root pass size (1440, 3036) I think subpass size is correct. If the subpass is larger than the parent, then checking the parent size will give incorrect result.
jonahwilliams
pushed a commit
to jonahwilliams/engine
that referenced
this pull request
Nov 1, 2023
…#46887) Fixes flutter/flutter#136507 . The cause of this bug is that we use different sizes to compute the clear color value when constructing the render target and when skipping entities: https://github.com/flutter/engine/blob/main/impeller/entity/entity_pass.cc#L632 https://github.com/flutter/engine/blob/main/impeller/entity/entity_pass.cc#L731 and https://github.com/flutter/engine/blob/main/impeller/entity/entity_pass.cc#L900C3-L900C3 Either the former should use the root pass size or the later should use the subpass size. This usually isn't an issue because if something covers the root pass size it generally always covers the subpass size. But during the page transition, the scaled subpass ends up slightly bigger than the root pass size and so these conditions are mismatched: Subpass size: (1550, 3188) root pass size (1440, 3036) I think subpass size is correct. If the subpass is larger than the parent, then checking the parent size will give incorrect result.
auto-submit bot
pushed a commit
that referenced
this pull request
Nov 2, 2023
) (#47577) Fixes flutter/flutter#136507 . The cause of this bug is that we use different sizes to compute the clear color value when constructing the render target and when skipping entities: https://github.com/flutter/engine/blob/main/impeller/entity/entity_pass.cc#L632 https://github.com/flutter/engine/blob/main/impeller/entity/entity_pass.cc#L731 and https://github.com/flutter/engine/blob/main/impeller/entity/entity_pass.cc#L900C3-L900C3 Either the former should use the root pass size or the later should use the subpass size. This usually isn't an issue because if something covers the root pass size it generally always covers the subpass size. But during the page transition, the scaled subpass ends up slightly bigger than the root pass size and so these conditions are mismatched: Subpass size: (1550, 3188) root pass size (1440, 3036) I think subpass size is correct. If the subpass is larger than the parent, then checking the parent size will give incorrect result.
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fixes flutter/flutter#136507 .
The cause of this bug is that we use different sizes to compute the clear color value when constructing the render target and when skipping entities:
https://github.com/flutter/engine/blob/main/impeller/entity/entity_pass.cc#L632
https://github.com/flutter/engine/blob/main/impeller/entity/entity_pass.cc#L731 and
https://github.com/flutter/engine/blob/main/impeller/entity/entity_pass.cc#L900C3-L900C3
Either the former should use the root pass size or the later should use the subpass size. This usually isn't an issue because if something covers the root pass size it generally always covers the subpass size. But during the page transition, the scaled subpass ends up slightly bigger than the root pass size and so these conditions are mismatched:
Subpass size: (1550, 3188) root pass size (1440, 3036)
I think subpass size is correct. If the subpass is larger than the parent, then checking the parent size will give incorrect result.