SVG clip-path is sometimes broken on stevejobsarchive.com #12845
Conversation
This file contains 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
|
EWS run on previous version of this PR (hash aadf5e6) |
shallawa
approved these changes
Apr 18, 2023
| @@ -167,7 +167,7 @@ bool RenderSVGResourceClipper::applyClippingToContext(GraphicsContext& context, | |||
| if (!clipperData.imageBuffer && pathOnlyClipping(context, animatedLocalTransform, objectBoundingBox, effectiveZoom)) | |||
| return true; | |||
|
|
|||
| if (clipperData.invalidate(computeInputs(renderer, objectBoundingBox, clippedContentBounds, effectiveZoom))) { | |||
| if (clipperData.invalidate(computeInputs(renderer, context, objectBoundingBox, clippedContentBounds, effectiveZoom))) { | |||
There was a problem hiding this comment.
Should not computeInputs() and applyClippingToContext() have the arguments: RenderElement and GraphicsContext in the same order: (const RenderElement&, const GraphicsContext& ...) or (const GraphicsContext&, const RenderElement& ...)?
| @@ -45,6 +45,7 @@ struct ClipperData { | |||
| FloatRect clippedContentBounds; | |||
| FloatSize scale; | |||
| float effectiveZoom = 1; | |||
| bool paintingDisabled { false }; | |||
There was a problem hiding this comment.
Alternatively ClipperData::invalidate() can check
context.hasPlatformContext() == imageBuffer->context().hasPlatformContext()
There was a problem hiding this comment.
That won't work; both the NullGraphicsContext and the GPUP-based GraphicsContext have no platform contexts.
shallawa
approved these changes
Apr 18, 2023
Comment on lines
37
to
40
| bool operator==(const Inputs& other) const | ||
| { | ||
| return std::tie(objectBoundingBox, clippedContentBounds, scale, effectiveZoom) == std::tie(other.objectBoundingBox, other.clippedContentBounds, other.scale, other.effectiveZoom); | ||
| return std::tie(objectBoundingBox, clippedContentBounds, scale, effectiveZoom, paintingDisabled) == std::tie(other.objectBoundingBox, other.clippedContentBounds, other.scale, other.effectiveZoom, other.paintingDisabled); | ||
| } |
There was a problem hiding this comment.
Can this just be
bool operator==(const Inputs&) const = default;
webkit-ews-buildbot
added
the
merging-blocked
smfr
force-pushed
the
eng/SVG-clip-path-is-sometimes-broken-on-stevejobsarchive-com
branch
from
aadf5e6
to
0ad7712
Compare
|
EWS run on current version of this PR (hash 0ad7712)
|
https://bugs.webkit.org/show_bug.cgi?id=255577 rdar://107885344 Reviewed by Said Abou-Hallawa. http://book.stevejobsarchive.com/ uses CSS clip-path with a reference to an SVG <clipPath> element which contains text. In this configuration, RenderSVGResourceClipper::applyClippingToContext() falls back to a code path that uses an ImageBuffer as a mask, and it caches the ImageBuffer between calls. This caused a problem when DOM Rendering in the GPU Process was enabled; this code is first hit for a "fake" paint with a NullGraphicsContext which is updating EventRegions, called out of `RenderLayerBacking::updateEventRegion()`. The NullGraphicsContext will make a local ImageBuffer. If we then hit this same code for actual painting with a painting GraphicsContext, we'll use that cached ImageBuffer, rather than creating a new one with appropriate GPU Process backing. Fix this by adding `isPaintingDisabled` to the criteria used to decide if the cached buffer can be re-used. * LayoutTests/svg/masking/masking-with-event-region-expected.html: Added. * LayoutTests/svg/masking/masking-with-event-region.html: Added. * Source/WebCore/rendering/svg/RenderSVGResourceClipper.cpp: (WebCore::RenderSVGResourceClipper::computeInputs): (WebCore::RenderSVGResourceClipper::applyClippingToContext): * Source/WebCore/rendering/svg/RenderSVGResourceClipper.h: (WebCore::ClipperData::Inputs::operator== const): Canonical link: https://commits.webkit.org/263087@main
webkit-commit-queue
force-pushed
the
eng/SVG-clip-path-is-sometimes-broken-on-stevejobsarchive-com
branch
from
0ad7712
to
e4d2a74
Compare
|
Committed 263087@main (e4d2a74): https://commits.webkit.org/263087@main Reviewed commits have been landed. Closing PR #12845 and removing active labels. |
webkit-commit-queue
removed
the
merge-queue
smfr
deleted the
eng/SVG-clip-path-is-sometimes-broken-on-stevejobsarchive-com
branch
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
e4d2a74
0ad7712
🧪 wpe-wk2🧪 ios-wk2🧪 ios-wk2-wpt🧪 gtk-wk2🧪 api-ios🧪 mac-AS-debug-wk2