Always make a LayerClipRecorder when painting ancestor clipping mask layers. [chromium/src : master]
3 views
Skip to first unread message
Chris Harrelson (Gerrit)
Jul 5, 2017, 4:55:24 PM7/5/17
to blink-revi...@chromium.org, blink-rev...@chromium.org, blink-...@chromium.org, eae+bli...@chromium.org, jchaffraix...@chromium.org, leviw+re...@chromium.org, pdr+renderi...@chromium.org, szager+la...@chromium.org, zol...@webkit.org, chromium...@chromium.org, Dongseong Hwang
Chris Harrelson uploaded patch set #3 to this change.
Always make a LayerClipRecorder when painting ancestor clipping mask layers.
PaintLayerPainter::NeedsToClip applies heuristics to avoid issuing unnecessary
clips (which are quite costly in terms of raster time & memory). However,
the heuristic there for checking HasRadius on the ClipRect is not accurate
for ancestor clipping mask layers, which are painted in a different coordinate
space and mode than other layers.
Bug: 736248
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
Change-Id: I5e91e3d2496c2992eb05304b0a6357d8aa055e3f
---
A third_party/WebKit/LayoutTests/compositing/overflow/composited-layer-under-border-radius-under-composited-layer-expected.png
A third_party/WebKit/LayoutTests/compositing/overflow/composited-layer-under-border-radius-under-composited-layer-expected.txt
A third_party/WebKit/LayoutTests/compositing/overflow/composited-layer-under-border-radius-under-composited-layer.html
M third_party/WebKit/Source/core/layout/compositing/CompositedLayerMapping.cpp
M third_party/WebKit/Source/core/layout/compositing/CompositedLayerMapping.h
M third_party/WebKit/Source/core/paint/PaintLayerPainter.cpp
M third_party/WebKit/Source/core/paint/PaintLayerPainter.h
7 files changed, 56 insertions(+), 16 deletions(-)
To view, visit change 559859. To unsubscribe, visit settings.
Chris Harrelson (Gerrit)
Jul 5, 2017, 4:56:51 PM7/5/17
to blink-revi...@chromium.org, blink-rev...@chromium.org, blink-...@chromium.org, eae+bli...@chromium.org, jchaffraix...@chromium.org, leviw+re...@chromium.org, pdr+renderi...@chromium.org, szager+la...@chromium.org, zol...@webkit.org, Xianda Sun, Stephen Chenney, chromium...@chromium.org, Dongseong Hwang
Chris Harrelson posted comments on this change.
Patch set 3:
Set Ready For Review
Chris Harrelson (Gerrit)
Jul 5, 2017, 6:05:48 PM7/5/17
to blink-revi...@chromium.org, blink-rev...@chromium.org, blink-...@chromium.org, eae+bli...@chromium.org, jchaffraix...@chromium.org, leviw+re...@chromium.org, pdr+renderi...@chromium.org, szager+la...@chromium.org, zol...@webkit.org, Commit Bot, Xianda Sun, Stephen Chenney, chromium...@chromium.org, Dongseong Hwang
Chris Harrelson posted comments on this change.
Patch set 4:Commit-Queue +1
Xianda Sun (Gerrit)
Jul 6, 2017, 10:38:41 AM7/6/17
to Chris Harrelson, blink-revi...@chromium.org, blink-rev...@chromium.org, blink-...@chromium.org, eae+bli...@chromium.org, jchaffraix...@chromium.org, leviw+re...@chromium.org, pdr+renderi...@chromium.org, szager+la...@chromium.org, zol...@webkit.org, Commit Bot, Stephen Chenney, chromium...@chromium.org, Dongseong Hwang
Xianda Sun posted comments on this change.
Patch set 4:Code-Review +1
Chris Harrelson (Gerrit)
Jul 6, 2017, 1:11:30 PM7/6/17
to blink-revi...@chromium.org, blink-rev...@chromium.org, blink-...@chromium.org, eae+bli...@chromium.org, jchaffraix...@chromium.org, leviw+re...@chromium.org, pdr+renderi...@chromium.org, szager+la...@chromium.org, zol...@webkit.org, Xianda Sun, Commit Bot, Stephen Chenney, chromium...@chromium.org, Dongseong Hwang
Chris Harrelson posted comments on this change.
Patch set 4:Commit-Queue +2
Commit Bot (Gerrit)
Jul 6, 2017, 1:16:42 PM7/6/17
to Chris Harrelson, blink-revi...@chromium.org, blink-rev...@chromium.org, blink-...@chromium.org, eae+bli...@chromium.org, jchaffraix...@chromium.org, leviw+re...@chromium.org, pdr+renderi...@chromium.org, szager+la...@chromium.org, zol...@webkit.org, Xianda Sun, Stephen Chenney, chromium...@chromium.org, Dongseong Hwang
Commit Bot merged this change.
Approvals:
Xianda Sun: Looks good to me
Chris Harrelson: Commit
Always make a LayerClipRecorder when painting ancestor clipping mask layers.
PaintLayerPainter::NeedsToClip applies heuristics to avoid issuing unnecessary
clips (which are quite costly in terms of raster time & memory). However,
the heuristic there for checking HasRadius on the ClipRect is not accurate
for ancestor clipping mask layers, which are painted in a different coordinate
space and mode than other layers.
Bug: 736248
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
Change-Id: I5e91e3d2496c2992eb05304b0a6357d8aa055e3f
Reviewed-on: https://chromium-review.googlesource.com/559859
Reviewed-by: Xianda Sun <su...@chromium.org>
Commit-Queue: Chris Harrelson <chri...@chromium.org>
Cr-Commit-Position: refs/heads/master@{#484644}
---
M third_party/WebKit/LayoutTests/FlagExpectations/enable-slimming-paint-v2
A third_party/WebKit/LayoutTests/compositing/overflow/composited-layer-under-border-radius-under-composited-layer-expected.png
A third_party/WebKit/LayoutTests/compositing/overflow/composited-layer-under-border-radius-under-composited-layer-expected.txt
A third_party/WebKit/LayoutTests/compositing/overflow/composited-layer-under-border-radius-under-composited-layer.html
M third_party/WebKit/Source/core/layout/compositing/CompositedLayerMapping.cpp
M third_party/WebKit/Source/core/layout/compositing/CompositedLayerMapping.h
M third_party/WebKit/Source/core/paint/PaintLayerPainter.cpp
M third_party/WebKit/Source/core/paint/PaintLayerPainter.h
8 files changed, 57 insertions(+), 16 deletions(-)
Reply all
Reply to author
Forward
0 new messages