Bug 116336 – Headers and footers will sometimes disappear and re-appear during page loads

• Component: Layout and Rendering
• Status: RESOLVED FIXED

[Beth Dakin]

We can fix this by making the RenderLayerCompositor add the header/footer layer as early as possible in its creation process if one has been set for the Page/WKView.

<rdar://problem/13886753>

[Beth Dakin]

Created attachment 202129 [details]
Patch

[WebKit Commit Bot]

Attachment 202129 [details] did not pass style-queue:

Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/WebCore/ChangeLog', u'Source/WebCore/WebCore.exp.in', u'Source/WebCore/page/ChromeClient.h', u'Source/WebCore/page/FrameView.cpp', u'Source/WebCore/page/FrameView.h', u'Source/WebCore/page/Page.cpp', u'Source/WebCore/page/Page.h', u'Source/WebCore/rendering/RenderLayerCompositor.cpp', u'Source/WebKit2/ChangeLog', u'Source/WebKit2/WebProcess/WebCoreSupport/WebChromeClient.cpp', u'Source/WebKit2/WebProcess/WebCoreSupport/WebChromeClient.h', u'Source/WebKit2/WebProcess/WebPage/PageBanner.cpp', u'Source/WebKit2/WebProcess/WebPage/PageBanner.h', u'Source/WebKit2/WebProcess/WebPage/mac/PageBannerMac.mm']" exit_code: 1
Source/WebKit2/WebProcess/WebPage/PageBanner.h:39:  Code inside a namespace should not be indented.  [whitespace/indent] [4]
Total errors found: 1 in 12 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Beth Dakin]

Created attachment 202130 [details]
Patch

[Early Warning System Bot]

Comment on attachment 202130 [details]
Patch

Attachment 202130 [details] did not pass qt-ews (qt):
Output: http://webkit-queues.appspot.com/results/494370

[Early Warning System Bot]

Comment on attachment 202130 [details]
Patch

Attachment 202130 [details] did not pass qt-wk2-ews (qt-wk2):
Output: http://webkit-queues.appspot.com/results/497073

[EFL EWS Bot]

Comment on attachment 202130 [details]
Patch

Attachment 202130 [details] did not pass efl-ews (efl):
Output: http://webkit-queues.appspot.com/results/486518

[Build Bot]

Comment on attachment 202130 [details]
Patch

Attachment 202130 [details] did not pass win-ews (win):
Output: http://webkit-queues.appspot.com/results/486519

[EFL EWS Bot]

Comment on attachment 202130 [details]
Patch

Attachment 202130 [details] did not pass efl-wk2-ews (efl-wk2):
Output: http://webkit-queues.appspot.com/results/482888

[kov's GTK+ EWS bot]

Comment on attachment 202130 [details]
Patch

Attachment 202130 [details] did not pass gtk-ews (gtk):
Output: http://webkit-queues.appspot.com/results/495383

[Beth Dakin]

Created attachment 202135 [details]
Patch

[Early Warning System Bot]

Comment on attachment 202135 [details]
Patch

Attachment 202135 [details] did not pass qt-wk2-ews (qt-wk2):
Output: http://webkit-queues.appspot.com/results/486528

[Early Warning System Bot]

Comment on attachment 202135 [details]
Patch

Attachment 202135 [details] did not pass qt-ews (qt):
Output: http://webkit-queues.appspot.com/results/497089

[EFL EWS Bot]

Comment on attachment 202135 [details]
Patch

Attachment 202135 [details] did not pass efl-ews (efl):
Output: http://webkit-queues.appspot.com/results/484787

[EFL EWS Bot]

Comment on attachment 202135 [details]
Patch

Attachment 202135 [details] did not pass efl-wk2-ews (efl-wk2):
Output: http://webkit-queues.appspot.com/results/487522

[Beth Dakin]

Created attachment 202138 [details]
Patch

[Early Warning System Bot]

Comment on attachment 202138 [details]
Patch

Attachment 202138 [details] did not pass qt-ews (qt):
Output: http://webkit-queues.appspot.com/results/498015

[Early Warning System Bot]

Comment on attachment 202138 [details]
Patch

Attachment 202138 [details] did not pass qt-wk2-ews (qt-wk2):
Output: http://webkit-queues.appspot.com/results/495393

[Beth Dakin]

Created attachment 202139 [details]
Patch

[WebKit Commit Bot]

Attachment 202139 [details] did not pass style-queue:

Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/WebCore/ChangeLog', u'Source/WebCore/WebCore.exp.in', u'Source/WebCore/page/ChromeClient.h', u'Source/WebCore/page/FrameView.cpp', u'Source/WebCore/page/FrameView.h', u'Source/WebCore/page/Page.cpp', u'Source/WebCore/page/Page.h', u'Source/WebCore/rendering/RenderLayerCompositor.cpp', u'Source/WebKit2/ChangeLog', u'Source/WebKit2/WebProcess/WebCoreSupport/WebChromeClient.cpp', u'Source/WebKit2/WebProcess/WebCoreSupport/WebChromeClient.h', u'Source/WebKit2/WebProcess/WebPage/PageBanner.cpp', u'Source/WebKit2/WebProcess/WebPage/PageBanner.h', u'Source/WebKit2/WebProcess/WebPage/mac/PageBannerMac.mm']" exit_code: 1
Source/WebKit2/ChangeLog:40:  ChangeLog entry has no bug number  [changelog/bugnumber] [5]
Total errors found: 1 in 12 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Beth Dakin]

Created attachment 202142 [details]
Patch

[Early Warning System Bot]

Comment on attachment 202142 [details]
Patch

Attachment 202142 [details] did not pass qt-ews (qt):
Output: http://webkit-queues.appspot.com/results/483873

[Early Warning System Bot]

Comment on attachment 202142 [details]
Patch

Attachment 202142 [details] did not pass qt-wk2-ews (qt-wk2):
Output: http://webkit-queues.appspot.com/results/493403

[Beth Dakin]

Created attachment 202146 [details]
Patch

[Early Warning System Bot]

Comment on attachment 202146 [details]
Patch

Attachment 202146 [details] did not pass qt-ews (qt):
Output: http://webkit-queues.appspot.com/results/490460

[Beth Dakin]

Created attachment 202153 [details]
Patch

[Simon Fraser (smfr)]

Comment on attachment 202153 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=202153&action=review

> Source/WebKit2/WebProcess/WebPage/mac/PageBannerMac.mm:77
> +    [parentLayer->platformLayer() addSublayer:m_layer.get()];

GraphicsLayers own their sublayer tree, so if changes to the parentLayer get flushed, that will nuke the sublayers. What prevents that from happening to headers and footers?

[Simon Fraser (smfr)]

Comment on attachment 202153 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=202153&action=review

>> Source/WebKit2/WebProcess/WebPage/mac/PageBannerMac.mm:77
>> +    [parentLayer->platformLayer() addSublayer:m_layer.get()];
> 
> GraphicsLayers own their sublayer tree, so if changes to the parentLayer get flushed, that will nuke the sublayers. What prevents that from happening to headers and footers?

We think this is OK because the banner GraphicsLayers are outside the normal flush tree.

[Beth Dakin]

Thank you! http://trac.webkit.org/changeset/150307