support device scale factor for GraphicsLayerWC #28842

r-otaka · 2024-05-21T08:10:22Z

`b6d35ab`

support device scale factor for GraphicsLayerWC
https://bugs.webkit.org/show_bug.cgi?id=196339

Reviewed by Fujii Hironori.

support device scale factor for GraphicsLayerWC (and WinCairo).
But, NOT using this function now because of following bugs.
so NOW, MiniBrowser sets page zoom to intrinsic device scale factor and
sets "custom device scale factor" to 1.
This simulates the correspondence of device scale factors.

WebInspector always uses intrinsic device scale factor, so if you use
fractional device scale factor, repaint noises may remain.

Remaining bugs
1. Using fractional device scale factor, rendering noises may remain when
   repaint display by JavaScript, scroll with scrollbar dragging or etc.

* Source/WebKit/GPUProcess/graphics/wc/WCScene.cpp:
* Source/WebKit/Shared/NativeWebMouseEvent.h:
* Source/WebKit/Shared/NativeWebWheelEvent.h:
* Source/WebKit/Shared/PlatformPopupMenuData.h:
* Source/WebKit/Shared/PlatformPopupMenuData.serialization.in:
* Source/WebKit/Shared/win/NativeWebMouseEventWin.cpp:
* Source/WebKit/Shared/win/NativeWebWheelEventWin.cpp:
* Source/WebKit/Shared/win/WebEventFactory.cpp:
* Source/WebKit/Shared/win/WebEventFactory.h:
* Source/WebKit/UIProcess/Automation/win/WebAutomationSessionWin.cpp:
* Source/WebKit/UIProcess/Inspector/win/WebInspectorUIProxyWin.cpp:
* Source/WebKit/UIProcess/cairo/BackingStoreCairo.cpp:
* Source/WebKit/UIProcess/wc/DrawingAreaProxyWC.cpp:
* Source/WebKit/UIProcess/wc/DrawingAreaProxyWC.h:
* Source/WebKit/UIProcess/win/PageClientImpl.cpp:
* Source/WebKit/UIProcess/win/WebContextMenuProxyWin.cpp:
* Source/WebKit/UIProcess/win/WebPopupMenuProxyWin.cpp:
* Source/WebKit/UIProcess/win/WebPopupMenuProxyWin.h:
* Source/WebKit/UIProcess/win/WebView.cpp:
* Source/WebKit/UIProcess/win/WebView.h:
* Source/WebKit/WebProcess/WebCoreSupport/win/WebPopupMenuWin.cpp:
* Source/WebKit/WebProcess/WebPage/DrawingArea.h:
* Source/WebKit/WebProcess/WebPage/DrawingArea.messages.in:
* Source/WebKit/WebProcess/WebPage/wc/DrawingAreaWC.cpp:
* Source/WebKit/WebProcess/WebPage/wc/DrawingAreaWC.h:
* Source/WebKit/WebProcess/WebPage/wc/GraphicsLayerWC.cpp:
* Source/WebKit/WebProcess/WebPage/wc/GraphicsLayerWC.h:
* Source/WebKit/WebProcess/WebPage/wc/WCUpdateInfo.h:
* Source/WebKit/WebProcess/WebPage/wc/WCUpdateInfo.serialization.in:
* Tools/MiniBrowser/win/WebKitBrowserWindow.cpp:

Canonical link: https://commits.webkit.org/279794@main

bb094f1

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
✅ 🧪 bindings	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	~~🧪 wpe-wk2~~	~~🧪 wincairo-tests~~
✅ 🧪 webkitperl	~~🧪 ios-wk2~~	✅ 🧪 api-mac	✅ 🧪 api-wpe
	~~🧪 ios-wk2-wpt~~	~~🧪 mac-wk1~~	✅ 🛠 wpe-cairo
	~~🧪 api-ios~~	~~🧪 mac-wk2~~	✅ 🛠 gtk
	✅ 🛠 tv	~~🧪 mac-AS-debug-wk2~~	~~🧪 gtk-wk2~~
	✅ 🛠 tv-sim	❌ 🧪 mac-wk2-stress	~~🧪 api-gtk~~
	✅ 🛠 watch
✅ 🛠 🧪 unsafe-merge	✅ 🛠 watch-sim

webkit-early-warning-system · 2024-05-21T08:10:29Z

EWS run on previous version of this PR (hash 0e9134b)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
✅ 🧪 bindings	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2	✅ 🧪 wincairo-tests
✅ 🧪 webkitperl	✅ 🧪 ios-wk2	✅ 🧪 api-mac	❌ 🧪 api-wpe
	✅ 🧪 ios-wk2-wpt	✅ 🧪 mac-wk1	✅ 🛠 wpe-cairo
	✅ 🧪 api-ios	✅ 🧪 mac-wk2	✅ 🛠 gtk
	✅ 🛠 tv	✅ 🧪 mac-AS-debug-wk2	✅ 🧪 gtk-wk2
	✅ 🛠 tv-sim	✅ 🧪 mac-wk2-stress	✅ 🧪 api-gtk
	✅ 🛠 watch
	✅ 🛠 watch-sim

khei4 · 2024-05-21T12:29:10Z

Source/WebKit/UIProcess/WebPageProxy.h

@@ -1283,7 +1283,7 @@ class WebPageProxy final : public API::ObjectImpl<API::Object::Type::Page>, publ
    double viewScaleFactor() const { return m_viewScaleFactor; }
    void scaleView(double scale);
    void setShouldScaleViewToFitDocument(bool);
-    
+


I guess this new line is added unintentionally, we could narrow this patch's scope into WC and Win if we remove this line. ;)

yeah, I will delete it, thank-you.

khei4 · 2024-05-21T21:08:20Z

Source/WebKit/UIProcess/win/WebView.cpp

+    if (!m_page)
+        m_viewSize = expandedIntSize(FloatSize(LOWORD(lParam), HIWORD(lParam)) / deviceScaleFactorForWindow(hwnd));
+    else
+        m_viewSize = expandedIntSize(FloatSize(LOWORD(lParam), HIWORD(lParam)) / m_page->deviceScaleFactor());


nits: We can reduce redundancy by using ternary op

Suggested change

if (!m_page)

m_viewSize = expandedIntSize(FloatSize(LOWORD(lParam), HIWORD(lParam)) / deviceScaleFactorForWindow(hwnd));

else

m_viewSize = expandedIntSize(FloatSize(LOWORD(lParam), HIWORD(lParam)) / m_page->deviceScaleFactor());

float deviceScaleFactor = m_page ? m_page->deviceScaleFactor() : deviceScaleFactorForWindow(hwnd);

m_viewSize = expandedIntSize(FloatSize(LOWORD(lParam), HIWORD(lParam)) / deviceScaleFactor);

My code has fewer variables than your suggested code.
So, I think that PCs can run my code faster than yours.
Does anyone have any ideas about which is good?

Of Course this is style nits, but it seems like the code emitted by clang/gcc don't have any difference between ternary version and original one, if we use any optimization level over 1. https://godbolt.org/z/c87YTf3na

I fixed this like yours.
Thank-you.

fujii · 2024-05-22T07:47:07Z

Source/WebKit/UIProcess/cairo/BackingStoreCairo.cpp

@@ -101,8 +101,8 @@ void BackingStore::incorporateUpdate(UpdateInfo&& updateInfo)
        return;

 #if ASSERT_ENABLED
-    IntSize updateSize = updateInfo.updateRectBounds.size();
-    updateSize.scale(m_deviceScaleFactor);
+    IntSize updateSize = { static_cast<int>(ceilf(updateInfo.updateRectBounds.size().width() * m_deviceScaleFactor)),


I think this can be IntSize updateSize = expandedIntSize(updateInfo.updateRectBounds.size() * m_deviceScaleFactor);

That seems smart!
I will change the codes.

fujii · 2024-05-22T23:42:11Z

Source/WebKit/WebProcess/WebPage/wc/GraphicsLayerWC.cpp

@@ -584,6 +591,7 @@ void GraphicsLayerWC::flushCompositingStateForThisLayerOnly()
        update.background.color = backgroundColor();
        if (drawsContent() && contentsAreVisible()) {
            update.background.hasBackingStore = true;
+            update.background.backingStoreSize = WebCore::expandedIntSize(scaled(size(), deviceScaleFactor()));


I think size() * deviceScaleFactor() is simpler than scaled(size(), deviceScaleFactor()) .

Okay, I'll change it.

webkit-early-warning-system · 2024-05-23T02:03:10Z

EWS run on previous version of this PR (hash 71654e2)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
✅ 🧪 bindings	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2	✅ 🧪 wincairo-tests
✅ 🧪 webkitperl	✅ 🧪 ios-wk2	✅ 🧪 api-mac	✅ 🧪 api-wpe
	✅ 🧪 ios-wk2-wpt	✅ 🧪 mac-wk1	✅ 🛠 wpe-cairo
	✅ 🧪 api-ios	✅ 🧪 mac-wk2	✅ 🛠 gtk
	✅ 🛠 tv	✅ 🧪 mac-AS-debug-wk2	✅ 🧪 gtk-wk2
	✅ 🛠 tv-sim	✅ 🧪 mac-wk2-stress	✅ 🧪 api-gtk
	✅ 🛠 watch
	✅ 🛠 watch-sim

webkit-early-warning-system · 2024-06-07T05:41:59Z

Starting EWS tests for bb094f1. Live statuses available at the PR page, #28842

https://bugs.webkit.org/show_bug.cgi?id=196339 Reviewed by Fujii Hironori. support device scale factor for GraphicsLayerWC (and WinCairo). But, NOT using this function now because of following bugs. so NOW, MiniBrowser sets page zoom to intrinsic device scale factor and sets "custom device scale factor" to 1. This simulates the correspondence of device scale factors. WebInspector always uses intrinsic device scale factor, so if you use fractional device scale factor, repaint noises may remain. Remaining bugs 1. Using fractional device scale factor, rendering noises may remain when repaint display by JavaScript, scroll with scrollbar dragging or etc. * Source/WebKit/GPUProcess/graphics/wc/WCScene.cpp: * Source/WebKit/Shared/NativeWebMouseEvent.h: * Source/WebKit/Shared/NativeWebWheelEvent.h: * Source/WebKit/Shared/PlatformPopupMenuData.h: * Source/WebKit/Shared/PlatformPopupMenuData.serialization.in: * Source/WebKit/Shared/win/NativeWebMouseEventWin.cpp: * Source/WebKit/Shared/win/NativeWebWheelEventWin.cpp: * Source/WebKit/Shared/win/WebEventFactory.cpp: * Source/WebKit/Shared/win/WebEventFactory.h: * Source/WebKit/UIProcess/Automation/win/WebAutomationSessionWin.cpp: * Source/WebKit/UIProcess/Inspector/win/WebInspectorUIProxyWin.cpp: * Source/WebKit/UIProcess/cairo/BackingStoreCairo.cpp: * Source/WebKit/UIProcess/wc/DrawingAreaProxyWC.cpp: * Source/WebKit/UIProcess/wc/DrawingAreaProxyWC.h: * Source/WebKit/UIProcess/win/PageClientImpl.cpp: * Source/WebKit/UIProcess/win/WebContextMenuProxyWin.cpp: * Source/WebKit/UIProcess/win/WebPopupMenuProxyWin.cpp: * Source/WebKit/UIProcess/win/WebPopupMenuProxyWin.h: * Source/WebKit/UIProcess/win/WebView.cpp: * Source/WebKit/UIProcess/win/WebView.h: * Source/WebKit/WebProcess/WebCoreSupport/win/WebPopupMenuWin.cpp: * Source/WebKit/WebProcess/WebPage/DrawingArea.h: * Source/WebKit/WebProcess/WebPage/DrawingArea.messages.in: * Source/WebKit/WebProcess/WebPage/wc/DrawingAreaWC.cpp: * Source/WebKit/WebProcess/WebPage/wc/DrawingAreaWC.h: * Source/WebKit/WebProcess/WebPage/wc/GraphicsLayerWC.cpp: * Source/WebKit/WebProcess/WebPage/wc/GraphicsLayerWC.h: * Source/WebKit/WebProcess/WebPage/wc/WCUpdateInfo.h: * Source/WebKit/WebProcess/WebPage/wc/WCUpdateInfo.serialization.in: * Tools/MiniBrowser/win/WebKitBrowserWindow.cpp: Canonical link: https://commits.webkit.org/279794@main

webkit-commit-queue · 2024-06-07T05:56:12Z

Committed 279794@main (b6d35ab): https://commits.webkit.org/279794@main

Reviewed commits have been landed. Closing PR #28842 and removing active labels.

webkit-early-warning-system · 2024-06-07T06:01:30Z

Starting EWS tests for bb094f1. Live statuses available at the PR page, #28842

r-otaka requested review from magomez, zdobersek, cdumez and JonWBedard as code owners May 21, 2024 08:10

Ahmad-S792 assigned r-otaka May 21, 2024

khei4 reviewed May 21, 2024

View reviewed changes

fujii reviewed May 22, 2024

View reviewed changes

r-otaka marked this pull request as draft May 22, 2024 08:10

fujii reviewed May 22, 2024

View reviewed changes

r-otaka force-pushed the eng/WebView-should-call-WebPageProxy_setIntrinsicDeviceScaleFactor-for-high-DPI branch from 0e9134b to 71654e2 Compare May 23, 2024 02:03

r-otaka force-pushed the eng/WebView-should-call-WebPageProxy_setIntrinsicDeviceScaleFactor-for-high-DPI branch from 71654e2 to bb094f1 Compare June 7, 2024 05:41

r-otaka marked this pull request as ready for review June 7, 2024 05:44

r-otaka requested review from dcrousso and patrickangle as code owners June 7, 2024 05:44

fujii approved these changes Jun 7, 2024

View reviewed changes

fujii added the unsafe-merge-queue Applied to send a pull request to merge-queue, but skip building and testing label Jun 7, 2024

webkit-commit-queue force-pushed the eng/WebView-should-call-WebPageProxy_setIntrinsicDeviceScaleFactor-for-high-DPI branch from bb094f1 to b6d35ab Compare June 7, 2024 05:56

webkit-commit-queue removed the unsafe-merge-queue Applied to send a pull request to merge-queue, but skip building and testing label Jun 7, 2024

webkit-commit-queue merged commit b6d35ab into WebKit:main Jun 7, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

support device scale factor for GraphicsLayerWC #28842

support device scale factor for GraphicsLayerWC #28842

Uh oh!

r-otaka commented May 21, 2024 •

edited by webkit-early-warning-system

Loading

Uh oh!

webkit-early-warning-system commented May 21, 2024 •

edited

Loading

Uh oh!

khei4 May 21, 2024 •

edited

Loading

Uh oh!

r-otaka May 23, 2024

Uh oh!

khei4 May 21, 2024

Uh oh!

r-otaka May 23, 2024

Uh oh!

khei4 May 23, 2024 •

edited

Loading

Uh oh!

r-otaka Jun 7, 2024

Uh oh!

fujii May 22, 2024

Uh oh!

r-otaka May 23, 2024

Uh oh!

fujii May 22, 2024

Uh oh!

r-otaka May 23, 2024

Uh oh!

webkit-early-warning-system commented May 23, 2024 •

edited

Loading

Uh oh!

webkit-early-warning-system commented Jun 7, 2024

Uh oh!

webkit-commit-queue commented Jun 7, 2024

Uh oh!

webkit-early-warning-system commented Jun 7, 2024

Uh oh!

Uh oh!

support device scale factor for GraphicsLayerWC #28842

support device scale factor for GraphicsLayerWC #28842

Uh oh!

Conversation

r-otaka commented May 21, 2024 • edited by webkit-early-warning-system Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

webkit-early-warning-system commented May 21, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

khei4 May 21, 2024 • 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!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

khei4 May 23, 2024 • 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!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

webkit-early-warning-system commented May 23, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

webkit-early-warning-system commented Jun 7, 2024

Uh oh!

webkit-commit-queue commented Jun 7, 2024

Uh oh!

webkit-early-warning-system commented Jun 7, 2024

Uh oh!

Uh oh!

r-otaka commented May 21, 2024 •

edited by webkit-early-warning-system

Loading

webkit-early-warning-system commented May 21, 2024 •

edited

Loading

khei4 May 21, 2024 •

edited

Loading

khei4 May 23, 2024 •

edited

Loading

webkit-early-warning-system commented May 23, 2024 •

edited

Loading