summaryrefslogtreecommitdiffstats
path: root/Source/WebCore/Android.mk
Commit message (Collapse)AuthorAgeFilesLines
* [HTML5] Implementation of WebSockets (WebKit).Oleg Smirnov2013-02-231-0/+12
| | | | | | | | Added binding to android port for Html5 WebSockets. Added external API for enabling\disabling from browser. SSL Web Sockets are not implemented in the current version. Change-Id: Ie37569107c429492ee91b153939ab1ef4e07435e
* external/webkit: use Cpu Upload pathMatthew Hui2013-01-211-0/+5
| | | | | | | Use Cpu Upload mode since the Gpu mode is not working and expected to be slower. Change-Id: Iacb6de4a4e054e4f6ebb07d8a8209228e78e5ee0
* DOM OptimizationsNaiem Shaik2013-01-201-0/+1
| | | | | | | | | | | | DOM traversal optimizations DOM Core optimizations Prefetch optimization for DOM Tree Traversal Conflicts: Source/WebKit/android/jni/WebViewCore.cpp Change-Id: Icbb8a7229ee9cff1a5401b57c8181f18b9a6d6e0
* WebGL implementation for AndroidPierre-Antoine LaFayette2013-01-201-2/+37
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Implementation of WebGL in Android WebKit. Exposes a low level 3D graphics API based on OpenGL ES 2.0 to JavaScript. WebGL is not compiled by default. It can be enabled by setting ENABLE_WEBGL = true in your device BoardConfig.mk or directly in external/webkit/Android.mk. Includes runtime enablement through Browser Settings -> Labs menu. Enable WebGL debug logs and FPS timing with: adb shell setprop debug.webgl 1 Includes Cross-origin resource sharing support. Includes fixes for the following Khronos WebGL 1.0.1. tests: - premultiplyalpha-test.html - struct-nesting-exceeds-maximum.html - index-validation.html - context-attributes-alpha-depth-stencil-antialias.html - program-test.html - object-deletion-behaviour.html Squashed patches: ----------------------------------------------------------- WebGL implementation for Android Implementation of WebGL in Android WebKit. Exposes a low level 3D graphics API based on OpenGL ES 2.0 to JavaScript. WebGL is not compiled by default. It can be enabled by setting ENABLE_WEBGL = true in your device BoardConfig.mk or directly in external/webkit/Android.mk. It is also disabled by default (in WebSettings.cpp) as required by Khronos until it reaches 100% conformance. -- From: Anders Edenbrandt <anders.edenbrandt@sonyericsson.com> Date: Thu, 26 Jan 2012 11:48:41 +0100 WebGL bug fixes and updates Some updates: - stability fixes - image decoder - redesign of how the FBOs are used - other bug fixes -- From: Anders Edenbrandt <anders.edenbrandt@sonyericsson.com> Date: Tue, 31 Jan 2012 17:20:13 +0100 WebGL code cleanup and bug fixes -- From: Anders Edenbrandt <anders.edenbrandt@sonyericsson.com> Date: Thu Mar 15 10:15:33 2012 +0100 More improvements and bug fixes Updates: - cleaned up buffer handling, allowing for arbitrary number of buffers - removed rect from invalidation - removed screen update request from drawGL - releasing buffers when the browser is paused - added missing method 'slice' for typed arrays - fixed bug in bindFramebuffer https://www.codeaurora.org/gitweb/quic/la/?p=platform/external/webkit.git;a=commit;h=c72ff2aa562941d94ceb51ba685d60809ca882a6 Updates to fix some test failures in Khronos Conformance Test Suite: - Increase max identifier length in shaders to 256 - Add length check on uniforms and attributes - Add minimal support for compressed textures (that is, just return the correct error codes) - Add support for Uint8ClampedByteArray - Modify how error checking is done on framebuffer operations - Activate the GL_OES_packed_depth_stencil extension - Activate the GL_OES_texture_float extension https://www.codeaurora.org/gitweb/quic/la/?p=platform/external/webkit.git;a=commit;h=5bfe05848c12a2894697fbb503cfa79981eb96bd Fix WebGL 1.0.1 premultiplyalpha-test conformance test Fixing bug with toDataURL when called against a Canvas in which WebGL content is being rendered and make sure paintRenderingResultsToImageData isn't used for the premultiplied case. ihttps://www.codeaurora.org/gitweb/quic/la/?p=platform/external/webkit.git;a=commit;h=5834a1e00f89d898a7a0039d47916b196e40a2c8 Update ANGLE to r1009 to fix WebGL 1.0.1 conformance test Fixes struct-nesting-exceeds-maximum Khronos WebGL conformance test. https://www.codeaurora.org/gitweb/quic/la/?p=platform/external/webkit.git;a=commit;h=ae2392db6360b41a5717f3770a9e63b1bcea45d6 vertexAttribPointer needs to reject large negative offsets https://bugs.webkit.org/show_bug.cgi?id=85117 Reviewed by Kenneth Russell. Source/WebCore: * html/canvas/WebGLRenderingContext.cpp: Use long long for GLsizeiptr and GLintptr (WebCore): (WebCore::WebGLRenderingContext::bufferData): (WebCore::WebGLRenderingContext::bufferSubData): (WebCore::WebGLRenderingContext::drawElements): (WebCore::WebGLRenderingContext::getVertexAttribOffset): (WebCore::WebGLRenderingContext::vertexAttribPointer): * html/canvas/WebGLRenderingContext.h: Ditto (WebGLRenderingContext): * html/canvas/WebGLRenderingContext.idl: Ditto LayoutTests: * fast/canvas/webgl/index-validation-expected.txt: * fast/canvas/webgl/index-validation.html: Add a test case for large negative offset. git-svn-id: http://svn.webkit.org/repository/webkit/trunk@116374 268f45cc-cd09-0410-ab3c-d52691b4dbfc https://www.codeaurora.org/gitweb/quic/la/?p=platform/external/webkit.git;a=commit;h=b3a02a0942a0e2c44d23961382145fad6016b2ef Fix for context-attributes-alpha-depth-stencil-antialias Support for alpha, depth and stencil to fix WebGL Khronos 1.0.1 conformance test. Report back that antialiasing is not supported on our platform. https://www.codeaurora.org/gitweb/quic/la/?p=platform/external/webkit.git;a=commit;h=22e98195ac73e7e12a59d5b9a42dfc4e2252b475 WebGLRenderingContext should defer caching program info https://bugs.webkit.org/show_bug.cgi?id=83513 Reviewed by Kenneth Russell. * html/canvas/WebGLProgram.cpp: (WebCore::WebGLProgram::WebGLProgram): (WebCore::WebGLProgram::numActiveAttribLocations): call cacheInfoIfNeeded(); (WebCore::WebGLProgram::getActiveAttribLocation): Ditto. (WebCore::WebGLProgram::isUsingVertexAttrib0): Ditto. (WebCore::WebGLProgram::getLinkStatus): Ditto. (WebCore): (WebCore::WebGLProgram::cacheActiveAttribLocations): (WebCore::WebGLProgram::cacheInfoIfNeeded): Cache link status, active attire locations, etc if needed. (WebCore::WebGLProgram::increaseLinkCount): also invalidate cached info. * html/canvas/WebGLProgram.h: (WebGLProgram): * html/canvas/WebGLRenderingContext.cpp: (WebCore): (WebCore::WebGLRenderingContext::linkProgram): Do not cache program info immediately. git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113915 268f45cc-cd09-0410-ab3c-d52691b4dbfc https://www.codeaurora.org/gitweb/quic/la/?p=platform/external/webkit.git;a=commit;h=66bc9c1b9eb5151b1b236199d0eeb17df0557b47 Runtime enablement of WebGL Hooks up with the Java side to enable/disable WebGL through the debug menu. https://www.codeaurora.org/gitweb/quic/la/?p=platform/external/webkit.git;a=commit;h=75aab57859de45a0aaec5a7cad41a12111e9a82e Support the usage of CORS for WebGL and the 2D canvas Factor CORS request preparation out of DocumentThreadableLoader https://bugs.webkit.org/show_bug.cgi?id=61209 2011-05-20 Adam Barth <abarth@webkit.org> Reviewed by Alexey Proskuryakov. Factor CORS request preparation out of DocumentThreadableLoader https://bugs.webkit.org/show_bug.cgi?id=61209 DocumentThreadableLoader has two jobs: 1) Proxy loads between threads. 2) Run the CORS state machine. This patch begins the work of separating those concerns, allowing CORS to be used elsewhere in the loading pipeline. In particular, this patch moves knowledge of how to prepare CORS requests out of DocumentThreadableLoder. * loader/CrossOriginAccessControl.cpp: (WebCore::isOnAccessControlSimpleRequestHeaderWhitelist): (WebCore::updateRequestForAccessControl): (WebCore::createAccessControlPreflightRequest): * loader/CrossOriginAccessControl.h: * loader/DocumentThreadableLoader.cpp: (WebCore::DocumentThreadableLoader::DocumentThreadableLoader): (WebCore::DocumentThreadableLoader::makeSimpleCrossOriginAccessRequest): (WebCore::DocumentThreadableLoader::makeCrossOriginAccessRequestWithPreflight): git-svn-id: http://svn.webkit.org/repository/webkit/trunk@86980 268f45cc-cd09-0410-ab3c-d52691b4dbfc Add rel type prerender to distinguish prerender from prefetch https://bugs.webkit.org/show_bug.cgi?id=61079 2011-05-21 Gavin Peters <gavinp@chromium.org> Reviewed by Adam Barth. Add rel type prerender to distinguish prerender from prefetch https://bugs.webkit.org/show_bug.cgi?id=61079 Chrome right now uses <link rel=prefetch ...> for one of two things, to warm the cache in the same way as firefox, or to launch a speculative rendering of a web page, for faster "loading" when the user navigates to it. This new rel type will let us distinguish the two cases; the rel type prerender isn't used on the web today, but the Google Web Search example prerendering application is ready to experiment with it. * fast/dom/HTMLLinkElement/prerender-expected.txt: Added. * fast/dom/HTMLLinkElement/prerender.html: Added. * platform/gtk/Skipped: * platform/mac/Skipped: * platform/qt/Skipped: * platform/win/Skipped: 2011-05-21 Gavin Peters <gavinp@chromium.org> Reviewed by Adam Barth. Add rel type prerender to distinguish prerender from prefetch https://bugs.webkit.org/show_bug.cgi?id=61079 Chrome right now uses <link rel=prefetch ...> for one of two things, to warm the cache in the same way as firefox, or to launch a speculative rendering of a web page, for faster "loading" when the user navigates to it. This new rel type will let us distinguish the two cases; the rel type prerender isn't used on the web today, but the Google Web Search example prerendering application is ready to experiment with it. Test: fast/dom/HTMLLinkElement/prerender.html * html/HTMLLinkElement.cpp: (WebCore::HTMLLinkElement::tokenizeRelAttribute): (WebCore::HTMLLinkElement::process): * html/HTMLLinkElement.h: (WebCore::HTMLLinkElement::RelAttribute::RelAttribute): * loader/cache/CachedResource.cpp: (WebCore::defaultPriorityForResourceType): * loader/cache/CachedResource.h: (WebCore::CachedResource::isLinkResource): * loader/cache/CachedResourceLoader.cpp: (WebCore::createResource): (WebCore::CachedResourceLoader::requestLinkResource): (WebCore::CachedResourceLoader::canRequest): * loader/cache/CachedResourceLoader.h: * loader/cache/CachedResourceRequest.cpp: (WebCore::cachedResourceTypeToTargetType): (WebCore::CachedResourceRequest::load): * platform/network/ResourceRequestBase.h: git-svn-id: http://svn.webkit.org/repository/webkit/trunk@87020 268f45cc-cd09-0410-ab3c-d52691b4dbfc Make CachedResource take a ResourceRequest instead of just a url string. 2011-05-24 Nate Chapin <japhet@chromium.org> Reviewed by Adam Barth. Change CachedResources to take a ResourceRequest instead of a url in their constructors and provide a very minimal set of cases for avoiding reusing a CachedResource. The ResourceRequest part of this change requires pushing responsibility for calling Document::completeURL() to the caller, instead of CachedResourceLoader, since ResourceRequest ASSERTs if it is constructed with an invalid url. https://bugs.webkit.org/show_bug.cgi?id=61318 Refactor, no new tests. * css/CSSFontSelector.cpp: (WebCore::CSSFontSelector::addFontFaceRule): * css/CSSImageValue.cpp: (WebCore::CSSImageValue::cachedImage): * css/CSSImportRule.cpp: (WebCore::CSSImportRule::insertedIntoParent): * dom/ProcessingInstruction.cpp: (WebCore::ProcessingInstruction::checkStyleSheet): * dom/ScriptElement.cpp: (WebCore::ScriptElement::requestScript): * html/HTMLLinkElement.cpp: (WebCore::HTMLLinkElement::process): * loader/ImageLoader.cpp: (WebCore::ImageLoader::updateFromElement): * loader/cache/CachedCSSStyleSheet.cpp: (WebCore::CachedCSSStyleSheet::CachedCSSStyleSheet): (WebCore::CachedCSSStyleSheet::didAddClient): (WebCore::CachedCSSStyleSheet::checkNotify): * loader/cache/CachedCSSStyleSheet.h: * loader/cache/CachedFont.cpp: (WebCore::CachedFont::CachedFont): * loader/cache/CachedFont.h: * loader/cache/CachedImage.cpp: (WebCore::CachedImage::CachedImage): (WebCore::CachedImage::checkShouldPaintBrokenImage): * loader/cache/CachedImage.h: * loader/cache/CachedResource.cpp: (WebCore::CachedResource::CachedResource): (WebCore::reuseRequest): (WebCore::CachedResource::allowReuseOfRequest): (WebCore::CachedResource::removeClient): (WebCore::CachedResource::canUseCacheValidator): * loader/cache/CachedResource.h: (WebCore::CachedResource::resourceRequest): (WebCore::CachedResource::url): * loader/cache/CachedResourceLoader.cpp: * loader/cache/CachedResourceLoader.h: * loader/cache/CachedResourceRequest.cpp: (WebCore::CachedResourceRequest::load): * loader/cache/CachedScript.cpp: (WebCore::CachedScript::CachedScript): * loader/cache/CachedScript.h: * loader/cache/CachedXSLStyleSheet.cpp: (WebCore::CachedXSLStyleSheet::CachedXSLStyleSheet): (WebCore::CachedXSLStyleSheet::didAddClient): (WebCore::CachedXSLStyleSheet::checkNotify): * loader/cache/CachedXSLStyleSheet.h: * svg/SVGFEImageElement.cpp: (WebCore::SVGFEImageElement::requestImageResource): * svg/SVGFontFaceUriElement.cpp: (WebCore::SVGFontFaceUriElement::loadFont): * xml/XSLImportRule.cpp: (WebCore::XSLImportRule::loadSheet): git-svn-id: http://svn.webkit.org/repository/webkit/trunk@87239 268f45cc-cd09-0410-ab3c-d52691b4dbfc Support cross-origin property for images 2011-05-26 Adam Barth <abarth@webkit.org> Reviewed by Eric Seidel. Support cross-origin property for images https://bugs.webkit.org/show_bug.cgi?id=61015 Test various cases involving CORS requests and canvas tainting. * http/tests/security/canvas-remote-read-remote-image-allowed-expected.txt: Added. * http/tests/security/canvas-remote-read-remote-image-allowed-with-credentials-expected.txt: Added. * http/tests/security/canvas-remote-read-remote-image-allowed-with-credentials.html: Added. * http/tests/security/canvas-remote-read-remote-image-allowed.html: Added. * http/tests/security/canvas-remote-read-remote-image-blocked-no-crossorigin-expected.txt: Added. * http/tests/security/canvas-remote-read-remote-image-blocked-no-crossorigin.html: Added. * http/tests/security/resources/abe-allow-credentials.php: Added. * http/tests/security/resources/abe-allow-star.php: Added. 2011-05-26 Adam Barth <abarth@webkit.org> Reviewed by Eric Seidel. Support cross-origin property for images https://bugs.webkit.org/show_bug.cgi?id=61015 This patch add support for the crossorigin attribute of images and teaches 2D canvas to respect that flag and not taint a canvas if the image drawn on the canvas is allowed by CORS. While I was editing this code, I couldn't resist a couple touch-up changes. Tests: http/tests/security/canvas-remote-read-remote-image-allowed-with-credentials.html http/tests/security/canvas-remote-read-remote-image-allowed.html http/tests/security/canvas-remote-read-remote-image-blocked-no-crossorigin.html * html/HTMLAttributeNames.in: * html/HTMLCanvasElement.cpp: (WebCore::HTMLCanvasElement::securityOrigin): * html/HTMLCanvasElement.h: * html/HTMLImageElement.idl: * html/canvas/CanvasRenderingContext.cpp: (WebCore::CanvasRenderingContext::checkOrigin): * html/canvas/CanvasRenderingContext2D.cpp: (WebCore::CanvasRenderingContext2D::createPattern): * loader/ImageLoader.cpp: (WebCore::ImageLoader::updateFromElement): * loader/cache/CachedResource.cpp: (WebCore::CachedResource::passesAccessControlCheck): * loader/cache/CachedResource.h: git-svn-id: http://svn.webkit.org/repository/webkit/trunk@87473 268f45cc-cd09-0410-ab3c-d52691b4dbfc HTMLVideoElement::currentSrc() should return a KURL 2011-05-27 Adam Barth <abarth@webkit.org> Reviewed by Eric Seidel. HTMLVideoElement::currentSrc() should return a KURL https://bugs.webkit.org/show_bug.cgi?id=61578 I suspect we got into this mess because the author of this code didn't know about the URL attribute in WebKit IDL, which is super useful! Bad news: The line of code in question seems to have another bug, which I've documented in a FIXME. Let the yak shaving continue! * html/HTMLMediaElement.cpp: (WebCore::urlForLogging): (WebCore::HTMLMediaElement::loadResource): (WebCore::HTMLMediaElement::isSafeToLoadURL): (WebCore::HTMLMediaElement::selectNextSourceChild): (WebCore::HTMLMediaElement::getPluginProxyParams): * html/HTMLMediaElement.h: (WebCore::HTMLMediaElement::currentSrc): (WebCore::HTMLMediaElement::currentURL): * html/canvas/CanvasRenderingContext.cpp: (WebCore::CanvasRenderingContext::checkOrigin): * rendering/HitTestResult.cpp: (WebCore::HitTestResult::absoluteMediaURL): - This complete URL call was unnecessary because currentSrc is already absolute. git-svn-id: http://svn.webkit.org/repository/webkit/trunk@87539 268f45cc-cd09-0410-ab3c-d52691b4dbfc Disallow use of cross-domain media (images, video) in WebGL 2011-06-09 Kenneth Russell <kbr@google.com> Reviewed by Adam Barth. Disallow use of cross-domain media (images, video) in WebGL https://bugs.webkit.org/show_bug.cgi?id=62257 Updated WebGL implementation to track recent spec updates in this area. Tests: http/tests/security/webgl-remote-read-remote-image-allowed-with-credentials.html http/tests/security/webgl-remote-read-remote-image-allowed.html http/tests/security/webgl-remote-read-remote-image-blocked-no-crossorigin.html * html/canvas/CanvasRenderingContext.cpp: (WebCore::CanvasRenderingContext::wouldTaintOrigin): (WebCore::CanvasRenderingContext::checkOrigin): * html/canvas/CanvasRenderingContext.h: (WebCore::CanvasRenderingContext::checkOrigin): * html/canvas/WebGLRenderingContext.cpp: (WebCore::WebGLRenderingContext::readPixels): (WebCore::WebGLRenderingContext::texImage2D): (WebCore::WebGLRenderingContext::videoFrameToImage): (WebCore::WebGLRenderingContext::texSubImage2D): * html/canvas/WebGLRenderingContext.h: 2011-06-09 Kenneth Russell <kbr@google.com> Reviewed by Adam Barth. Disallow use of cross-domain media (images, video) in WebGL https://bugs.webkit.org/show_bug.cgi?id=62257 Updated origin-clean-conformance.html to track upstream version in Khronos repository. Added new layout tests mirroring those added in bug 61015 which verify that new CORS support for images is working in the context of WebGL. Verified new tests in WebKit and Chromium. Skipped tests on platforms where WebGL is disabled. * http/tests/canvas/webgl/origin-clean-conformance-expected.txt: * http/tests/canvas/webgl/origin-clean-conformance.html: * http/tests/security/webgl-remote-read-remote-image-allowed-expected.txt: Added. * http/tests/security/webgl-remote-read-remote-image-allowed-with-credentials-expected.txt: Added. * http/tests/security/webgl-remote-read-remote-image-allowed-with-credentials.html: Added. * http/tests/security/webgl-remote-read-remote-image-allowed.html: Added. * http/tests/security/webgl-remote-read-remote-image-blocked-no-crossorigin-expected.txt: Added. * http/tests/security/webgl-remote-read-remote-image-blocked-no-crossorigin.html: Added. * platform/gtk/Skipped: * platform/mac-leopard/Skipped: * platform/mac-wk2/Skipped: * platform/qt/Skipped: git-svn-id: http://svn.webkit.org/repository/webkit/trunk@88489 268f45cc-cd09-0410-ab3c-d52691b4dbfc https://www.codeaurora.org/gitweb/quic/la/?p=platform/external/webkit.git;a=commit;h=0ddd7c1d91c08fbee6c99b79fa9971a7ac914384 Runtime enablement of WebGL logs Allows enabling WebGL method level logging using: adb shell setprop debug.webgl 1 https://www.codeaurora.org/gitweb/quic/la/?p=platform/external/webkit.git;a=commit;h=307d4a2b5b20f2609497ceaea1eca2e29a8a781f Adding WebGL FPS timing to logs Samples WebGL average FPS every 15 frames and outputs to WebGL debug log. TODO: Add setting to browser debug settings that overlays FPS on web pages. Postpone deleteRenderbuffer/deleteTexture until all framebuffer attachment points are removed. https://bugs.webkit.org/show_bug.cgi?id=74741 Reviewed by Kenneth Russell. Source/WebCore: Use WebGLObject's attachment count mechanism to track if a renderbuffer/texture is still attached to framebuffers, and if its deletion should be delated or not. * html/canvas/WebGLFramebuffer.cpp: (WebCore::WebGLFramebuffer::setAttachmentForBoundFramebuffer): (WebCore::WebGLFramebuffer::getAttachment): (WebCore::WebGLFramebuffer::removeAttachmentFromBoundFramebuffer): (WebCore::WebGLFramebuffer::deleteObjectImpl): (WebCore::WebGLFramebuffer::isBound): * html/canvas/WebGLFramebuffer.h: LayoutTests: * fast/canvas/webgl/object-deletion-behaviour-expected.txt: * fast/canvas/webgl/object-deletion-behaviour.html: synced with khronos side. git-svn-id: http://svn.webkit.org/repository/webkit/trunk@103272 268f45cc-cd09-0410-ab3c-d52691b4dbfc https://www.codeaurora.org/gitweb/quic/la/?p=platform/external/webkit.git;a=commit;h=9def386340c74f2a745fb041b1cb11daa30d1a82 ----------------------------------------------------------- (cherry-picked from commit 6d9d732ff06a6b265d02b18d7034068a68ef0fde) Conflicts: Android.mk Source/WebCore/ChangeLog Change-Id: I3bbf993fe5a3d6cea53e019c8fa3912ecd2bd429
* Switch GraphicsOperation to LinearAllocatorJohn Reck2012-07-311-0/+1
| | | | Change-Id: I9afb489c55a7acfdc35cfe8655d46ce47923b3df
* Don't use SkPicture for DrawText recordingJohn Reck2012-07-271-0/+1
| | | | Change-Id: I5b6872bc37eefded7decae767fc39f1cef858dac
* DO NOT MERGE Push everything into an R-TreeJohn Reck2012-07-271-1/+0
| | | | | | Cherry pick from master Change-Id: I0ef32777f33104d440459d9d4d3c0d61f75a11a2
* Use a linear allocator for the RTreeJohn Reck2012-07-171-1/+2
| | | | Change-Id: I5e04c28a0c65378a26b1b99bcfeed4331591e265
* New R-Tree implementationNicolas Roard2012-07-131-0/+1
| | | | | Change-Id: If46f35856279ab2a356cb7f99671fcd15f4ddf37 (cherry picked from commit 19d56fe70a9ea33c3ce2ad4f0b547e35453aeafd)
* Hook up client-based Geolocation mockSteve Block2012-05-221-2/+0
| | | | | | | | | | | | | | | | | | | This simply involves updating GeolocationMock, which is called by DRT from Java, to call the GeolocationClientMock via GeolocationManager, rather than the old GeolocationServiceMock. Note that GeolocationClientMock is tied to a specific page, whereas GeolocationServiceMock had static methods, so we need to pass the WebViewCore from Java. We can now safely exclude the non-client-based files from the build. We pass all Geolocation LayoutTests, except for permission-denied-already-[error|success].html, which are flaky due to https://bugs.webkit.org/show_bug.cgi?id=87033. See corresponding framework change I6d88d5dce5c2148812b191a5b452718bf0854aeb. Bug: 6511338 Change-Id: Ib74a3c05991593e75c3138415d4ac0bf0c9aefa9
* Switch to client-based GeolocationSteve Block2012-05-221-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The motivation for this switch is that it is required to enable PREEMPT_GEOLOCATION_PERMISSION. See b/4500947. The switch involves changing lots of Android's WebKit plumbing to use a new GeolocationClientAndroid in WebKit (called from WebCore::GeolocationController), rather than the existing GeolocationServiceAndroid in WebCore/platform (called from WebCore::Geolocation). This is complicated by the need to support mocking for LayoutTests. The pattern used is very similar to that used for DeviceOrientation and DeviceMotion. This patch ... - Enables CLIENT_BASED_GEOLOCATION - Leaves PREEMPT_GEOLOCATION_PERMISSION disabled for now - Removes GeolocationServiceAndroid - Adds a new GeolocationClientAndroid - Just a proxy to real client - Adds a new GeolocationManager - Manages switching between the real and mock clients - Provides the appropriate client to GeolocationClientAndroid - Plumbs to the real client calls for permissions and for suspending/resuming - Adds a new GeolocationClientImpl - The real client - Based heavily on GeolocationServiceAndroid - Delegates to GeolocationServiceBridge, as GeolocationServiceAndroid did - Calls back directly to GeolocationController - Owns the existing GeolocationPermissions - Updates WebViewCore to call the GeolocationManager for permissions and for suspending/resuming - Removes superfluous ChromeClient methods for permissions Note that support for mocking will be added in a later change. See bug for details. Bug: 6511338 Change-Id: I4cdbf55846e76b37c161834b83135b4dc48dbcdc
* Move GeolocationServiceBridge to WebKit/android/jni and extract Listener ↵Steve Block2012-05-181-1/+0
| | | | | | | | | | | | | | interface Also modify GeolocationServiceBridge to take a WebViewCore rather than a Frame. This will simplify the transition to client-based implementation, which will re-use GeolocationServiceBridge. No functional change. Bug: 6511338 Change-Id: I2de9dfed51d785c85b40f362f20189f208adf6ec
* Rewrite PictureSet with TURBO!John Reck2012-05-021-1/+1
| | | | | | | | | | | | | | | | | | | | This changes how partial invals are done by adding a hybrid mode. What we used to do is generate a SkPicture for the new area. This SkPicture would possibly be larger than the actual inval, depending on various merge rules (more SkPictures == slower to draw a tile) The new code rewrites PictureSet entirely, preserving many of the old rules but cleans up the code and adds the concept of a "PrerenderedInval". This is a partial inval that WebKit has rasterized. By having WebKit produce both a SkPicture and a SkBitmap, we avoid needing to play back the picture and avoid overdrawing. We take this SkBitmap, and simply update the front textures with it. This gives us full partial invals through the entire system without hitting any driver bugs, and with minimal copies. And while the SkPicture may be larger than the inval, the SkBitmap that is rasterized is not - it matches the area webkit has said is dirty. Change-Id: Ieb7ecc9db0d4f679102fda004a43399f9b319ebc
* Fix Gradient & Pattern handlingJohn Reck2012-04-131-1/+1
| | | | | | | | | | | Bug: 6300357 Switch to GradientSkia (fixes the crash) Fix GraphicsContextAndroid to delay setting the fill & stroke shaders until immediately before they are needed, as they can be updated between when they are set and when they are used, which SkShader doesn't support. Change-Id: I85c81ad5764495f45547d2cef08bcf7ed7a885c1
* Reorganize platform/graphics/androidNicolas Roard2012-04-061-51/+55
| | | | Change-Id: Idc67155cfa99784dcd931e705336bfa063ecae46
* Implements a recording GraphicsContextNicolas Roard2012-04-061-2/+6
| | | | Change-Id: I41feadb23dce25af321331c459eb159c6141831b
* Rename several classesChris Craik2012-04-041-4/+5
| | | | | | | | | | | | | | | | | | | | BaseTile,BaseTileTexture -> Tile,TileTexture Used in layers other than the base LayerGroup -> Surface Renamed to convey that it is a member of the SurfaceCollection, and that the layers grouped inside are painted together (if at all) DualTiledTexture -> SurfaceBacking Better conveys that this is the raster backing for a surface that can paint. It may be implemented with two tiled textures for now, but that isn't as important as its relationship to the surface. TiledTexture -> TileGrid Renamed to make it more clear that it was a container of tiles, and to be less similar to TileTexture Change-Id: I843f8603a2080cfe5a7313ba1c2eff10620f8aa2
* reimplement BaseLayerAndroid as a LayerAndroid subclassChris Craik2012-03-221-2/+0
| | | | | | Breaks low res tiling, and expanded rendering bounds. Change-Id: Iec3ba2adceb6bd95399594d29fd8c0b18c1122ce
* Visual Indicator tweaksJohn Reck2012-03-211-1/+0
| | | | Change-Id: I2652ca8d6d3008bb97c2d9ad795240880fd83006
* Canvas on a textureJohn Reck2012-03-201-0/+2
| | | | Change-Id: I841b3e021298738c91701068992798a55290a520
* Introduce a LayerContent interfaceNicolas Roard2012-03-091-0/+2
| | | | | | | | | Layers can now use a LayerContent object to draw their content. We currently have two subclasses, one using an SkPicture (currently used for composited layers), the other using a PictureSet (that we use for the base layer). First step toward unification... Change-Id: I5e7fd06a653f02f8721613fd3a39d36fb64a8614
* Change the way we apply fixed position to layersNicolas Roard2012-03-081-1/+1
| | | | | | | | | With the latest refactoring we lost the possibility of having a layer other than LayerAndroid be fixed positioned. This CL remove the FixedLayerAndroid class (rename it into FixedPositioning) of the hierarchy, and use delegation to apply fixed position. Change-Id: Ib291fcaefe6a4431849ccfe2cf458fac6cac58aa
* Refactor LayerAndroid to add iframe subclassesNicolas Roard2012-03-071-0/+2
| | | | Change-Id: Ia3810e83a2d919ff5fa37b8fcd92e6cbbb63d569
* Refactor LayerAndroid and add a FixedLayerAndroid classNicolas Roard2012-03-061-0/+2
| | | | Change-Id: Ic94e67678e7f5783bf6690308d0a13ce2202d1f2
* layer rendering cleanup/rewriteChris Craik2012-02-281-2/+3
| | | | | | | | | | | | | | | | | | | | | * added 'LayerGroup' class separate the painting/tiled features of a layer (eventually multiple layers) tiled drawing/preparing/counting details are abstracted out of LayerAndroid * added 'SurfaceCollection' class incorporates both the BaseLayerAndroid/LayerAndroid tree, as well as the vector of tiled LayerGroup * renamed 'TreeManager' to 'SurfaceManager' * Removes PaintedSurface was mostly a thin wrapper around DualTiledTexture * Combines TilePainter/SurfacePainter * Simplified ref counting Change-Id: I92c5c75f48d92e0e28812c56de00102399fc02ee
* Delete CursorRingJohn Reck2012-02-271-1/+0
| | | | | | Bug: 6074032 Change-Id: I74f80da22bebbfac556728bba25bcddc2333629a
* Support address detectionJohn Reck2012-02-231-0/+1
| | | | Change-Id: I88b6ebfb4c967e954a2420592a58af3ed188226d
* Merge "Clean up all the old code paths before TransferQueue"Teng-Hui Zhu2011-12-151-2/+0
|\
| * Clean up all the old code paths before TransferQueueTeng-Hui Zhu2011-12-151-2/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | TransferQueue is using one GL texture per tile, which is the best in terms of memory usage. The old code paths include EGLImage mode and SurfaceTexture per tile mode. At the same time, all the locks related to EGLImage are deleted too. There should be no funcationality change here. By deleting the DoubleBufferTexture, the only thing need to be added into BaseTileTexture is the TextureInfo, which can be optimized further in the future. Change-Id: If27fbca7cb55c73d94fbde3ba02d6208707ad8f0
* | Don't pass pictures with no content to UI threadChris Craik2011-12-151-0/+1
|/ | | | | | | | | | | | bug:5748182 Adjusted text detection logic to also simply detect content, and moved to a separate class. No content means a picture is not attached to the layer, and a PaintedSurface isn't allocated to the layer on the UI side. Most commonly, this avoids allocating textures for the clipping element for an overscroll div, but only if it has nothing to paint. Change-Id: I5e57d497780c0532e0f28ac94b7ced3de9718e92
* Add support for Float64Array.Ben Murdoch2011-12-021-0/+1
| | | | | | | | | | | | | | Cherry pick from WebKit r87197 to add support for Float64Array. Note that this requires at least V8 3.3. In particular, this CL means that we will consume less memory executing Emscripten code. See http://trac.webkit.org/changeset/87197/ Bug: 5558474 Change-Id: I686d59acfdb56d03496f407d114826f35b4ff2c8
* resolved conflicts for merge of 4eaaa2a0Chris Craik2011-11-171-1/+1
|\ | | | | | | Change-Id: I5da59f75c0ea1cdb7818dd3f8dc557fbf1d29885
| * synchronous layer updates, and animation deferral during paintChris Craik2011-11-161-1/+1
| | | | | | | | | | | | | | | | bug:5522081 bug:5239801 bug:5297563 Change-Id: I600f66999e093f720a8ea97ef3e15d3d1d297a8f
* | Fix debug buildSteve Block2011-10-281-0/+1
|/ | | | | | | | | | | | | | | | | | | | | | | - FrameLoaderClientAndroid::dispatchDidClearWindowObjectInWorld() FrameLoader::url() was removed in http://trac.webkit.org/changeset/76872. We should use Document::url() instead. - ImageBufferAndroid See http://trac.webkit.org/changeset/77286 and http://trac.webkit.org/changeset/77398 - Add PluginDebug.cpp to build - ChromiumIncludes.h Don't build Chromium code without NDEBUG set - Frame.cpp We should not include WebKit files from WebCore. WebViewCore.h introduces problems due to the change to ChromiumIncludes.h and is not required here. Bug: 5448972 Change-Id: I82c5c95591bf01fb65179282d03f865df4255c89
* am 9bd21466: Merge "Refactor the image sharing code" into ics-mr0Nicolas Roard2011-10-131-0/+1
|\ | | | | | | | | * commit '9bd214662e776fb18e43d1ad32ce274803ec309a': Refactor the image sharing code
| * Refactor the image sharing codeNicolas Roard2011-10-121-0/+1
| | | | | | | | | | | | | | | | | | We add a separate Images Manager, working both in GL and software rendering. bug:5425148 Change-Id: Ib5bc88ad94d49c7f1eb969a716691ac98dc10b77
* | Fallback handle if vertical glyphs are missing.claireho2011-10-111-0/+1
|/ | | | | | | | | | | | | | Bug 5094208 : browser VerticalWritingMode support. Some punctuation has to change its glyph in vertical writing mode. For example, dash in vertical text is a vertical bar. In current Android system fonts and fallback fonts, none of them have the substitute tables for vertical writing mode. We call a conversion function that converts the punctuation to the Unicode Vertical Form (0xFE10 - 0xFE19) or to the closest glyphs as alternative when vertical substitute tables are absent in fonts. Otherwise, we use Harfbuzz to do the vertical glyph shapping. Change-Id: If1d31bb72aaaba45ed50bbb75d6246cfa30e3428
* Merge "Double buffering for layers"Chris Craik2011-10-101-0/+1
|\
| * Double buffering for layersChris Craik2011-10-101-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | bug:5309934 Defer picture changes and invalidations when: 1) a tiled texture hasn't completed painting 2) the user is scrolling (done in PaintedSurface) Remaining issues: Should block updates on zooming Change-Id: I60ee722e651f96e95a4ac0e819999fc1589037c3
* | Switch find on page to GLJohn Reck2011-10-071-0/+1
|/ | | | | | | Bug: 5354191 Refactor how extras are drawn in GL Change-Id: Ifb61c73abecd2d074903a7fb44fe4b82aacc64ac
* Add shared images for layersNicolas Roard2011-09-261-0/+1
| | | | | | bug:5242595 bug:5218173 Change-Id: I37d395e85441671312aac3e236cc8276019aa990
* Refactoring of BaseLayerAndroid and GLWebViewState:Nicolas Roard2011-08-191-0/+1
| | | | | | | | | - we extract the GL set up GLWebViewState - we extract the zoom management to a new ZoomManager class bug:3392331 Change-Id: If42523a7192bab2f8c9d1f00a50721380340230d
* Shared surface textureTeng-Hui Zhu2011-08-021-0/+1
| | | | | | | | | | | Instead of assigning each tile a Surface Texture, now just a normal GL texture. The content from skia bitmap will first drawn into a queue of Surface Textures, at draw time, blit them into each tile's GL texture. Added extra lock to protect the queue operation and Surface Texture operation. bug:5044597 Change-Id: I25f46228b93a3e99526daf52d8bd0d8d8fcc2879
* Implements tiled layers. We remove all the existing code in LayerAndroidNicolas Roard2011-07-141-2/+2
| | | | | | | | | | | related to the old texture management and introduce two new classes: - PaintedSurface, used as a placeholder ui-side for the regularly changing LayerAndroid - TiledTexture, implementing the tiling of PaintedSurface using a set of BaseTile bug:4580444 bug:3392331 Change-Id: I0daa079d2acce49f3baa8c312ea0b0dd457ccd5c
* Tile profiling backendChris Craik2011-07-131-0/+1
| | | | | | Added tile profiling member (TilesProfiler) to TilesManager to track rendering status of tiles. Change-Id: Ied828c6ad2a15588965e6641edecb162bd76bc62
* Merge "Refactoring: Moved 'SkLayer' to 'Layer' class in webkit"Chris Craik2011-06-301-0/+1
|\
| * Refactoring: Moved 'SkLayer' to 'Layer' class in webkitChris Craik2011-06-301-0/+1
| | | | | | | | Change-Id: Ie1c24e5e402c539e0359810cfdf872178fa083c1
* | Add initial support for Ganesh renderer.Derek Sollenberger2011-06-301-0/+2
|/ | | | | | | The GaneshRenderer can now be swapped in for the RasterRenderer in BaseTile. A following CL will enable this swap at runtime. Change-Id: Id42040fce2895ab00cccb8f11dfd0489df861655
* Add BaseRenderer to handle generic rendering tasks.Derek Sollenberger2011-06-291-0/+1
| | | | Change-Id: I5cf0e2da4d4f48be5daf176b5aac8222b9919ba9
* Cleanup Skia related rendering code for raster rendering.Derek Sollenberger2011-06-281-1/+1
| | | | | | | | | | | | | | In the existing code we created a sharedBitmap for BaseTiles and never used it instead opting for createing bitmaps on the stack to enable partial invalidation. This CL removes the sharedBitmap concept and instead creates all bitmaps on the stack. This means that BackedDoubleBufferedTexture no longer needs to have a bitmap member, but instead simply takes the bitmap from the caller and uploads it to the appropriate texture. To make this upload clean we now pass the bitmap via const references instead of pointers. Change-Id: Ie218c4b4564e5574ca6e404d4857904ab41a3a5c
generated by cgit v1.1 at 2024-05-12 08:56:13 +0000