OSDN Git Service

android-x86/external-webkit.git
14 years agoCherry-pick WebKit change 56112 to add WORKERS guards to V8 bindings
Steve Block [Wed, 17 Mar 2010 15:01:37 +0000 (15:01 +0000)]
Cherry-pick WebKit change 56112 to add WORKERS guards to V8 bindings

The motivation for this change is to allow us to disable workers on Android,
because V8 on Android does not have the required locking.

See http://trac.webkit.org/changeset/56112

Change-Id: Ieee214f4765af3188bbdc51c56f14883f6a6779c

14 years agoCherry-pick WebKit change 55084 to add SHARED_WORKERS guard in V8 globalObjectPrototy...
Steve Block [Wed, 17 Mar 2010 14:36:17 +0000 (14:36 +0000)]
Cherry-pick WebKit change 55084 to add SHARED_WORKERS guard in V8 globalObjectPrototypeIsDOMWindow

The motivation for this change is to allow us to disable workers on Android,
because V8 on Android does not have the required locking.

See http://trac.webkit.org/changeset/55084

Change-Id: If72cb478fb7dad6be44eaa2c1ca19e9790a9dda0

14 years agoCherry-pick WebKit change 54972 to merge DOMWindow and WorkerContext object wrapping...
Steve Block [Wed, 17 Mar 2010 15:25:43 +0000 (15:25 +0000)]
Cherry-pick WebKit change 54972 to merge DOMWindow and WorkerContext object wrapping in V8 bindings

The motivation for this change is to disable workers on Android, because V8 on
Android does not have the required locking.

The current version of WebKit in Android (54731) uses workers features without
ENABLE(WORKERS) guards in the V8 bindings. In particular,
V8DOMWrapper::instantiateV8ObjectInWorkerContext is used from generated code
without guards.

Fixing the code generator would be difficult and would be wasted effort as
V8DOMWrapper::instantiateV8ObjectInWorkerContext no longer exists in tip-of-tree
WebKit. Instead, we cherry-pick the WebKit change which removes this method.

See http://trac.webkit.org/changeset/54972

Change-Id: Ie6365073d7a4d92aa7c1553d87a7e1c1ed514118

14 years agoMerge "Compute the position of the fixed elements to be relative to the screen and...
Nicolas Roard [Wed, 17 Mar 2010 03:34:40 +0000 (20:34 -0700)]
Merge "Compute the position of the fixed elements to be relative to the screen and not the virtual viewport."

14 years agoCompute the position of the fixed elements to be relative to the
Nicolas Roard [Wed, 17 Mar 2010 02:06:20 +0000 (02:06 +0000)]
Compute the position of the fixed elements to be relative to the
screen and not the virtual viewport.

This fixes Bug:2515587 and Bug:2457215

Change-Id: Ib987787cf9bf74b760b857dbbb397f85af334dce

14 years agoTo enable per file log, only need to define PLUGIN_DEBUG_LOCAL
Grace Kloba [Sat, 13 Mar 2010 02:45:47 +0000 (18:45 -0800)]
To enable per file log, only need to define PLUGIN_DEBUG_LOCAL

14 years agoEnsure a plugin is not shown if it is completely obscured by other
Derek Sollenberger [Tue, 16 Mar 2010 20:29:40 +0000 (16:29 -0400)]
Ensure a plugin is not shown if it is completely obscured by other
html content.

Change-Id: Iffcc30d47428708842ac5dddee41cc611d879c1c

14 years agoMerge "Revert "Fix V8 crashes with plugins. The extra call to _NPN_ReleaseObject...
Ben Murdoch [Tue, 16 Mar 2010 18:41:46 +0000 (11:41 -0700)]
Merge "Revert "Fix V8 crashes with plugins. The extra call to _NPN_ReleaseObject inside forgetV8ObjectForNPObject is causing a double deletion.""

14 years agoMerge "Fix the V8 NPObject reference count problem."
Grace Kloba [Tue, 16 Mar 2010 18:40:52 +0000 (11:40 -0700)]
Merge "Fix the V8 NPObject reference count problem."

14 years agoRevert "Fix V8 crashes with plugins. The extra call to _NPN_ReleaseObject inside...
Ben Murdoch [Tue, 16 Mar 2010 18:37:48 +0000 (18:37 +0000)]
Revert "Fix V8 crashes with plugins. The extra call to _NPN_ReleaseObject inside forgetV8ObjectForNPObject is causing a double deletion."

Right problem, wrong fix.

This reverts commit 9e0d526d0170e26b1dc2d0da8347cca28b5f3507.

14 years agoFix the V8 NPObject reference count problem.
Grace Kloba [Tue, 16 Mar 2010 18:19:34 +0000 (11:19 -0700)]
Fix the V8 NPObject reference count problem.

To match JSC version, after getNPObject() is called,
the object should have ref count 2. One will be released
in ~PluginView(). Another one will be released by
the plugin.

This should fix the crash of Flash using V8.

Fix http://b/issue?id=2511963

14 years agoMerge "Fix V8 crashes with plugins. The extra call to _NPN_ReleaseObject inside forge...
Ben Murdoch [Tue, 16 Mar 2010 15:58:16 +0000 (08:58 -0700)]
Merge "Fix V8 crashes with plugins. The extra call to _NPN_ReleaseObject inside forgetV8ObjectForNPObject is causing a double deletion."

14 years agoset up the loader so the plugin will receive the first block of data
Cary Clark [Tue, 16 Mar 2010 12:58:14 +0000 (08:58 -0400)]
set up the loader so the plugin will receive the first block of data

this makes our port follow the pattern of the gtk port

fixes http://b/2484987

Change-Id: I7119a4b84c12f5ed69a9f260148536a3b76016d6

14 years agoFix V8 crashes with plugins. The extra call to _NPN_ReleaseObject inside forgetV8Obje...
Ben Murdoch [Tue, 16 Mar 2010 11:28:32 +0000 (11:28 +0000)]
Fix V8 crashes with plugins. The extra call to _NPN_ReleaseObject inside forgetV8ObjectForNPObject is causing a double deletion.

In ScriptController::clearScriptObjects, we call Unregister (which without this change causes Release to get called in forgetV8ObjectForNpObject).
If the ref count gets to 0, we free the object in Release. Then when Unregister returns, we call Release again in the ScriptController with the
same just-free'd pointer. If we're unlucky, then the Deallocate and free gets called again. We're also trying to access members such as the reference
count and deallocation function through a dead pointer.

Also, in the case where the Flash does not clear it's window object, we call Deallocate directly inside ScriptController::clearScriptObjects. This causes
Unregister to get called (as the object is still alive) which calls Release through forgetV8ObjectForNPObject, which results in a recursive call to Deallocate,
which frees the memory. Then the stack unwinds and we happily try to free again from the first call to Deallocate.

Fix these issues by removing the extra call to NPN_ReleaseObject in forgetV8ObjectForNpObject.

Change-Id: I7f6d21cd4ff38e29bd1a178e4816d023560b6b16

14 years agoPutting Grace's change back in now that we have the new apk from Adobe.
Bart Sears [Mon, 15 Mar 2010 22:50:21 +0000 (15:50 -0700)]
Putting Grace's change back in now that we have the new apk from Adobe.

Revert "Temporary revert the new android_npapi. Waiting for"

This reverts commit 31ba9c5b345211c7c89d08a762c09582a3975738.

14 years agoMerge "read old focus before it's thrown away"
Cary Clark [Mon, 15 Mar 2010 20:50:26 +0000 (13:50 -0700)]
Merge "read old focus before it's thrown away"

14 years agoread old focus before it's thrown away
Cary Clark [Mon, 15 Mar 2010 20:34:21 +0000 (16:34 -0400)]
read old focus before it's thrown away

Change-Id: Iab6a64aec5f146e8b0286d1249d353567b49ea49
http://b/2515977

14 years agoIf the plugin is in an IFrame we need to reposition the plugins surface
Derek Sollenberger [Mon, 15 Mar 2010 18:22:08 +0000 (14:22 -0400)]
If the plugin is in an IFrame we need to reposition the plugins surface
every time the IFrame changes position within the document.

Change-Id: I03090bda439c3a3cac95dce42b3ea37901dc42f8

14 years agoMerge "Clean up full-screen mode plugins now that they handle touch in their java...
Derek Sollenberger [Mon, 15 Mar 2010 17:32:31 +0000 (10:32 -0700)]
Merge "Clean up full-screen mode plugins now that they handle touch in their java view."

14 years agoMerge "Adds some logging to JObjectWrapper"
Steve Block [Mon, 15 Mar 2010 17:31:10 +0000 (10:31 -0700)]
Merge "Adds some logging to JObjectWrapper"

14 years agoAdds some logging to JObjectWrapper
Steve Block [Mon, 15 Mar 2010 16:15:47 +0000 (16:15 +0000)]
Adds some logging to JObjectWrapper

This is an attempt to track down the cause of a crashing bug in the WebCore
bridge, which is only occasionally triggered by the monkey.

Bug: 2435360
Change-Id: I3171eb1978de285fcf258f908aa0a08d95b72968

14 years agoMerge "adjust cache layer position by offsets"
Cary Clark [Mon, 15 Mar 2010 12:40:48 +0000 (05:40 -0700)]
Merge "adjust cache layer position by offsets"

14 years agoClean up full-screen mode plugins now that they handle touch
Derek Sollenberger [Tue, 9 Mar 2010 22:40:44 +0000 (17:40 -0500)]
Clean up full-screen mode plugins now that they handle touch
in their java view.

There is a matching commit in frameworks/base

Change-Id: Ibe76eb22c2b5ccc7fb80e4807edcbbc4d25e4793

14 years agoTemporary revert the new android_npapi. Waiting for
Grace Kloba [Sat, 13 Mar 2010 02:45:47 +0000 (18:45 -0800)]
Temporary revert the new android_npapi. Waiting for
the new apk from Adobe.

14 years agoMerge "Use the bounds we layout the page for screenRect. This should fix the problem...
Grace Kloba [Fri, 12 Mar 2010 22:13:39 +0000 (14:13 -0800)]
Merge "Use the bounds we layout the page for screenRect. This should fix the problem where links in google.com news panel show up in the subwindow."

14 years agoAdd requestCenterFitZoom to android_npapi so that
Grace Kloba [Fri, 12 Mar 2010 21:34:26 +0000 (13:34 -0800)]
Add requestCenterFitZoom to android_npapi so that
Flash can trigger the smart-zoom

Fix http://b/issue?id=2510670

14 years agoUse the bounds we layout the page for screenRect.
Grace Kloba [Fri, 12 Mar 2010 21:54:21 +0000 (13:54 -0800)]
Use the bounds we layout the page for screenRect.
This should fix the problem where links in google.com
news panel show up in the subwindow.

Fix http://b/issue?id=2509597

14 years agoiterate through the children of the label to find its text
Cary Clark [Fri, 12 Mar 2010 20:54:48 +0000 (15:54 -0500)]
iterate through the children of the label to find its text

Change-Id: Ib4d713fd5797d9ee5c0a3080085a2db438c2bee0
http://b/2502091

14 years agoadjust cache layer position by offsets
Cary Clark [Fri, 12 Mar 2010 19:31:50 +0000 (14:31 -0500)]
adjust cache layer position by offsets

Nodes in layers need to know where they are relative to the layers
so that the cursor rings are correctly positioned as the layer
moves (in document coordinates). In addition to tracking the
global offset to make the coordinates relative to any parent
frames, the node is offset relative to the body.

Sometimes layers have zero height or width; in this case, don't
treat them as clips.

Change-Id: Id2811c31a4a0674d316aadda210570ec93311013
http://b/2503096

14 years agoMerge "Outset the rectangle containing the match after checking the clip."
Leon Scroggins [Fri, 12 Mar 2010 16:09:10 +0000 (08:09 -0800)]
Merge "Outset the rectangle containing the match after checking the clip."

14 years agoOutset the rectangle containing the match after checking the clip.
Leon Scroggins [Thu, 11 Mar 2010 21:53:18 +0000 (16:53 -0500)]
Outset the rectangle containing the match after checking the clip.

When considering whether a match is contained by the current clip,
do not outset it first.  Check to see if it is contained by the
clip, and then add the outset so the drawn rectangle is larger.

Fix for http://b/issue?id=2507893

Change-Id: I900d92432d412396e8c5b9e9e341085656ed0a2f

14 years agoMerge "fix nav algorithm when node is clipped"
Cary Clark [Thu, 11 Mar 2010 18:33:39 +0000 (10:33 -0800)]
Merge "fix nav algorithm when node is clipped"

14 years agoMerge "Prepare touch key modifiers for upstreaming."
Ben Murdoch [Thu, 11 Mar 2010 17:39:27 +0000 (09:39 -0800)]
Merge "Prepare touch key modifiers for upstreaming."

14 years agofix nav algorithm when node is clipped
Cary Clark [Thu, 11 Mar 2010 17:25:40 +0000 (12:25 -0500)]
fix nav algorithm when node is clipped

A fix for http://b/2319610 incorrectly retried finding
nodes if the node was clipped, possibly causing infinite
recursion. Leave all of the prior fix but only try once.

I tested the prior test case and ensured that it still
works.

Change-Id: Ib229b6fc0ba57c131a8c1f33350982ac22a445c0
http://b/2501914

14 years agoMerge "Actually call the script when plugins try to invoke scripts from v8 http:...
Leon Clarke [Thu, 11 Mar 2010 17:22:34 +0000 (09:22 -0800)]
Merge "Actually call the script when plugins try to invoke scripts from v8 b/issue?id=2506729"

14 years agoPrepare touch key modifiers for upstreaming.
Ben Murdoch [Mon, 1 Mar 2010 12:14:11 +0000 (12:14 +0000)]
Prepare touch key modifiers for upstreaming.

Fix two compiler warnings in EventHandler.cpp

The webkit bug tracking this was https://bugs.webkit.org/show_bug.cgi?id=35521
and was landed to the WebKit tree as r55843.

Change-Id: Id49ba00d49bf98797a79b6f55b1cb1f906f5e0ad

14 years agoMerge "Marks remaining Android-specific changes to Geolocation with ANDROID"
Steve Block [Thu, 11 Mar 2010 17:06:39 +0000 (09:06 -0800)]
Merge "Marks remaining Android-specific changes to Geolocation with ANDROID"

14 years agoActually call the script when plugins try to invoke scripts from v8
Leon Clarke [Thu, 11 Mar 2010 15:48:11 +0000 (15:48 +0000)]
Actually call the script when plugins try to invoke scripts from v8
http://b/issue?id=2506729

14 years agoset the cached focus state when a node is in a fixed layer
Cary Clark [Wed, 10 Mar 2010 21:47:32 +0000 (16:47 -0500)]
set the cached focus state when a node is in a fixed layer

The newly built nav cache needs the layer information set
before it attempts to set the focus so the nodes' bounds
are set correctly.

Change-Id: I72ff03988c267388da5921ac85b0ac0f6fb2374a
http://2503096

14 years agoMerge "Update Android to reflect final version submitted to webkit.org for Geolocatio...
Steve Block [Thu, 11 Mar 2010 12:02:20 +0000 (04:02 -0800)]
Merge "Update Android to reflect final version submitted to webkit.org for Geolocation maximumAge"

14 years agoMarks remaining Android-specific changes to Geolocation with ANDROID
Steve Block [Thu, 11 Mar 2010 11:48:12 +0000 (11:48 +0000)]
Marks remaining Android-specific changes to Geolocation with ANDROID

Change-Id: I23e2ef2ec23b340e0bfe8bbf0e71e2c675fda676

14 years agoUpdate Android to reflect final version submitted to webkit.org for Geolocation maxim...
Steve Block [Wed, 24 Feb 2010 11:11:00 +0000 (11:11 +0000)]
Update Android to reflect final version submitted to webkit.org for Geolocation maximumAge

See http://trac.webkit.org/changeset/55841

Change-Id: Ib504563b7be40b0ff794cdcc625b07f494bb6483

14 years agoAdd clean steps for webkit changes
Kenny Root [Wed, 10 Mar 2010 21:17:22 +0000 (13:17 -0800)]
Add clean steps for webkit changes

Change-Id: Ia45d3c18906da5fcb0c4a089f6c6278dc1074b6b

14 years agoMerge "Use a paint with 80% alpha instead of saveLayer for disabled elements."
Leon Scroggins [Wed, 10 Mar 2010 20:07:27 +0000 (12:07 -0800)]
Merge "Use a paint with 80% alpha instead of saveLayer for disabled elements."

14 years agoMerge "Switch default build to use V8 on Passion and Sholes. A small step in the...
Bart Sears [Wed, 10 Mar 2010 19:04:55 +0000 (11:04 -0800)]
Merge "Switch default build to use V8 on Passion and Sholes. A small step in the makefile, a giant step for the browser."

14 years agoUse a paint with 80% alpha instead of saveLayer for disabled elements.
Leon Scroggins [Wed, 10 Mar 2010 18:51:22 +0000 (13:51 -0500)]
Use a paint with 80% alpha instead of saveLayer for disabled elements.

Fix for http://b/issue?id=2504048

Change-Id: Icb01fd90e428fde9f82ae6416dd4f9ee66e76ed6

14 years agoMerge "When starting the Geolocation service, we need to make sure that we have a...
Steve Block [Wed, 10 Mar 2010 17:19:11 +0000 (09:19 -0800)]
Merge "When starting the Geolocation service, we need to make sure that we have a valid FrameView before querying the Java side to determine whether the WebView is paused."

14 years agoSwitch default build to use V8 on Passion and Sholes.
Andrei Popescu [Wed, 10 Mar 2010 14:19:32 +0000 (14:19 +0000)]
Switch default build to use V8 on Passion and Sholes.
A small step in the makefile, a giant step for the browser.

14 years agoUpdate android_npapi to remove the special code
Grace Kloba [Tue, 9 Mar 2010 22:11:37 +0000 (14:11 -0800)]
Update android_npapi to remove the special code
added for mwc.

14 years agoIf we do change to document focus in mouse down,
Grace Kloba [Tue, 9 Mar 2010 22:04:10 +0000 (14:04 -0800)]
If we do change to document focus in mouse down,
set default handled so that EventHandler won't try
to change the document focus.

This should address the problem Adobe saw when tapping
on a object may cause both GainFocus and LostFocus
life cycle events.

14 years agoWhen starting the Geolocation service, we need to make sure that we have a valid...
Steve Block [Tue, 9 Mar 2010 21:24:23 +0000 (21:24 +0000)]
When starting the Geolocation service, we need to make sure that we have a valid FrameView before querying the Java side to determine whether the WebView is paused.

Bug: 2501254
Change-Id: Ideb86ce4c35efb7bc7595c4205dc1e8d5c1cf106

14 years agoOnly call plugin touch code if it is in full screen
Leon Scroggins [Fri, 5 Mar 2010 16:39:01 +0000 (11:39 -0500)]
Only call plugin touch code if it is in full screen
mode or it is the document focused node.

Remove Android special code added to support prevent
default for double tap and long press. Long press
and double tap needs to call preventDefault if it
is needed.

Remove Android special code added to pass event time.

Need a separate CL for android_npapi.

14 years agoFix bug 'Children of fixed elements do not always remain fixed themselves'
Nicolas Roard [Fri, 5 Mar 2010 14:41:28 +0000 (14:41 +0000)]
Fix bug 'Children of fixed elements do not always remain fixed themselves'
The problem was twofold:
- webkit didn't create composited layers of the children div unless they were
  intersecting with the fixed layer
- the children divs layers are siblings, not children of the fixed layer

The solution is to:
1/ mark layers as needed to be composited if their ancestor is a
   fixed element (in RenderLayerCompositor)
2/ as the GraphicsLayer/LayerAndroid hierarchy is based on the RenderLayer
   hierarchy (z-order..) and not the display hierarchy, we need to a way of
   updating the position of the contained layers when a fixed layer move.

We do that by:
- marking layers contained in a fixed layer as being linked to the
  fixed layer (GraphicsLayerAndroid::syncFixedDescendants), and set the
  offset between the layer and the fixed layer.
- when pushing the layers tree to the UI side, we ensure that such layers
  are linked to their corresponding fixed layer
  (LayerAndroid::ensureFixedLayersForDescendants)
- when we draw, we do a first pass to update the fixed layers position
  (LayerAndroid::updateFixedLayersPositions) then update the rest of the
  layers (LayerAndroid::updatePositions). The layers that are linked
  to the fixed layers will then update their position relative to it,
  using the original offset between the fixed layer and the layer.

Bug:2470701
Change-Id: I512966df94de6a5f84aff335c5d09b3f027bc2c3

14 years agobuild-fix for the no-layers config
Mike Reed [Tue, 9 Mar 2010 17:58:55 +0000 (12:58 -0500)]
build-fix for the no-layers config

Change-Id: I4994cb5dc61cb845db76060fbd7680f6f5f9ab88

14 years agorequestChildRectangleOnScreen takes layers into account
Cary Clark [Tue, 9 Mar 2010 14:17:51 +0000 (09:17 -0500)]
requestChildRectangleOnScreen takes layers into account

Make native LayerAndroid::subtractLayers visible to java so
the layers can be removed from the visible portion of the screen
when computing how much to scroll.

companion fix in frameworks/base

Change-Id: I3bc0760aaeb515415e90996ec1257d24f9f8705a
http://b/2453841

14 years agoMerge "fix to ensure plugins are not shown when they are marked hidden by the DOM."
Derek Sollenberger [Tue, 9 Mar 2010 13:47:23 +0000 (05:47 -0800)]
Merge "fix to ensure plugins are not shown when they are marked hidden by the DOM."

14 years agoMerge "Adds some logging to JObjectWrapper"
Steve Block [Tue, 9 Mar 2010 13:45:55 +0000 (05:45 -0800)]
Merge "Adds some logging to JObjectWrapper"

14 years agoAdds some logging to JObjectWrapper
Steve Block [Mon, 8 Mar 2010 18:13:21 +0000 (18:13 +0000)]
Adds some logging to JObjectWrapper

This is an attempt to track down the cause of a crashing bug in the WebCore JSC
bridge, which is only occasionaly triggered by the monkey.

Bug: 2435360
Change-Id: I35069c2aff3d9cdf2cfd49264715c418911c9d9c

14 years agofix to ensure plugins are not shown when they are marked hidden by the DOM.
Derek Sollenberger [Mon, 8 Mar 2010 21:11:27 +0000 (16:11 -0500)]
fix to ensure plugins are not shown when they are marked hidden by the DOM.

Change-Id: Ic32359b699d71dcd1603ac9295c95100014fc484

14 years agoAdd an empty CleanSpec.mk
Jean-Baptiste Queru [Tue, 9 Mar 2010 02:04:47 +0000 (18:04 -0800)]
Add an empty CleanSpec.mk

Change-Id: Iafcd71c043fab0ae26f5d38dba1197c50a7279df

14 years agoMerge "Serialize/unserialize screen width scale, which is used when restoring sites."
Shimeng (Simon) Wang [Mon, 8 Mar 2010 17:11:46 +0000 (09:11 -0800)]
Merge "Serialize/unserialize screen width scale, which is used when restoring sites."

14 years agocheck for null node to prevent monkey crash
Cary Clark [Mon, 8 Mar 2010 14:55:23 +0000 (09:55 -0500)]
check for null node to prevent monkey crash

Change-Id: I11625535de5e18f94d293081416eaa9bfa0a8c3c
http://b/2492244

14 years agoscroll into view considers layers
Cary Clark [Fri, 5 Mar 2010 22:51:19 +0000 (17:51 -0500)]
scroll into view considers layers

LayerAndroid.*
- Add a utility that takes the given rectangle and subtracts
  layers that overlay it.

FindCanvas.*
- Add function that returns if the current match is on the
  main page or in a layer.

WebView.cpp
- Add jni caller for WebView.java calcOurContentVisibleRect.
  Call it instead of getVisibleRect, which has the side effect
  of sending messages back to webkit.
- Remove jni caller for getViewMetrics, since it is redundant.
- Only call scrollRectOnScreen for non-layers.

companion fix in framework/base

http://b/2485168

14 years agoSerialize/unserialize screen width scale, which is used when restoring sites.
Shimeng (Simon) Wang [Fri, 5 Mar 2010 23:17:45 +0000 (15:17 -0800)]
Serialize/unserialize screen width scale, which is used when restoring sites.

Also change the scales' default value to 0, so in WebViewCore.java, in case
the scales are not set inside webkit, the default logic can be used.
This helps in case a mobile site is loaded but no scales is saved, and when
restored, the mobile site will have wrong restored scales because default
values are not 0.

14 years agoSet the proper frame's focus controller to active and focused.
Leon Scroggins [Fri, 5 Mar 2010 16:39:01 +0000 (11:39 -0500)]
Set the proper frame's focus controller to active and focused.

When editing text, set the FocusController to both active and
focused, and use the FocusController associated with the focus.

Fixes http://b/issue?id=2251399

14 years agoMerge "Fixes a leaked JNI local reference in MIMETypeRegistry::getMIMETypeForExtension"
Steve Block [Fri, 5 Mar 2010 18:29:52 +0000 (10:29 -0800)]
Merge "Fixes a leaked JNI local reference in MIMETypeRegistry::getMIMETypeForExtension"

14 years agoFixes a leaked JNI local reference in MIMETypeRegistry::getMIMETypeForExtension
Steve Block [Thu, 4 Mar 2010 16:37:50 +0000 (16:37 +0000)]
Fixes a leaked JNI local reference in MIMETypeRegistry::getMIMETypeForExtension

This method was leaking a Java class object. This caused layout test
fast/dom/prototype-inheritance-2.html to crash because the VM ran out of local
references.

Bug: 2218794
Change-Id: I15d6bb8bf16ec58ecd6470ef63e6cd3635d766ef

14 years agoMerge "Remove unnecessary call to ChromeClientAndroid::scheduleCompositingLayerSync()."
Andrei Popescu [Fri, 5 Mar 2010 18:02:54 +0000 (10:02 -0800)]
Merge "Remove unnecessary call to ChromeClientAndroid::scheduleCompositingLayerSync()."

14 years agoRemove unnecessary call to ChromeClientAndroid::scheduleCompositingLayerSync().
Andrei Popescu [Fri, 5 Mar 2010 16:26:29 +0000 (16:26 +0000)]
Remove unnecessary call to ChromeClientAndroid::scheduleCompositingLayerSync().

Fix b: 2471383

14 years agoMerge "Fix history restore."
Patrick Scott [Fri, 5 Mar 2010 16:18:06 +0000 (08:18 -0800)]
Merge "Fix history restore."

14 years agoMerge "forgot header file ; broke build in 43447"
Cary Clark [Fri, 5 Mar 2010 15:29:58 +0000 (07:29 -0800)]
Merge "forgot header file ; broke build in 43447"

14 years agoforgot header file ; broke build in 43447
Cary Clark [Fri, 5 Mar 2010 15:28:17 +0000 (10:28 -0500)]
forgot header file ; broke build in 43447

14 years agoMerge "fixes with find drawing in layers"
Cary Clark [Fri, 5 Mar 2010 15:06:09 +0000 (07:06 -0800)]
Merge "fixes with find drawing in layers"

14 years agofixes with find drawing in layers
Cary Clark [Fri, 5 Mar 2010 14:29:34 +0000 (09:29 -0500)]
fixes with find drawing in layers

Clipping the match picture before drawing prevents layer
pictures from drawing outside of their intended bounds.

Rearranging the layer picture draw call causes it to restore
the canvas matrix, which causes extras like finds and cursor
rings to draw correctly.

Also, added more layer debugging, and a function to show
layer state in the console window.

http://b/2488516
http://b/2488509

14 years agoFix history restore.
Patrick Scott [Thu, 4 Mar 2010 21:15:43 +0000 (16:15 -0500)]
Fix history restore.

goToItem now does a lot of what we were doing and was causing the page to not
load because the current item matched where we were trying to move.

14 years agoMerge "Use the complete url when blocking images."
Patrick Scott [Thu, 4 Mar 2010 16:20:19 +0000 (08:20 -0800)]
Merge "Use the complete url when blocking images."

14 years agoMerge "disable visiblerect debugging printfs"
Mike Reed [Thu, 4 Mar 2010 14:10:31 +0000 (06:10 -0800)]
Merge "disable visiblerect debugging printfs"

14 years agodisable visiblerect debugging printfs
Mike Reed [Thu, 4 Mar 2010 14:09:41 +0000 (09:09 -0500)]
disable visiblerect debugging printfs

14 years agoUse the complete url when blocking images.
Patrick Scott [Thu, 4 Mar 2010 12:54:25 +0000 (07:54 -0500)]
Use the complete url when blocking images.

Image urls that are relative to a base url need to be resolved before looking at
the scheme.

Bug: 2455482

14 years agoDon't leak global weak references when running layout tests
Leon Clarke [Tue, 23 Feb 2010 10:17:11 +0000 (10:17 +0000)]
Don't leak global weak references when running layout tests

14 years agoam fa65eb8a: (-s ours) am 9a77bf56: Revert "Use STLPort instead of our stripped versi...
Jean-Baptiste Queru [Wed, 3 Mar 2010 22:53:45 +0000 (14:53 -0800)]
am fa65eb8a: (-s ours) am 9a77bf56: Revert "Use STLPort instead of our stripped version." DO NOT MERGE.

Merge commit 'fa65eb8a4fe0c48e4c8630850b2492051e5adf5a'

* commit 'fa65eb8a4fe0c48e4c8630850b2492051e5adf5a':
  Revert "Use STLPort instead of our stripped version." DO NOT MERGE.

14 years agoam 9a77bf56: Revert "Use STLPort instead of our stripped version." DO NOT MERGE.
Jean-Baptiste Queru [Wed, 3 Mar 2010 22:49:33 +0000 (14:49 -0800)]
am 9a77bf56: Revert "Use STLPort instead of our stripped version." DO NOT MERGE.

Merge commit '9a77bf563cbe404d041d0df3530cb57ae67013aa' into eclair-mr2-plus-aosp

* commit '9a77bf563cbe404d041d0df3530cb57ae67013aa':
  Revert "Use STLPort instead of our stripped version." DO NOT MERGE.

14 years agoRevert "Use STLPort instead of our stripped version." DO NOT MERGE.
Jean-Baptiste Queru [Wed, 3 Mar 2010 22:27:34 +0000 (14:27 -0800)]
Revert "Use STLPort instead of our stripped version." DO NOT MERGE.

This reverts commit f0073a99fda2bf1acd49fa66d0dd4272d09a1e61.

14 years agoam 381c572c: (-s ours) am f0073a99: Use STLPort instead of our stripped version....
Jean-Baptiste Queru [Wed, 3 Mar 2010 21:51:21 +0000 (13:51 -0800)]
am 381c572c: (-s ours) am f0073a99: Use STLPort instead of our stripped version. DO NOT MERGE.

Merge commit '381c572c6f77708151c42fbda8c8c1ed8e67c9c5'

* commit '381c572c6f77708151c42fbda8c8c1ed8e67c9c5':
  Use STLPort instead of our stripped version. DO NOT MERGE.

14 years agoam dd77efd5: (-s ours) am 60326ce1: Android ICU4.2.1 upgrade. DO NOT MERGE. ...
Claire Ho [Wed, 3 Mar 2010 21:51:18 +0000 (13:51 -0800)]
am dd77efd5: (-s ours) am 60326ce1:     Android ICU4.2.1 upgrade. DO NOT MERGE.     Added "include <config.h>" into WebKit/android/icu/unicode/ucnv.cpp to avoid     compiler error.

Merge commit 'dd77efd5572bacfcd92fc815484a87297403a20d'

* commit 'dd77efd5572bacfcd92fc815484a87297403a20d':
  Android ICU4.2.1 upgrade. DO NOT MERGE.

14 years agoam f0073a99: Use STLPort instead of our stripped version. DO NOT MERGE.
Jean-Baptiste Queru [Wed, 3 Mar 2010 21:40:16 +0000 (13:40 -0800)]
am f0073a99: Use STLPort instead of our stripped version. DO NOT MERGE.

Merge commit 'f0073a99fda2bf1acd49fa66d0dd4272d09a1e61' into eclair-mr2-plus-aosp

* commit 'f0073a99fda2bf1acd49fa66d0dd4272d09a1e61':
  Use STLPort instead of our stripped version. DO NOT MERGE.

14 years agoUse STLPort instead of our stripped version. DO NOT MERGE.
Jean-Baptiste Queru [Wed, 3 Mar 2010 21:08:11 +0000 (13:08 -0800)]
Use STLPort instead of our stripped version. DO NOT MERGE.

This fixes the simulator build and allows us to get closer to sharing code with
chromium.

STLPort was copied with minor edits from the Android port of gears.

14 years agoMerge "stop parsing text for addresses if input element is found"
Cary Clark [Wed, 3 Mar 2010 20:48:58 +0000 (12:48 -0800)]
Merge "stop parsing text for addresses if input element is found"

14 years agostop parsing text for addresses if input element is found
Cary Clark [Wed, 3 Mar 2010 20:41:45 +0000 (15:41 -0500)]
stop parsing text for addresses if input element is found

http://b/2420076

14 years agoMerge "add optional preftable to image codecs, for more control over the resulting...
Mike Reed [Wed, 3 Mar 2010 20:38:38 +0000 (12:38 -0800)]
Merge "add optional preftable to image codecs, for more control over the resulting bitmap config"

14 years agoMerge "Call setCurrentIndex to trigger the client callback."
Patrick Scott [Wed, 3 Mar 2010 20:24:51 +0000 (12:24 -0800)]
Merge "Call setCurrentIndex to trigger the client callback."

14 years agoCall setCurrentIndex to trigger the client callback.
Patrick Scott [Wed, 3 Mar 2010 20:18:19 +0000 (15:18 -0500)]
Call setCurrentIndex to trigger the client callback.

14 years agoMerge "Revive my change to not add anchor navigations to the history in some cases."
Patrick Scott [Wed, 3 Mar 2010 19:05:39 +0000 (11:05 -0800)]
Merge "Revive my change to not add anchor navigations to the history in some cases."

14 years agoadd optional preftable to image codecs, for more control over the resulting bitmap...
Mike Reed [Wed, 3 Mar 2010 18:25:57 +0000 (13:25 -0500)]
add optional preftable to image codecs, for more control over the resulting bitmap config

http://b/issue?id=2267010

14 years agoMerge "Adds Android-specific expected results for some storage LayoutTests when using V8"
Steve Block [Wed, 3 Mar 2010 18:54:53 +0000 (10:54 -0800)]
Merge "Adds Android-specific expected results for some storage LayoutTests when using V8"

14 years agoam 60326ce1: Android ICU4.2.1 upgrade. DO NOT MERGE. Added "include <config...
Claire Ho [Wed, 3 Mar 2010 18:51:53 +0000 (10:51 -0800)]
am 60326ce1:     Android ICU4.2.1 upgrade. DO NOT MERGE.     Added "include <config.h>" into WebKit/android/icu/unicode/ucnv.cpp to avoid     compiler error.

Merge commit '60326ce181ad373e08112435e159a9c71f85a08e' into eclair-mr2-plus-aosp

* commit '60326ce181ad373e08112435e159a9c71f85a08e':
  Android ICU4.2.1 upgrade. DO NOT MERGE.

14 years agofix build
Andrei Popescu [Wed, 3 Mar 2010 18:46:13 +0000 (18:46 +0000)]
fix build

14 years agoMerge "Fix V8 build when instrumentation is on"
Andrei Popescu [Wed, 3 Mar 2010 18:41:52 +0000 (10:41 -0800)]
Merge "Fix V8 build when instrumentation is on"

14 years agoFix V8 build when instrumentation is on
Andrei Popescu [Wed, 3 Mar 2010 18:16:12 +0000 (18:16 +0000)]
Fix V8 build when instrumentation is on

14 years ago Android ICU4.2.1 upgrade. DO NOT MERGE.
Claire Ho [Tue, 24 Nov 2009 22:03:26 +0000 (14:03 -0800)]
Android ICU4.2.1 upgrade. DO NOT MERGE.
    Added "include <config.h>" into WebKit/android/icu/unicode/ucnv.cpp to avoid
    compiler error.

    Bug: 1823720
Reviewed-by: benm@google.com,enh@google.com
    CC:
    Signed-off-by: