OSDN Git Service

android-x86/external-webkit.git
13 years agoam 80aad5ee: (-s ours) am 622d72be: Clear DOM storage when clearing other cached...
Ben Murdoch [Tue, 12 Apr 2011 20:18:31 +0000 (13:18 -0700)]
am 80aad5ee: (-s ours) am 622d72be: Clear DOM storage when clearing other cached data. Do not merge.

* commit '80aad5eedfe9e84cabb8540f818f7d1971ecadc3':
  Clear DOM storage when clearing other cached data. Do not merge.

13 years agoam 622d72be: Clear DOM storage when clearing other cached data. Do not merge.
Ben Murdoch [Tue, 12 Apr 2011 10:13:59 +0000 (03:13 -0700)]
am 622d72be: Clear DOM storage when clearing other cached data. Do not merge.

* commit '622d72beb96a024bb18591feb39d4b21d0a34bd7':
  Clear DOM storage when clearing other cached data. Do not merge.

13 years agoClear DOM storage when clearing other cached data. Do not merge.
Ben Murdoch [Mon, 11 Apr 2011 19:08:03 +0000 (20:08 +0100)]
Clear DOM storage when clearing other cached data. Do not merge.

Empty out DOM local and session storage in addition to HTML5
database and app cache when we get a request from java to
delete cached data.

Cherry pick of I5b6ee075d2a8fb44ee373ad4462a33623c9c2460

Bug: 4267038

Change-Id: Id2dece1963d443c3758710e9c3e871054991ae32

13 years agoam 15692f0c: Merge "Fix regression for superpudu animations" into honeycomb-mr1
Nicolas Roard [Fri, 8 Apr 2011 00:27:03 +0000 (17:27 -0700)]
am 15692f0c: Merge "Fix regression for superpudu animations" into honeycomb-mr1

* commit '15692f0cb85c86ff494b94d4e61ad0757f5238b4':
  Fix regression for superpudu animations

13 years agoMerge "Fix regression for superpudu animations" into honeycomb-mr1
Nicolas Roard [Fri, 8 Apr 2011 00:25:49 +0000 (17:25 -0700)]
Merge "Fix regression for superpudu animations" into honeycomb-mr1

13 years agoFix regression for superpudu animations
Nicolas Roard [Thu, 7 Apr 2011 23:45:06 +0000 (16:45 -0700)]
Fix regression for superpudu animations

bug:4181605 bug:3389597
Change-Id: I2717cdfdeea12df434a119bb8dcc9d2f17f07a7e
(cherry picked from commit 14866c7d6aaa0930627503d0b49bde8e21e90a6e)

13 years agoam b6ac9086: Merge "Fix for bug:4183801" into honeycomb-mr1
Nicolas Roard [Fri, 8 Apr 2011 00:24:07 +0000 (17:24 -0700)]
am b6ac9086: Merge "Fix for bug:4183801" into honeycomb-mr1

* commit 'b6ac908669555fedfaf5dc648367d4ad459fc53e':
  Fix for bug:4183801

13 years agoMerge "Fix for bug:4183801" into honeycomb-mr1
Nicolas Roard [Fri, 8 Apr 2011 00:21:50 +0000 (17:21 -0700)]
Merge "Fix for bug:4183801" into honeycomb-mr1

13 years agoFix for bug:4183801
Nicolas Roard [Tue, 5 Apr 2011 01:40:11 +0000 (18:40 -0700)]
Fix for bug:4183801

The computation for the invalidated / clipping rects were wrong,
sometimes causing the linked bug when layers had a transparent
background (they were not obeying the clipping rect passed by
the framework).

java counterpart CL:
https://android-git.corp.google.com/g/#change,105503

Change-Id: I60769e7cbf1a3a939724c57b3d3ce63a6f87aa87

13 years agoam 63f5370e: Fix race condition
John Reck [Thu, 7 Apr 2011 23:44:32 +0000 (16:44 -0700)]
am 63f5370e: Fix race condition

* commit '63f5370e1736a4e699dd9ccd643525772d67ffef':
  Fix race condition

13 years agoFix race condition
John Reck [Thu, 7 Apr 2011 23:25:54 +0000 (16:25 -0700)]
Fix race condition

 Bug: 4262018

Change-Id: I0ddda2348e72484cd57fc1909efed47bf8e2222d

13 years agoam e4cbbfc0: Merge "Show the fullscreen button all the time" into honeycomb-mr1
Teng-Hui Zhu [Thu, 7 Apr 2011 19:02:02 +0000 (12:02 -0700)]
am e4cbbfc0: Merge "Show the fullscreen button all the time" into honeycomb-mr1

* commit 'e4cbbfc01b360475415d07bc3a4276fc46e828b9':
  Show the fullscreen button all the time

13 years agoMerge "Show the fullscreen button all the time" into honeycomb-mr1
Teng-Hui Zhu [Thu, 7 Apr 2011 18:59:42 +0000 (11:59 -0700)]
Merge "Show the fullscreen button all the time" into honeycomb-mr1

13 years agoShow the fullscreen button all the time
Teng-Hui Zhu [Wed, 6 Apr 2011 22:01:42 +0000 (15:01 -0700)]
Show the fullscreen button all the time

bug:4232704
Change-Id: I64833baeab708fabe88873573235a896b86b6098

13 years agoam 7404fb69: Merge "Race condition fix" into honeycomb-mr1
John Reck [Wed, 6 Apr 2011 01:12:21 +0000 (18:12 -0700)]
am 7404fb69: Merge "Race condition fix" into honeycomb-mr1

* commit '7404fb69c0389c04c5d20d9d2dff7a4aba1db63a':
  Race condition fix

13 years agoam bd064363: Merge "Fix the precision loss issue" into honeycomb-mr1
Teng-Hui Zhu [Wed, 6 Apr 2011 01:10:24 +0000 (18:10 -0700)]
am bd064363: Merge "Fix the precision loss issue" into honeycomb-mr1

* commit 'bd064363b650efb80cde0685b2260a854311a3bb':
  Fix the precision loss issue

13 years agoMerge "Race condition fix" into honeycomb-mr1
John Reck [Wed, 6 Apr 2011 01:08:35 +0000 (18:08 -0700)]
Merge "Race condition fix" into honeycomb-mr1

13 years agoMerge "Fix the precision loss issue" into honeycomb-mr1
Teng-Hui Zhu [Wed, 6 Apr 2011 01:05:09 +0000 (18:05 -0700)]
Merge "Fix the precision loss issue" into honeycomb-mr1

13 years agoRace condition fix
John Reck [Tue, 5 Apr 2011 22:39:47 +0000 (15:39 -0700)]
Race condition fix

 Bug: 4232693

Change-Id: I273671aea3dc379858f11546e6459fd9e1443d4f

13 years agoFix the precision loss issue
Teng-Hui Zhu [Tue, 5 Apr 2011 00:22:51 +0000 (17:22 -0700)]
Fix the precision loss issue

Instead of scaling the final rect to compare, scaling the dirtyrect.
Because the dirtyrect will be discarded anyway.
This can better preserve the precision the final rect.

bug:4186567
Change-Id: Icade7a0b1ff13cd36791c7a3c0e2d624870fb44a

13 years agoam a33d4903: Fix tile vanishing act
John Reck [Tue, 5 Apr 2011 02:00:14 +0000 (19:00 -0700)]
am a33d4903: Fix tile vanishing act

* commit 'a33d490302a04453e2c671114d30a4ee93cb08f1':
  Fix tile vanishing act

13 years agoFix tile vanishing act
John Reck [Tue, 5 Apr 2011 01:18:37 +0000 (18:18 -0700)]
Fix tile vanishing act

 Bug: 4184047
 This is caused by textures being marked as unused by swapPages().
 Limiting when swapPages() releases textures slightly fixes the issue.

Change-Id: I8ac94f1370b023355f4b6f5877c3929b5d6d3ef6

13 years agoam ca84feeb: Merge "Fix animations bug" into honeycomb-mr1
Nicolas Roard [Fri, 1 Apr 2011 18:05:30 +0000 (11:05 -0700)]
am ca84feeb: Merge "Fix animations bug" into honeycomb-mr1

* commit 'ca84feebe782358e319da8bd13938179e941f9b4':
  Fix animations bug

13 years agoMerge "Fix animations bug" into honeycomb-mr1
Nicolas Roard [Fri, 1 Apr 2011 17:53:22 +0000 (10:53 -0700)]
Merge "Fix animations bug" into honeycomb-mr1

13 years agoFix animations bug
Nicolas Roard [Fri, 1 Apr 2011 01:56:54 +0000 (18:56 -0700)]
Fix animations bug

This solves the problem on the nytimes.com/skimmer website.
We were applying the last animation step twice.

bug:4181605
Change-Id: Id2c003b5a2b6a8c56aa3d6be76d4a6cccfb86ad8

13 years agoam 39ff342c: Fix regression impacting Flash performances
Nicolas Roard [Thu, 31 Mar 2011 23:14:55 +0000 (16:14 -0700)]
am 39ff342c: Fix regression impacting Flash performances

* commit '39ff342cb3755bb7529c775a6d15b1ccbbe9be6a':
  Fix regression impacting Flash performances

13 years agoFix regression impacting Flash performances
Nicolas Roard [Thu, 31 Mar 2011 22:54:41 +0000 (15:54 -0700)]
Fix regression impacting Flash performances

We marked layers dirty when they did not
have a texture -- even the layers that
did not *need* a texture...

bug:4195690
Change-Id: Ic2a0fdf9462277d060f07ef7d281f5d27e7e8ed5

13 years agoam bd14a6fc: Merge "try faster dashing" into honeycomb-mr1
Bart Sears [Fri, 25 Mar 2011 20:54:22 +0000 (13:54 -0700)]
am bd14a6fc: Merge "try faster dashing" into honeycomb-mr1

* commit 'bd14a6fcc48d3d5713574a34c441f84296a03c0b':
  try faster dashing

13 years agoMerge "try faster dashing" into honeycomb-mr1
Bart Sears [Fri, 25 Mar 2011 20:52:28 +0000 (13:52 -0700)]
Merge "try faster dashing" into honeycomb-mr1

13 years agotry faster dashing
Mike Reed [Fri, 25 Mar 2011 15:51:24 +0000 (11:51 -0400)]
try faster dashing

if we're drawing a horizontal line, try using a bitmap-shader to simulate the
dash, as this can be much faster than the general SkDashPathEffect.

bug:4163023
Change-Id: I362543d6efb83ebf395cbe92c2d889c590a7c2df

13 years agoam 92c3acf6: Fix error where mediaLayer is derefed on UI thread prior to being derefe...
Derek Sollenberger [Fri, 25 Mar 2011 17:35:14 +0000 (10:35 -0700)]
am 92c3acf6: Fix error where mediaLayer is derefed on UI thread prior to being derefed in Webkit.

* commit '92c3acf6b42939215be09fe379bfef07ab703216':
  Fix error where mediaLayer is derefed on UI thread prior to being derefed in Webkit.

13 years agoFix error where mediaLayer is derefed on UI thread prior to being derefed in Webkit.
Derek Sollenberger [Fri, 25 Mar 2011 14:43:12 +0000 (10:43 -0400)]
Fix error where mediaLayer is derefed on UI thread prior to being derefed in Webkit.

bug: 4171470
Change-Id: I6afe957b2eccae249afa46baede822c34f8f1a7f

13 years agoam 557855a0: Fix regression blocking the page cycler The bug caused a continuous...
Nicolas Roard [Thu, 24 Mar 2011 22:17:56 +0000 (15:17 -0700)]
am 557855a0: Fix regression blocking the page cycler The bug caused a continuous redraw if you had a webview without content; the Activity containing the webview would thus not be marked as idle. The pagecycler is launching such an activity and wait until it\'s idle to

* commit '557855a0cf753d49915b11849f3fa4a2c23c2c18':
  Fix regression blocking the page cycler The bug caused a continuous redraw if you had a webview without content; the Activity containing the webview would thus not be marked as idle. The pagecycler is launching such an activity and wait until it's idle to determinate that it's usable to loade urls...

13 years agoFix regression blocking the page cycler
Nicolas Roard [Thu, 24 Mar 2011 21:30:55 +0000 (14:30 -0700)]
Fix regression blocking the page cycler
The bug caused a continuous redraw if you had a webview without content;
the Activity containing the webview would thus not be marked as idle.
The pagecycler is launching such an activity and wait until it's idle
to determinate that it's usable to loade urls...

bug:4172381
Change-Id: I6c838c5f6757186d01e8e476a8fc5101ea00ed15

13 years agoam 909e9c7e: Improve performances for the base surface rendering
Nicolas Roard [Thu, 24 Mar 2011 00:56:40 +0000 (17:56 -0700)]
am 909e9c7e: Improve performances for the base surface rendering

* commit '909e9c7e41ff3d446960cd840fb9aa055fef6812':
  Improve performances for the base surface rendering

13 years agoImprove performances for the base surface rendering
Nicolas Roard [Thu, 24 Mar 2011 00:33:29 +0000 (17:33 -0700)]
Improve performances for the base surface rendering

- don't queue large number of tiles when scrolling/zooming
- while scrolling/zooming, schedule the tiles generation at the front of the queue

bug:4136077
Change-Id: I652464d01f3000e88e42fad1e98402436a85dce4

13 years agoam 0cdc1296: Fix the repainting problems with overflow-scroll div elements
Nicolas Roard [Wed, 23 Mar 2011 23:01:27 +0000 (16:01 -0700)]
am 0cdc1296: Fix the repainting problems with overflow-scroll div elements

* commit '0cdc129636d794dec2649b50a470e03f727258f9':
  Fix the repainting problems with overflow-scroll div elements

13 years agoFix the repainting problems with overflow-scroll div elements
Nicolas Roard [Wed, 23 Mar 2011 22:47:27 +0000 (15:47 -0700)]
Fix the repainting problems with overflow-scroll div elements

bug:4163243 bug:4163233 bug:4159272 bug:4159270 bug:3442104 bug:4163237
Change-Id: Iab71d1da3f0ca6c52b23ddfb8b86c58b118b60e0

13 years agoam 5bdfd148: Merge "Regression fix from bug:4136077 Plus animations improvement bug...
Nicolas Roard [Wed, 23 Mar 2011 22:46:41 +0000 (15:46 -0700)]
am 5bdfd148: Merge "Regression fix from bug:4136077 Plus animations improvement bug:3389597" into honeycomb-mr1

* commit '5bdfd148249f6196a184550910f1cd8cb138fe8c':
  Regression fix from bug:4136077 Plus animations improvement bug:3389597

13 years agoMerge "Regression fix from bug:4136077 Plus animations improvement bug:3389597" into...
Nicolas Roard [Wed, 23 Mar 2011 21:59:41 +0000 (14:59 -0700)]
Merge "Regression fix from bug:4136077 Plus animations improvement bug:3389597" into honeycomb-mr1

13 years agoRegression fix from bug:4136077
Nicolas Roard [Wed, 23 Mar 2011 20:37:26 +0000 (13:37 -0700)]
Regression fix from bug:4136077
Plus animations improvement bug:3389597

Change-Id: I07319910ae366d3e237bf3a1042cf28217434437

13 years agoam 9d342774: Merge "Possible fix for bug 4155219 IPv6 support: DNS is not reset when...
Bart Sears [Wed, 23 Mar 2011 20:07:17 +0000 (13:07 -0700)]
am 9d342774: Merge "Possible fix for bug 4155219 IPv6 support: DNS is not reset when switching from IPv4/IPv6 network to IPv4 network" into honeycomb-mr1

* commit '9d3427742106914aa11f06f5cf967a0d538f8d24':
  Possible fix for bug 4155219 IPv6 support: DNS is not reset when switching from IPv4/IPv6 network to IPv4 network

13 years agoMerge "Possible fix for bug 4155219 IPv6 support: DNS is not reset when switching...
Bart Sears [Wed, 23 Mar 2011 20:05:05 +0000 (13:05 -0700)]
Merge "Possible fix for bug 4155219 IPv6 support: DNS is not reset when switching from IPv4/IPv6 network to IPv4 network" into honeycomb-mr1

13 years agoam f32b2ce6: Fix GPU leak where plugin textures in the WebKit thread were not being...
Derek Sollenberger [Wed, 23 Mar 2011 19:43:46 +0000 (12:43 -0700)]
am f32b2ce6: Fix GPU leak where plugin textures in the WebKit thread were not being deleted.

* commit 'f32b2ce68c45663b1193cfe45a10d2889fd61c81':
  Fix GPU leak where plugin textures in the WebKit thread were not being deleted.

13 years agoFix GPU leak where plugin textures in the WebKit thread were not being deleted.
Derek Sollenberger [Wed, 23 Mar 2011 15:19:05 +0000 (11:19 -0400)]
Fix GPU leak where plugin textures in the WebKit thread were not being deleted.

bug: 4165067
Change-Id: I2f50f600133300cec712b0177ead1a7afbbee2c3

13 years agoPossible fix for bug 4155219 IPv6 support: DNS is not reset when switching from IPv4...
Kristian Monsen [Wed, 23 Mar 2011 18:13:52 +0000 (18:13 +0000)]
Possible fix for bug 4155219 IPv6 support: DNS is not reset when switching from IPv4/IPv6 network to IPv4 network

Enable network change notifier

Change-Id: I89919c8a8bc54aaf0b2e1b013fbdf43abe657295

13 years agoam ef25442f: don\'t call bounder when uninitialized
Cary Clark [Wed, 23 Mar 2011 15:35:28 +0000 (08:35 -0700)]
am ef25442f: don\'t call bounder when uninitialized

* commit 'ef25442f9f65d5f8283b2a5c2b90e63bf875c694':
  don't call bounder when uninitialized

13 years agodon't call bounder when uninitialized
Cary Clark [Wed, 23 Mar 2011 14:02:01 +0000 (10:02 -0400)]
don't call bounder when uninitialized

If the canvas' device list has never been set up, or
has been set to an empty clip, the bounder's clip will not be
initialized, and checking bounding rect may fail.

Check to see if the bounder was ever called for an individual
glyph before checking the result.

bug:4155270

Change-Id: I625e885d13208e82c44259277c74b8f0cef5b0cb

13 years agoam 77274212: Fix performance problems with layers
Nicolas Roard [Wed, 23 Mar 2011 07:01:46 +0000 (00:01 -0700)]
am 77274212: Fix performance problems with layers

* commit '772742124c1e5c700120019397dc8af29ced74c0':
  Fix performance problems with layers

13 years agoFix performance problems with layers
Nicolas Roard [Tue, 22 Mar 2011 21:29:12 +0000 (14:29 -0700)]
Fix performance problems with layers

We were preventing the layers' positions to be updated
while waiting for the base surface to render.
This also fix some repainting issues.

bug:4136077
Change-Id: I6870faf8296e205c6ee58e80b1cf95833a67814b

13 years agoam 55c8e9b1: Merge "Workaround for poor dashed line performance" into honeycomb-mr1
John Reck [Wed, 23 Mar 2011 02:31:35 +0000 (19:31 -0700)]
am 55c8e9b1: Merge "Workaround for poor dashed line performance" into honeycomb-mr1

* commit '55c8e9b1b75caff31e804a6f4da404da9f67af93':
  Workaround for poor dashed line performance

13 years agoMerge "Workaround for poor dashed line performance" into honeycomb-mr1
John Reck [Wed, 23 Mar 2011 02:29:49 +0000 (19:29 -0700)]
Merge "Workaround for poor dashed line performance" into honeycomb-mr1

13 years agoWorkaround for poor dashed line performance
John Reck [Wed, 23 Mar 2011 02:14:16 +0000 (19:14 -0700)]
Workaround for poor dashed line performance

 Bug: 4163023

Change-Id: Ie76a2f5855130ff174a80d1c7e9187bea89f3aa5

13 years agoam 50e0a1d1: Fix a browser ANR
Nicolas Roard [Tue, 22 Mar 2011 20:51:06 +0000 (13:51 -0700)]
am 50e0a1d1: Fix a browser ANR

* commit '50e0a1d1948f6b53b9b6ea82f865812a7b42221b':
  Fix a browser ANR

13 years agoFix a browser ANR
Nicolas Roard [Tue, 22 Mar 2011 18:30:52 +0000 (11:30 -0700)]
Fix a browser ANR

Layer release textures (setting their owner to zero). Texture with no
owners are routinely destroyed by the TilesManager.
When a texture is busy, the release is delayed until the end of the
painting; if the owner of the texture is still the same, we then
remove it.

The problem was that a layer could release a busy texture, thereby
delaying the release; and in a next cycle re-acquire that same texture.
The delayed release would then kicks in, and proceed, as the owner
was still the same one as the one that originally asked for the release.

We would then have a situation where the layer thought it acquired a
texture, while that texture's owner was ultimately set to zero.
The TilesManager would then proceed to destroy that texture.
The next time the layer tried to use the texture, it would either crash
or more often block on acquiring a lock on it, causing an ANR.

bug:3472320
Change-Id: I84a064b1aa6636a18e0518e34c8572cbea3eb558

13 years agoam 983c9f05: Merge "Fix invals for CSS3D" into honeycomb-mr1
Nicolas Roard [Tue, 22 Mar 2011 01:15:12 +0000 (18:15 -0700)]
am 983c9f05: Merge "Fix invals for CSS3D" into honeycomb-mr1

* commit '983c9f05fadaaceaa011dec0e5bcff2bd2b4199f':
  Fix invals for CSS3D

13 years agoMerge "Fix invals for CSS3D" into honeycomb-mr1
Nicolas Roard [Tue, 22 Mar 2011 01:02:28 +0000 (18:02 -0700)]
Merge "Fix invals for CSS3D" into honeycomb-mr1

13 years agoFix invals for CSS3D
Nicolas Roard [Tue, 22 Mar 2011 00:22:17 +0000 (17:22 -0700)]
Fix invals for CSS3D

bug:4129388
Change-Id: I52963d38412d0fd69c4695a281ad6558c27999ed

13 years agoam 47817e2e: Merge "Support drawing paused and loading image as GL texture" into...
Teng-Hui Zhu [Mon, 21 Mar 2011 21:19:59 +0000 (14:19 -0700)]
am 47817e2e: Merge "Support drawing paused and loading image as GL texture" into honeycomb-mr1

* commit '47817e2ea523c5798c192ec9005238611b3910eb':
  Support drawing paused and loading image as GL texture

13 years agoMerge "Support drawing paused and loading image as GL texture" into honeycomb-mr1
Teng-Hui Zhu [Mon, 21 Mar 2011 21:01:06 +0000 (14:01 -0700)]
Merge "Support drawing paused and loading image as GL texture" into honeycomb-mr1

13 years agoam 4c6c5a28: Merge "Turns out args should not be decoded twice." into honeycomb-mr1
Bart Sears [Mon, 21 Mar 2011 18:27:40 +0000 (11:27 -0700)]
am 4c6c5a28: Merge "Turns out args should not be decoded twice." into honeycomb-mr1

* commit '4c6c5a28199f4770d1422214fcc76a61b5afa42c':
  Turns out args should not be decoded twice.

13 years agoMerge "Turns out args should not be decoded twice." into honeycomb-mr1
Bart Sears [Mon, 21 Mar 2011 18:23:32 +0000 (11:23 -0700)]
Merge "Turns out args should not be decoded twice." into honeycomb-mr1

13 years agoSupport drawing paused and loading image as GL texture
Teng-Hui Zhu [Fri, 18 Mar 2011 22:41:20 +0000 (15:41 -0700)]
Support drawing paused and loading image as GL texture

bug:4142131

Change-Id: I871924cb7587077f50fdf4f9a00056f795a8daea

13 years agoam e1c3e599: Fix native crash Don\'t save the extras in the picture when drawing.
Nicolas Roard [Mon, 21 Mar 2011 14:32:59 +0000 (07:32 -0700)]
am e1c3e599: Fix native crash Don\'t save the extras in the picture when drawing.

* commit 'e1c3e59945c9789e93c417e064a9abed1057ce42':
  Fix native crash Don't save the extras in the picture when drawing.

13 years agoFix native crash
Nicolas Roard [Sun, 20 Mar 2011 02:01:15 +0000 (19:01 -0700)]
Fix native crash
Don't save the extras in the picture when drawing.

bug:4126884 bug:4132721
Change-Id: I52c46a33f847e64c1f8245a0bb84445a948d72a4

13 years agoam a5ec3c20: Fix ANR introduced by recent CL... before locking a texture, it needs...
Nicolas Roard [Sat, 19 Mar 2011 21:33:33 +0000 (14:33 -0700)]
am a5ec3c20: Fix ANR introduced by recent CL... before locking a texture, it needs to be ready.

* commit 'a5ec3c20d40b576e2f6878c615cc7c63eebe3a3f':
  Fix ANR introduced by recent CL... before locking a texture, it needs to be ready.

13 years agoFix ANR introduced by recent CL...
Nicolas Roard [Sat, 19 Mar 2011 00:41:11 +0000 (17:41 -0700)]
Fix ANR introduced by recent CL...
before locking a texture, it needs to be ready.

bug:4136077
Change-Id: Ib8235694e1a43524c52a02a5f288235d56088f46

13 years agoam 93a2b8c5: Fix for layers textures selection
Nicolas Roard [Sat, 19 Mar 2011 00:06:52 +0000 (17:06 -0700)]
am 93a2b8c5: Fix for layers textures selection

* commit '93a2b8c50e4f5f46e9d345c259e20192e47443a5':
  Fix for layers textures selection

13 years agoFix for layers textures selection
Nicolas Roard [Fri, 18 Mar 2011 23:58:13 +0000 (16:58 -0700)]
Fix for layers textures selection

bug:4136077
Change-Id: I2dabdf612e29c4e4c9a0b744ddecaeda7a537ba5

13 years agoam ce96fa32: Merge "Do not recreate textures if not needed for layers..." into honeyc...
Nicolas Roard [Fri, 18 Mar 2011 23:54:00 +0000 (16:54 -0700)]
am ce96fa32: Merge "Do not recreate textures if not needed for layers..." into honeycomb-mr1

* commit 'ce96fa326fe12ce86fdb1a7b42032a86c1c2382a':
  Do not recreate textures if not needed for layers...

13 years agoam b1444e52: Fix for FBO offset when drawing with alpha
Russell Brenner [Fri, 18 Mar 2011 23:53:58 +0000 (16:53 -0700)]
am b1444e52: Fix for FBO offset when drawing with alpha

* commit 'b1444e524d3f5e888d1ae3300b2658321fc56349':
  Fix for FBO offset when drawing with alpha

13 years agoam 9af1c3b9: Merge "Fix for missing invals" into honeycomb-mr1
Nicolas Roard [Fri, 18 Mar 2011 23:53:25 +0000 (16:53 -0700)]
am 9af1c3b9: Merge "Fix for missing invals" into honeycomb-mr1

* commit '9af1c3b94981315f36e34dea401adedbbb98d7da':
  Fix for missing invals

13 years agoMerge "Do not recreate textures if not needed for layers..." into honeycomb-mr1
Nicolas Roard [Fri, 18 Mar 2011 23:46:56 +0000 (16:46 -0700)]
Merge "Do not recreate textures if not needed for layers..." into honeycomb-mr1

13 years agoDo not recreate textures if not needed for layers...
Nicolas Roard [Fri, 18 Mar 2011 23:32:34 +0000 (16:32 -0700)]
Do not recreate textures if not needed for layers...

bug:4136077
Change-Id: Ia975ac78f724450be6f36a61316c872c4d88d7a0

13 years agoFix for FBO offset when drawing with alpha
Russell Brenner [Fri, 18 Mar 2011 20:57:51 +0000 (13:57 -0700)]
Fix for FBO offset when drawing with alpha

Use new DrawGlInfo struct to detect when drawing to an FBO and adjust
topLeft of destination to (0,0).

Bug: 3275491
Change-Id: Ia3ef85d6a21475dbd27743a15edd9d16d0d9cc9e

13 years agoMerge "Fix for missing invals" into honeycomb-mr1
Nicolas Roard [Fri, 18 Mar 2011 22:50:54 +0000 (15:50 -0700)]
Merge "Fix for missing invals" into honeycomb-mr1

13 years agoFix for missing invals
Nicolas Roard [Fri, 18 Mar 2011 22:42:12 +0000 (15:42 -0700)]
Fix for missing invals

bug:4124433
Change-Id: I8cc7203dad408eff30da33f1c9a0a77dd7c97d66

13 years agoam 885c7dee: Merge "DO NOT MERGE Use libhwui\'s new private header. Bug #3275491...
Russell Brenner [Fri, 18 Mar 2011 22:34:22 +0000 (15:34 -0700)]
am 885c7dee: Merge "DO NOT MERGE Use libhwui\'s new private header. Bug #3275491" into honeycomb-mr1

* commit '885c7dee9944991f07e34aefe0b303d5b98ef560':
  DO NOT MERGE Use libhwui's new private header. Bug #3275491

13 years agoMerge "DO NOT MERGE Use libhwui's new private header. Bug #3275491" into honeycomb-mr1
Russell Brenner [Fri, 18 Mar 2011 22:28:44 +0000 (15:28 -0700)]
Merge "DO NOT MERGE Use libhwui's new private header. Bug #3275491" into honeycomb-mr1

13 years agoDO NOT MERGE Use libhwui's new private header.
Romain Guy [Thu, 17 Mar 2011 18:08:18 +0000 (11:08 -0700)]
DO NOT MERGE Use libhwui's new private header.
Bug #3275491

The DrawGlInfo struct is used by the OpenGL UI renderer and WebView to exchange
information. This new version of the struct adds information about the render
target and the current transform set on the Canvas.

Change-Id: Icc3bcc278898493ef4fe0f4574b231e6cfd6aca5

13 years agoam 776c5d44: Limit the SkPicture size we use to paint. This improves performances...
Nicolas Roard [Fri, 18 Mar 2011 02:05:56 +0000 (19:05 -0700)]
am 776c5d44: Limit the SkPicture size we use to paint. This improves performances by a visible factor on some websites.

* commit '776c5d4485a2b75fb4e143b0a5984cba47aaac4d':
  Limit the SkPicture size we use to paint. This improves performances by a visible factor on some websites.

13 years agoLimit the SkPicture size we use to paint.
Nicolas Roard [Fri, 18 Mar 2011 01:20:38 +0000 (18:20 -0700)]
Limit the SkPicture size we use to paint.
This improves performances by a visible factor on some websites.

bug:4124418
Change-Id: I9b24408b686e24cf323eb5bae977d74a0f2e2a34

13 years agoam 1ae2d07b: Merge "Enabled shadow drawing" into honeycomb-mr1
Russell Brenner [Thu, 17 Mar 2011 21:45:30 +0000 (14:45 -0700)]
am 1ae2d07b: Merge "Enabled shadow drawing" into honeycomb-mr1

* commit '1ae2d07bc763fbb9dc3df9a62631d2f8fb707ce2':
  Enabled shadow drawing

13 years agoMerge "Enabled shadow drawing" into honeycomb-mr1
Russell Brenner [Thu, 17 Mar 2011 21:43:06 +0000 (14:43 -0700)]
Merge "Enabled shadow drawing" into honeycomb-mr1

13 years agoEnabled shadow drawing
Russell Brenner [Thu, 17 Mar 2011 19:06:52 +0000 (12:06 -0700)]
Enabled shadow drawing

Reworked the logic of setupForText() and setupShadowPaint() to render
correctly, as exhibited by http://acid3.acidtests.org.

Bug: 4075135

Change-Id: I595456b658d0872183dedcbc726be44783440f89

13 years agoam 439cfed4: Fix Browser ANR
Nicolas Roard [Thu, 17 Mar 2011 18:54:26 +0000 (11:54 -0700)]
am 439cfed4: Fix Browser ANR

* commit '439cfed476441572caa4206a622e3ef3c188798b':
  Fix Browser ANR

13 years agoTurns out args should not be decoded twice.
Patrick Scott [Thu, 17 Mar 2011 17:43:29 +0000 (13:43 -0400)]
Turns out args should not be decoded twice.

This should help out in some cases with redirect/login loops.

Bug: 4110115
Change-Id: I42fff7e9227423b9b5ce94234ad6d606234fe252

13 years agoFix Browser ANR
Nicolas Roard [Thu, 17 Mar 2011 17:31:57 +0000 (10:31 -0700)]
Fix Browser ANR

The problem was that when attempting to forcefully destroy a texture
in TilesManager::cleanupLayersTextures(), the call to
LayerAndroid::removeTexture() may fail if the texture was busy being
painted -- the call to BackedDoubleBufferedTexture::release() would
return false, and the layer may thus still keep a pointer to the texture.
But the release() call, while indicating it failed, was only delaying
the release -- as soon as the texture was marked as not busy, it
could set its owner to nil.

We could thus have a situation where the layer did not reset its
texture pointers because the owner of the texture was not yet
changed, but the texture would then reset its owner to nil as soon
as it was not busy painting.

In TilesManager::cleanupLayersTexture() the next step before deleting
a texture is to check that the texture does not have an owner -- but
by then the texture could have been marked as not busy, and removed
its owner, letting TilesManager destroying it.

bug:3472320
Change-Id: I3bcf169b30dfacba1773d3b79a3c0d205bf3cbdb

13 years agoam 9f5143f9: Implement partial screen invalidations
Nicolas Roard [Thu, 17 Mar 2011 06:51:33 +0000 (23:51 -0700)]
am 9f5143f9: Implement partial screen invalidations

* commit '9f5143f9ae49a8e5fdb7ea626c4efad66096b020':
  Implement partial screen invalidations

13 years agoImplement partial screen invalidations
Nicolas Roard [Wed, 16 Mar 2011 01:20:42 +0000 (18:20 -0700)]
Implement partial screen invalidations

bug:3461349

Change-Id: Id654d176c58027c67be7cb604b87c0ec68984525

13 years agoam 8e1d1088: Pass the URL at the enterfullscreen time
Teng-Hui Zhu [Thu, 17 Mar 2011 06:03:10 +0000 (23:03 -0700)]
am 8e1d1088: Pass the URL at the enterfullscreen time

* commit '8e1d10880da3108f7eb53ae81682a7c4192a2e3f':
  Pass the URL at the enterfullscreen time

13 years agoPass the URL at the enterfullscreen time
Teng-Hui Zhu [Tue, 15 Mar 2011 01:16:58 +0000 (18:16 -0700)]
Pass the URL at the enterfullscreen time

bug:2126902

Change-Id: I84a27fabccc32ef85c8d78d6702d22ffca497db7

13 years agoam 9863c30d: Merge "Fix page jump after clicking links in nytimes.com" into honeycomb-mr1
Shimeng (Simon) Wang [Thu, 17 Mar 2011 00:26:52 +0000 (17:26 -0700)]
am 9863c30d: Merge "Fix page jump after clicking links in nytimes.com" into honeycomb-mr1

* commit '9863c30dfa5b5a03577c0ad069bb04d70435ac2d':
  Fix page jump after clicking links in nytimes.com

13 years agoam 6e9c5336: Merge "Pass the clip rect to WebView to optimize rendering" into honeyco...
Romain Guy [Thu, 17 Mar 2011 00:26:46 +0000 (17:26 -0700)]
am 6e9c5336: Merge "Pass the clip rect to WebView to optimize rendering" into honeycomb-mr1

* commit '6e9c5336ef1a230b1cabc7c55ef19aeee6c00f55':
  Pass the clip rect to WebView to optimize rendering

13 years agoMerge "Fix page jump after clicking links in nytimes.com" into honeycomb-mr1
Shimeng (Simon) Wang [Thu, 17 Mar 2011 00:24:44 +0000 (17:24 -0700)]
Merge "Fix page jump after clicking links in nytimes.com" into honeycomb-mr1

13 years agoMerge "Pass the clip rect to WebView to optimize rendering" into honeycomb-mr1
Romain Guy [Thu, 17 Mar 2011 00:24:35 +0000 (17:24 -0700)]
Merge "Pass the clip rect to WebView to optimize rendering" into honeycomb-mr1

13 years agoFix page jump after clicking links in nytimes.com
Shimeng (Simon) Wang [Thu, 17 Mar 2011 00:13:12 +0000 (17:13 -0700)]
Fix page jump after clicking links in nytimes.com

This reverts parts of change in
https://android-git.corp.google.com/g/#change,96777

issue: 4024447
Change-Id: I29b5de1f97ed9d7cd70786cdcfeb91b53563e66e

13 years agoam d6f0aa4a: Merge "Fix crash in WebCore::GlyphPageTreeNode::getChild" into honeycomb-mr1
Russell Brenner [Wed, 16 Mar 2011 23:02:03 +0000 (16:02 -0700)]
am d6f0aa4a: Merge "Fix crash in WebCore::GlyphPageTreeNode::getChild" into honeycomb-mr1

* commit 'd6f0aa4a5a76d43d382dad30b31541fc7372206b':
  Fix crash in WebCore::GlyphPageTreeNode::getChild

13 years agoMerge "Fix crash in WebCore::GlyphPageTreeNode::getChild" into honeycomb-mr1
Russell Brenner [Wed, 16 Mar 2011 23:00:54 +0000 (16:00 -0700)]
Merge "Fix crash in WebCore::GlyphPageTreeNode::getChild" into honeycomb-mr1

13 years agoFix crash in WebCore::GlyphPageTreeNode::getChild
Russell Brenner [Fri, 4 Mar 2011 02:29:26 +0000 (18:29 -0800)]
Fix crash in WebCore::GlyphPageTreeNode::getChild

Using existing Skia APIs, FontCache::createFontPlatformData can detect
when the retrieved font is the Skia default and will reject that choice,
allowing WebCore to try the next font on the CSS fallback list. If the
requested family is a generic family, e.g. "serif" or "monotype", the
Skia default will be accepted.

A prior attempt at this fix was not properly handling bold and italic
stylings, which could eventually lead to dangling references to freed
font data used for layout while custom fonts were loading, causing
segfaults in getChild and elsewhere during page rendering.

Bug: 3469204
Bug: 2720133

Change-Id: I3a9e746e18fd27848d0c4b9e1232392df33254a6