OSDN Git Service
Derek Sollenberger [Thu, 24 Feb 2011 13:27:46 +0000 (05:27 -0800)]
Merge "Skia Merge (revision 808)"
Steve Block [Thu, 24 Feb 2011 12:54:25 +0000 (04:54 -0800)]
Merge "Copy constructor for AutoJObject must increase the ref count on the underlying object"
Steve Block [Wed, 23 Feb 2011 13:49:10 +0000 (13:49 +0000)]
Copy constructor for AutoJObject must increase the ref count on the underlying object
This is needed to balance the derefs in any temporary instance's d'tor.
See https://chromereview.prom.corp.google.com/
2013001
Change-Id: I8329f475ecbd1bc8669d0e8d368f1a385d891fb3
Derek Sollenberger [Tue, 22 Feb 2011 19:00:57 +0000 (14:00 -0500)]
Skia Merge (revision 808)
This is a companion CL to the one found in /external/skia
Change-Id: I469b8845a88f24d972fd57ee5c9cab505a5b83aa
Patrick Scott [Tue, 22 Feb 2011 19:06:12 +0000 (11:06 -0800)]
Merge "Scroll the RenderLayer during touch events."
Cary Clark [Tue, 22 Feb 2011 18:01:47 +0000 (10:01 -0800)]
Merge "fix select text space detection"
Cary Clark [Tue, 22 Feb 2011 17:31:24 +0000 (12:31 -0500)]
fix select text space detection
Some helpers like isSpace() are used in two contexts: while the
picture is parsed, and after the best answer in the picture has
been found. The proper paint must be used in each case.
Associate paints with candidate and best glyphs, so that the
appropriate paint can be used as needed.
bug:
3470564
Change-Id: I28bcc82dbe727ee3ef0716e7657e61a982467a40
Patrick Scott [Tue, 22 Feb 2011 16:20:56 +0000 (11:20 -0500)]
Scroll the RenderLayer during touch events.
Keep track of the owning layer for each LayerAndroid. No longer need to unadjust
the node bounds since the node will be scrolled into view.
Bug:
3442108
Change-Id: I7c9604d347af326ccfb86e6d3f2b95b7ce1b97c1
Kristian Monsen [Mon, 21 Feb 2011 15:22:52 +0000 (07:22 -0800)]
Merge "Fix for bug
3424689 Corrupt (truncated) downloads for apk from browser"
Steve Block [Mon, 21 Feb 2011 13:10:33 +0000 (05:10 -0800)]
Merge "Fix a leaked JNI LocaRef in WebViewCore::jsPrompt()"
Steve Block [Mon, 21 Feb 2011 11:12:10 +0000 (11:12 +0000)]
Fix a leaked JNI LocaRef in WebViewCore::jsPrompt()
Bug:
3470906
Change-Id: I8fb9ddcf4ff3dee35c2225a18a86b4015002da96
Nicolas Roard [Sat, 19 Feb 2011 15:39:19 +0000 (07:39 -0800)]
am
d3b2919d: (-s ours) am
634f229e: Do not merge: Cherry-pick Fix ANR in the browser from master
* commit '
d3b2919d307e4b0ea98b7b5c67991c853c99b562':
Do not merge: Cherry-pick Fix ANR in the browser from master
Nicolas Roard [Sat, 19 Feb 2011 15:35:45 +0000 (07:35 -0800)]
am
634f229e: Do not merge: Cherry-pick Fix ANR in the browser from master
* commit '
634f229e4f0075b052e110ba0a89554c46f6c510':
Do not merge: Cherry-pick Fix ANR in the browser from master
Nicolas Roard [Sat, 19 Feb 2011 09:22:16 +0000 (01:22 -0800)]
Do not merge: Cherry-pick Fix ANR in the browser from master
Sometimes we were not releasing textures as they were busy; they could
still be deleted when swapping the layers trees, and as they were
also still present in the LayerTexture Hashmap this was causing an ANR
(at best -- the texture was already deallocated, the LayerTexture dtor
was then trying to release() them...)
bug:
3398660
Change-Id: I3b08af94a6a7041d45d373ebaec0ec4ba59dac82
Nicolas Roard [Sat, 19 Feb 2011 09:20:06 +0000 (01:20 -0800)]
am
d3f53495: (-s ours) am
928bc355: Merge "Do not merge: Cherry-pick 2nd part of GL_BLEND CL from master" into honeycomb
* commit '
d3f534955e1e80e5a9192c84db518f8f0eff5b00':
Do not merge: Cherry-pick 2nd part of GL_BLEND CL from master
Nicolas Roard [Sat, 19 Feb 2011 09:16:48 +0000 (01:16 -0800)]
am
928bc355: Merge "Do not merge: Cherry-pick 2nd part of GL_BLEND CL from master" into honeycomb
* commit '
928bc35505a3fa29f1598842d49124c06703f972':
Do not merge: Cherry-pick 2nd part of GL_BLEND CL from master
Nicolas Roard [Sat, 19 Feb 2011 09:13:56 +0000 (01:13 -0800)]
Merge "Do not merge: Cherry-pick 2nd part of GL_BLEND CL from master" into honeycomb
Nicolas Roard [Sat, 19 Feb 2011 05:09:18 +0000 (21:09 -0800)]
am
ffef3d3a: (-s ours) am
6f7a546a: Do not merge: Cherry-pick Flash performance CL from Master
* commit '
ffef3d3a1baae44ba44696024ddc12f92337d518':
Do not merge: Cherry-pick Flash performance CL from Master
Derek Sollenberger [Sat, 19 Feb 2011 05:09:16 +0000 (21:09 -0800)]
am
b1c4dba8: (-s ours) am
b1f94474: Do Not Merge: Cherry-pick GL_BLEND CL from master
* commit '
b1c4dba88c350dca089fc918e923d77052df750c':
Do Not Merge: Cherry-pick GL_BLEND CL from master
Nicolas Roard [Sat, 19 Feb 2011 05:09:14 +0000 (21:09 -0800)]
am
6e36bd5d: (-s ours) am
dc8ac1db: DO NOT MERGE Fix GL_BLEND problem. glBindAttribLocation() will only work after a shader has been linked; This was working previously due to the way the nvidia driver work when using blending (it generates another shader).
* commit '
6e36bd5dadbf7af2b01b92ebb38e23a22b9718d1':
DO NOT MERGE Fix GL_BLEND problem. glBindAttribLocation() will only work after a shader has been linked; This was working previously due to the way the nvidia driver work when using blending (it generates another shader).
Nicolas Roard [Sat, 19 Feb 2011 05:09:11 +0000 (21:09 -0800)]
am
dd17d165: (-s ours) am
bdbce22a: DO NOT MERGE Clip layers if the texture size is too large for the GPU
* commit '
dd17d165c8aefdc123104908cef4cb5f742f2296':
DO NOT MERGE Clip layers if the texture size is too large for the GPU
Nicolas Roard [Sat, 19 Feb 2011 05:09:07 +0000 (21:09 -0800)]
am
6c92fe34: am
227fffee: Fix potential bug in setMaxTextureCount and limit the number of textures to 154
* commit '
6c92fe3486d362d49c3d2a0399611797d35173d1':
Fix potential bug in setMaxTextureCount and limit the number of textures to 154
Shimeng (Simon) Wang [Sat, 19 Feb 2011 05:09:03 +0000 (21:09 -0800)]
am
e69a5d5d: (-s ours) am
704eaca6: Set correct max texture allocations.
* commit '
e69a5d5da77aef92f09588057db3ec1f5be5c71b':
Set correct max texture allocations.
Nicolas Roard [Sat, 19 Feb 2011 03:58:00 +0000 (19:58 -0800)]
am
6f7a546a: Do not merge: Cherry-pick Flash performance CL from Master
* commit '
6f7a546a8676227fcd810149bce26090ab1e49eb':
Do not merge: Cherry-pick Flash performance CL from Master
Nicolas Roard [Fri, 18 Feb 2011 23:54:32 +0000 (15:54 -0800)]
Do not merge: Cherry-pick Flash performance CL from Master
Bug:
3471589
Fix some of the performance issues when using the flash plugin embedded.
What happened was that MediaLayer was always telling us to repaint the
screen; we could repaint the same frame a dozen of times unecessarily.
This in itself was wasteful but should have been ok, but the stream
of commands plus the compositing caused the GPU commands used by flash
to stall until they were executed...
The compounded impact drasticaly lowered the performances.
This is not a full fix as those conditions (us repainting the screen
a lot) happens anyway when zooming or scrolling; but this should improve
performances in the general case.
Change-Id: I79a75c759fd5968cb822616eba0caa8c77e75835
Derek Sollenberger [Sat, 19 Feb 2011 03:48:34 +0000 (19:48 -0800)]
am
b1f94474: Do Not Merge: Cherry-pick GL_BLEND CL from master
* commit '
b1f94474f57c247527261dbb10de63b7a4285d60':
Do Not Merge: Cherry-pick GL_BLEND CL from master
Nicolas Roard [Sat, 19 Feb 2011 03:48:30 +0000 (19:48 -0800)]
am
dc8ac1db: DO NOT MERGE Fix GL_BLEND problem. glBindAttribLocation() will only work after a shader has been linked; This was working previously due to the way the nvidia driver work when using blending (it generates another shader).
* commit '
dc8ac1db6f8a14dbac0eb104ecfc18292f6e9a03':
DO NOT MERGE Fix GL_BLEND problem. glBindAttribLocation() will only work after a shader has been linked; This was working previously due to the way the nvidia driver work when using blending (it generates another shader).
Nicolas Roard [Sat, 19 Feb 2011 03:48:25 +0000 (19:48 -0800)]
am
bdbce22a: DO NOT MERGE Clip layers if the texture size is too large for the GPU
* commit '
bdbce22a3ad0e8abea6b824e7ec9fd5ae3ea9637':
DO NOT MERGE Clip layers if the texture size is too large for the GPU
Nicolas Roard [Sat, 19 Feb 2011 03:48:20 +0000 (19:48 -0800)]
am
227fffee: Fix potential bug in setMaxTextureCount and limit the number of textures to 154
* commit '
227fffee9ee9b07919a140866fda1ee7445d3f5a':
Fix potential bug in setMaxTextureCount and limit the number of textures to 154
Shimeng (Simon) Wang [Sat, 19 Feb 2011 03:48:16 +0000 (19:48 -0800)]
am
704eaca6: Set correct max texture allocations.
* commit '
704eaca6f77288587c438291bc15a967515eaa8e':
Set correct max texture allocations.
Derek Sollenberger [Fri, 18 Feb 2011 19:04:07 +0000 (14:04 -0500)]
Do not merge: Cherry-pick 2nd part of GL_BLEND CL from master
Ensure non-ready tiles are painted with the page background color.
bug:
3465059
bug:
3471680
Change-Id: Id0cc94f181882d1308471c9bd898fe005101d6f8
Derek Sollenberger [Mon, 14 Feb 2011 16:26:30 +0000 (11:26 -0500)]
Do Not Merge: Cherry-pick GL_BLEND CL from master
Selectively enable and disable GL_BLEND for better performance.
Bug:
3471680
Change-Id: I065b07f72a097418b81e5922574579574e7ca91a
Nicolas Roard [Fri, 11 Feb 2011 19:38:49 +0000 (11:38 -0800)]
DO NOT MERGE Fix GL_BLEND problem.
glBindAttribLocation() will only work after a shader has been linked;
This was working previously due to the way the nvidia driver work
when using blending (it generates another shader).
Change-Id: Ie5d28ba451e7b37fd9c6a76c14fc3e4e525d9bf0
Nicolas Roard [Fri, 11 Feb 2011 18:00:52 +0000 (10:00 -0800)]
DO NOT MERGE Clip layers if the texture size is too large for the GPU
bug:
3436987
Change-Id: Ibf42685ee07c6d0dd8a1a2b9c8b57da7f2a422db
Nicolas Roard [Sat, 19 Feb 2011 00:35:20 +0000 (16:35 -0800)]
Fix potential bug in setMaxTextureCount and limit
the number of textures to 154
Change-Id: I39409e938b6b4bff23d6f1f0dc44648d63443f59
Nicolas Roard [Sat, 19 Feb 2011 00:35:20 +0000 (16:35 -0800)]
Fix potential bug in setMaxTextureCount and limit
the number of textures to 154
Change-Id: Ie77935015d744fcfda9bfffd751df6c7f3cd10bb
Shimeng (Simon) Wang [Tue, 15 Feb 2011 22:30:49 +0000 (14:30 -0800)]
Set correct max texture allocations.
This fixes core dump or dead lock in GL code when rotating screen.
Change-Id: Ia7bfa71125f1d24158084b6192d805b5b08b60f8
Nicolas Roard [Sat, 19 Feb 2011 00:02:07 +0000 (16:02 -0800)]
Merge "Fix some of the performance issues when using the flash plugin embedded."
Nicolas Roard [Fri, 18 Feb 2011 23:54:32 +0000 (15:54 -0800)]
Fix some of the performance issues when using the flash plugin embedded.
What happened was that MediaLayer was always telling us to repaint the
screen; we could repaint the same frame a dozen of times unecessarily.
This in itself was wasteful but should have been ok, but the stream
of commands plus the compositing caused the GPU commands used by flash
to stall until they were executed...
The compounded impact drasticaly lowered the performances.
This is not a full fix as those conditions (us repainting the screen
a lot) happens anyway when zooming or scrolling; but this should improve
performances in the general case.
Change-Id: I3391c2c5a5198a73863ace53bc10f623cdc00921
Derek Sollenberger [Fri, 18 Feb 2011 20:28:21 +0000 (12:28 -0800)]
Merge "Ensure non-ready tiles are painted with the page background color."
Derek Sollenberger [Fri, 18 Feb 2011 19:04:07 +0000 (14:04 -0500)]
Ensure non-ready tiles are painted with the page background color.
bug:
3465059
Change-Id: Ia6e5ec7e9f68d641f5bf0df34c04c5c9958a4046
Kristian Monsen [Fri, 18 Feb 2011 15:58:41 +0000 (15:58 +0000)]
Part of fix for bug
3424424 Browser hanging
Always calling complete, and adding an assert if Wait finishes
without incrementing the count.
Change-Id: Iceec36c2b82f225ee88cd67eeb18d5c5064c6bfb
Cary Clark [Fri, 18 Feb 2011 16:14:49 +0000 (08:14 -0800)]
Merge "use measure instead of bounder to find spaces"
Cary Clark [Fri, 18 Feb 2011 15:55:25 +0000 (10:55 -0500)]
use measure instead of bounder to find spaces
Pictures do not include white space characters in their
text spans, so they must be deduced. The old method
used drawing through the canvas, compared against
drawing the text directly.
The new method measure and rounds the text to match
how webkit adjusts characters' widths in WidthIterator.
bug:348629
Change-Id: I39fecd274fe0b195fbad13a6083656daa62ed852
Steve Block [Fri, 18 Feb 2011 11:18:37 +0000 (11:18 +0000)]
Fix memory allocation bug in convertV8ObjectToNPVariant() for strings
This is a cherry-pick of WebKit change 78994
See http://trac.webkit.org/changeset/78994
Change-Id: I1994bbbe89490e68025f9bbaa0606cf9766f2ca3
Teng-Hui Zhu [Fri, 18 Feb 2011 01:03:29 +0000 (17:03 -0800)]
Merge "Fixed element in iframe should have a layer"
Teng-Hui Zhu [Thu, 17 Feb 2011 23:40:08 +0000 (15:40 -0800)]
Fixed element in iframe should have a layer
bug:
3466931
Change-Id: I3ce5aaef47b6135b191749b40869913f6420ce7f
Steve Block [Thu, 17 Feb 2011 17:06:53 +0000 (09:06 -0800)]
Merge "Merge WebKit at Chromium 9.0.597.106: Update WebKit merge revision"
Steve Block [Thu, 17 Feb 2011 17:06:39 +0000 (09:06 -0800)]
Merge "Merge WebKit at Chromium 9.0.597.106: Initial merge by Git"
Huahui Wu [Thu, 17 Feb 2011 04:58:42 +0000 (20:58 -0800)]
b/
3452201 Set certificate only when it's from the main page.
Change-Id: Ia81109b5afe993bac2ee4a3b2da232c19f750439
Teng-Hui Zhu [Wed, 16 Feb 2011 19:26:46 +0000 (11:26 -0800)]
Adding a debug setting to enable visual indicator for GL
[This is the webkit part]
The idea is to turn on the visual indicator without building the code.
The implementation included:
1. Setup the UI on browser side to check whether or not this is enabled.
2. Transfer the info from browser setting to web setting.
3. Send this info down from WebView to webkit.
4. In the webkit, we save this info in TilesManager.
5. At texture generation time, we query this info to decide whether or not
add the visual indicator on the texture.
One design decision we made is we don't want to restart the browser for
debugging purpose. This is better preserving the browser current activity,
the only pitfall is that the visual indicator is NOT updated on different
textures simultaneously.
The dependent webView change is #change,97055
bug:
3458721
Change-Id: I0ce51d7ac7de66ccdb95c9bf1f0edc7ec688622f
Derek Sollenberger [Wed, 16 Feb 2011 20:48:46 +0000 (12:48 -0800)]
Merge "Selectively enable and disable GL_BLEND for better performance."
Derek Sollenberger [Mon, 14 Feb 2011 16:26:30 +0000 (11:26 -0500)]
Selectively enable and disable GL_BLEND for better performance.
Change-Id: I473e2bff91c2e0b1ec2436f0666506023a7c0945
Patrick Scott [Wed, 16 Feb 2011 20:00:23 +0000 (12:00 -0800)]
Merge "Clip nodes to the visible rect."
Patrick Scott [Wed, 16 Feb 2011 18:35:37 +0000 (13:35 -0500)]
Clip nodes to the visible rect.
Do not build scrollable layers if the layer does not draw. No need to ask for a
sync since the rebuild happens during a sync.
Bug:
3429982
Change-Id: I7740cd4de4e1658e760da7391de708cbdf05cd44
Cary Clark [Wed, 16 Feb 2011 15:57:03 +0000 (07:57 -0800)]
Merge "match UI text selection background color"
Cary Clark [Wed, 16 Feb 2011 15:56:22 +0000 (07:56 -0800)]
Merge "apply essential security patches"
Cary Clark [Wed, 16 Feb 2011 14:45:27 +0000 (09:45 -0500)]
match UI text selection background color
bug:
3451973
Change-Id: Ie334c54ebb6b5b1083616739f9a90b8e8f3b6693
Kristian Monsen [Wed, 16 Feb 2011 14:40:18 +0000 (14:40 +0000)]
Fix for bug
3424689 Corrupt (truncated) downloads for apk from browser
Sending the url of the response instead of the request, so using
the new url if there is a redirect.
Change-Id: Ic314d315a6e9b5be837c302cdcfd740cc185466d
Steve Block [Wed, 16 Feb 2011 13:13:01 +0000 (13:13 +0000)]
Merge WebKit at Chromium 9.0.597.106: Update WebKit merge revision
Note that we are tracking the Chromium 9.0.597 release branch,
which is WebKit r72805 + stability cherry picks.
This corresponds to r78455 on the 597 release branch.
Change-Id: I79be576710a57ed6d72c7c4dd37408c61c159768
Steve Block [Wed, 16 Feb 2011 13:06:58 +0000 (13:06 +0000)]
Merge WebKit at Chromium 9.0.597.106: Initial merge by Git
Note that we are tracking the Chromium 9.0.597 release branch,
which is WebKit r72805 + stability cherry picks.
This corresponds to r78455 on the 597 release branch.
Change-Id: I72375d9b61a767449086f0c9dc4105b2a6b62ddc
Nicolas Roard [Wed, 16 Feb 2011 01:57:48 +0000 (17:57 -0800)]
Merge "Improve layers repaint speed. Computing the layers texture size is heavy; we delay this computation if one was done recently."
Nicolas Roard [Wed, 16 Feb 2011 01:48:19 +0000 (17:48 -0800)]
Improve layers repaint speed.
Computing the layers texture size is heavy; we delay this
computation if one was done recently.
bug:
3375416
Change-Id: I0ce79cdfcaa15cd1521370a1d5b8df5a5e5ea2c5
Shimeng (Simon) Wang [Tue, 15 Feb 2011 22:46:14 +0000 (14:46 -0800)]
Merge "Set correct max texture allocations."
Shimeng (Simon) Wang [Tue, 15 Feb 2011 22:30:49 +0000 (14:30 -0800)]
Set correct max texture allocations.
This fixes core dump or dead lock in GL code when rotating screen.
Change-Id: I5da2e5a41dfdb18f98f126d2a2fac999210ce7c7
Cary Clark [Tue, 15 Feb 2011 20:55:15 +0000 (12:55 -0800)]
Merge "fix buttons in GL"
Cary Clark [Tue, 15 Feb 2011 19:57:24 +0000 (14:57 -0500)]
fix buttons in GL
Buttons failed to draw when selected with the keyboard, and
sometimes when selected by touch, for three reasons:
1) GL short-circuits if the last extra rectangle and the
curent extra rectangle match. When buttons change color,
the rectangles do match, but the picture still needs to
be redrawn.
2) Buttons do not have cursor rings, so the button's bounds
was not passed to GL.
3) The timeout logic to switch from a focused state to a normal
state was incorrect.
This adds a flag to GLWebViewState::setExtra that can skip the
equal rectangle case. Note that the equal rectangle test hides
a more serious bug where the setExtra is called repeatedly -- so
it would be worthwhile to examine this further.
The logic that determines if the cursor is a button was split
out so that the cursor ring can be associated with a button even
if the cursor is hidden.
bug:
3364248
Change-Id: I75944cb045486c128743aa13922d20758ccf783f
Patrick Scott [Tue, 15 Feb 2011 16:21:28 +0000 (11:21 -0500)]
A couple of fixes for scrolling layers.
Update the layer structure during compositing sync. Make sure to turn off
scrolling if it changes.
Since I replace the content layer with a different structure, append the
children to the foreground layer so they move with the content. Remove the
clipping layer in the compositor for scrollable layers.
For children of scrollable layers, add back in the content scroll offset so that
layers are positioned relative to 0,0.
Bug:
3416512
Change-Id: I74de90aac2dcf67bd969d0b85f440343123ecab9
Cary Clark [Mon, 14 Feb 2011 21:10:15 +0000 (16:10 -0500)]
apply essential security patches
bug:
3341457
bug:
3341449
Change-Id: I7ba58e63a6bcf992b76eb2f5557c26ff30e63ef6
Cary Clark [Mon, 14 Feb 2011 20:10:25 +0000 (12:10 -0800)]
Merge "don't combine single image bitmaps"
Cary Clark [Mon, 14 Feb 2011 19:26:57 +0000 (14:26 -0500)]
don't combine single image bitmaps
Anchors and other links often contain multiple bitmaps, which must
be stitched together when analyzing the picture data to determine
if the anchor is fully visible, partially occluded, or fully hidden.
For instance, nine-patches in WebKit appear as adjacent bitmaps.
If a website design has two or more bitmaps adjacent to each other,
these may be mistakenly joined. By tracking whether the original
node contained exactly one img element, combining bitmaps based
on their location is avoided unncessarily.
bug:
3373743
Change-Id: I5bcbdaec3097a2b839ee2444e512a0def6a016d0
Svetoslav Ganov [Mon, 14 Feb 2011 18:17:23 +0000 (10:17 -0800)]
Merge "Generate well-formed makrup for accessibility events."
Svetoslav Ganov [Fri, 11 Feb 2011 02:27:54 +0000 (18:27 -0800)]
Generate well-formed makrup for accessibility events.
bug:
3444147
1. Removed the code that incorrectly strips CSS spans appended
by WebKit. This CSS information is now propagated to the
client applications such as TalkBack.
2. Renamed getImplicitAnchorOrOffsetNode to getImplicitBoundaryNode.
Change-Id: I6618be4d208fafd223ba09fb90c393046121ca88
Cary Clark [Mon, 14 Feb 2011 16:22:37 +0000 (08:22 -0800)]
Merge "remove unused helper class"
Cary Clark [Mon, 14 Feb 2011 15:50:57 +0000 (10:50 -0500)]
remove unused helper class
Change-Id: Icfb3ccfb8f89486604f1d63ab40e13d28d00cf7c
Kristian Monsen [Mon, 14 Feb 2011 15:36:51 +0000 (07:36 -0800)]
Merge "Adding debug information about requests. #ifdef out by default"
Kristian Monsen [Mon, 14 Feb 2011 15:36:39 +0000 (07:36 -0800)]
Merge "Partly fix for bug
3410366"
Nicolas Roard [Fri, 11 Feb 2011 19:38:49 +0000 (11:38 -0800)]
Fix GL_BLEND problem.
glBindAttribLocation() will only work after a shader has been linked;
This was working previously due to the way the nvidia driver work
when using blending (it generates another shader).
Change-Id: Ia8bf578e214a02c7a7b3ad47600fca87fce73ac8
Nicolas Roard [Fri, 11 Feb 2011 18:00:52 +0000 (10:00 -0800)]
Clip layers if the texture size is too large for the GPU
bug:
3436987
Change-Id: I180b9f54b9db5d190ca2832bea8b9ff210a0619a
Nicolas Roard [Fri, 11 Feb 2011 03:33:59 +0000 (19:33 -0800)]
Merge "Fix ANR in the browser Sometimes we were not releasing textures as they were busy; they could still be deleted when swapping the layers trees, and as they were also still present in the LayerTexture Hashmap this was causing an ANR (at best -- the texture was already deallocated, the LayerTexture dtor was then trying to release() them...)"
Nicolas Roard [Fri, 11 Feb 2011 01:05:41 +0000 (17:05 -0800)]
Fix ANR in the browser
Sometimes we were not releasing textures as they were busy; they could
still be deleted when swapping the layers trees, and as they were
also still present in the LayerTexture Hashmap this was causing an ANR
(at best -- the texture was already deallocated, the LayerTexture dtor
was then trying to release() them...)
bug:
3398660
Change-Id: I2747dd4630a2a7dc18eae4c5fde36fd14c461747
Huahui Wu [Thu, 10 Feb 2011 23:00:15 +0000 (15:00 -0800)]
Merge "b/
3416864 Pass the certificate info from native to JAVA."
Huahui Wu [Thu, 10 Feb 2011 22:39:12 +0000 (14:39 -0800)]
b/
3416864 Pass the certificate info from native to JAVA.
Change-Id: I797ab2d6ce6a07561c710d8b3262e548770eb477
Shimeng (Simon) Wang [Thu, 10 Feb 2011 22:18:41 +0000 (14:18 -0800)]
Merge "Only update history position if scrolling is done by user."
Steve Block [Thu, 10 Feb 2011 21:59:04 +0000 (13:59 -0800)]
Merge "Do not accept cookies for file scheme URLs by default"
Kristian Monsen [Mon, 6 Dec 2010 13:43:13 +0000 (13:43 +0000)]
Do not accept cookies for file scheme URLs by default
This change also adds a getter and setter to CookieMananager to allow
this policy to be configured.
Bug:
2944150
Change-Id: I48668b0d0f2b81581b98933a9399719626da6033
Shimeng (Simon) Wang [Thu, 10 Feb 2011 21:38:41 +0000 (13:38 -0800)]
Only update history position if scrolling is done by user.
issue:
3433562
Change-Id: I520f50cc6657f0a244860e01d38e3f71607fe665
Cary Clark [Thu, 10 Feb 2011 17:36:20 +0000 (12:36 -0500)]
remove obsolete nav cache dump statement
The fix for the bug below to remove part of CachedLayer wasn't
reflected in the nav cache dump.
bug:
3398340
Change-Id: I77723680bed88ecf2f9d11cd56fada111a4a785b
Teng-Hui Zhu [Thu, 10 Feb 2011 17:11:10 +0000 (09:11 -0800)]
Merge changes I0c422ecd,Ia06decf9
* changes:
Improve the fixed element position calculation
Revert "A temp WAR for fixed element position update"
Teng-Hui Zhu [Wed, 9 Feb 2011 02:33:22 +0000 (18:33 -0800)]
Improve the fixed element position calculation
In the previous WAR, we get the screen size from UI thread at every draw call.
That is awkward.
Now with Java side fix, we can get the screen size from setSize call now.
However, forcing a full re-layout seems like an expensive operation.
What we need to make the webkit to know the update can be done by just
updating the fixed position elements only.
So I have done these:
1. Take away the WAR by forcing the update at the convertToLayerCoord.
2. Get rid of the UI thread stuff used to tell WebViewCore the screen size.
And get rid of the redundant variables, too.
3. update the fixed element at the setSize time.
BTW, I also rename the PlatformBridge functions name.
bug:
3397602
Change-Id: I0c422ecdb570de89aecb6e568d5067acf18ecfc6
Patrick Scott [Wed, 9 Feb 2011 23:44:28 +0000 (15:44 -0800)]
Merge "Consolidate scrolling methods."
Patrick Scott [Fri, 4 Feb 2011 11:44:22 +0000 (06:44 -0500)]
Consolidate scrolling methods.
Call through to a single java function to scroll the page. Always use scrollTo
so that there isn't a problem with using scrollBy on an old or out-of-sync
scroll position.
Bug:
3187015
Change-Id: I4a75256687e53ed18c1f73970e9032f476b81859
Svetoslav Ganov [Sun, 6 Feb 2011 21:38:51 +0000 (13:38 -0800)]
WevView accessibility no JavaScript - clean up, null poiner error.
bug:
3421690
1. Added explicit checks for error/null pointer at places
where such check were missing.
2. Fixed skipping of som einput cotrols while traversing backwards.
3. Factored out common code.
4. Input controls were not correctly selected after scrolling
the view port.
5. Imroved node visibility checking.
Change-Id: Idfec7e6a17ac8531f6b073ad6d858932e0921767
Teng-Hui Zhu [Wed, 9 Feb 2011 01:49:45 +0000 (17:49 -0800)]
Revert "A temp WAR for fixed element position update"
This reverts commit
968791c5f1c465f0633c618fda279246ba213fdf.
Ben Murdoch [Tue, 8 Feb 2011 17:44:29 +0000 (17:44 +0000)]
Guard autofill correctly.
Bug:
3411200
Change-Id: I6e80795413d68ac1b7b9ada4ac44901916d6efbf
Ben Murdoch [Mon, 7 Feb 2011 22:28:35 +0000 (14:28 -0800)]
am
43f07e98: am
a046a139: Merge "Fix crash in RenderLayer due to null pointer" into honeycomb
* commit '
43f07e986c22b72e93c01c715d00755bb30f6e74':
Fix crash in RenderLayer due to null pointer
Ben Murdoch [Mon, 7 Feb 2011 22:26:24 +0000 (14:26 -0800)]
am
a046a139: Merge "Fix crash in RenderLayer due to null pointer" into honeycomb
* commit '
a046a139c49a1c3e9e85101aeca4b68b89c396ca':
Fix crash in RenderLayer due to null pointer
Ben Murdoch [Mon, 7 Feb 2011 22:24:20 +0000 (14:24 -0800)]
Merge "Fix crash in RenderLayer due to null pointer" into honeycomb
Russell Brenner [Mon, 7 Feb 2011 21:52:14 +0000 (13:52 -0800)]
am
b50b0e16: am
4a181ab9: Merge "Fix
3366514 (again): browser emulator crash" into honeycomb
* commit '
b50b0e165c9c308b9a4a1824db6ef52a00acadf5':
Fix
3366514 (again): browser emulator crash
Derek Sollenberger [Mon, 7 Feb 2011 21:50:58 +0000 (13:50 -0800)]
am
ad1398b8: am
79f07c99: Fix to post an inval when plugins request a native video surface.
* commit '
ad1398b85d7cef3bd7c3dc833921fcdb3c7f06f6':
Fix to post an inval when plugins request a native video surface.