| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
|
|
|
|
|
|
|
|
| |
The current JNI generator does not support using an inner class of an
object that is in the same package as the class being processed. This CL
add a new annotation to handle this case.
R=bulach@chromium.org
Review URL: https://codereview.chromium.org/290403002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@271376 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The underlying rule is now "long" by default.
Mechanical change:
git grep -l jni_generator_ptr_type | xargs sed -i /jni_generator_ptr_type/d
Relands crrev.com/270750 (reverted by 270849)
BUG=317523
TBR=darin@chromium.org
Review URL: https://codereview.chromium.org/288223005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@271047 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Suspected to cause Android Tests bot failure: http://build.chromium.org/p/chromium.linux/builders/Android%20Tests/builds/13592
> Android: cleans up jni_generator_ptr_type.
>
> The underlying rule is now "long" by default.
> Mechanical change:
> git grep -l jni_generator_ptr_type | xargs sed -i /jni_generator_ptr_type/d
>
> BUG=317523
> TBR=darin@chromium.org
>
> Review URL: https://codereview.chromium.org/288963002
TBR=bulach@chromium.org
Review URL: https://codereview.chromium.org/292433004
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@270849 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The underlying rule is now "long" by default.
Mechanical change:
git grep -l jni_generator_ptr_type | xargs sed -i /jni_generator_ptr_type/d
BUG=317523
TBR=darin@chromium.org
Review URL: https://codereview.chromium.org/288963002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@270750 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This change adds code to get gamepad data from java objects and
provide these updates to the GamepadProvider which then writes to
GamepadHardwareBuffer which in turn is read by SharedMemoryReader and
later returned by JS to the web page.
Added singleton class GamepadsReader which is responsible for
communication with java class and accessing gamepad data. It adds
methods for communication with singleton java GamepadList class to
get gamepads data.
This also adds new framework classes/methods required for Gamepad API
support. Frameworks changes are responsible for :
- Identifying gamepad devices and their capabilities.
- Managing connected gamepad devices
- Map the connected gamepad devices to standard Gamepad format.
- Keeping gamepads axes/buttons data up-to-date and returning it to
native whenever requested. In android we cannot get gamepad data
directly from sources, so framework is modified to capture
gamepad key and motion events and extract gamepad data from these
events.
* Class GamepadPlatformDataFetcherAndroid : Android specific
implementation of gamepad data fetcher.
* Class ContentViewCore : Manages gamepad list and notifies of new
key/motion event for gamepads.
* Class GamepadList : A new class to manage connected gamepad devices
* Class GamepadDevice : A new class to manage information related to
each gamepad device.
* Class GamepadMappings : This class is responsible for mapping of
known gamepads to the standard gamepad.
This change enables gamepad API by default.
Adds support for parsing float array return type in JNI generator.
NVIDIA Shield and XBox360 gamepads are mapped to the standard gamepad
BUG=330094
TEST=http://www.html5rocks.com/en/tutorials/doodles/gamepad/gamepad-tester/tester.html
R=tsepez@chromium.org
R=darin@chromium.org
Review URL: https://codereview.chromium.org/133943002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@270620 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://code.google.com/p/chromium/issues/detail?id=355595
There is a race between the crazy linker running on the native thread and the system linker running at the same time on the main UI thread, It can show up in several places, but in its simplest incarnation appears as:
--- crazy linker on native thread
mprotect page writable
--- system linker on main thread
mprotect page writable
write it
mprotect page readonly
--- crazy linker on native thread
write it [ <- sigsegv, page readonly (code=2) ]
mprotect page readonly [ not reached ]
The patch moves the crazy linker's r_map update calls so that they execute on the main UI thread. This ensures that the crazy linker sets these pages writable and then readonly once done synchronously with system linker activity.
Companion to:
https://googleplex-android-review.git.corp.google.com/#/c/448556/
BUG=355595
Review URL: https://codereview.chromium.org/228023005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@270084 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
ContentViewCore.java:
Before: real 1m25.705s
After: real 0m0.257s
There was a few overly cautious "non-greedy" terminations, and a few optional
matches outside their respective optional blocks.
BUG=371520
Review URL: https://codereview.chromium.org/277753004
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@269575 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The JNI generator can't distinguish between a class in the same package
and an implicitly imported class, resulting in a runtime error.
Make it a build time error so those classes will have to be explicitly
imported.
BUG=368942
Review URL: https://codereview.chromium.org/267833002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@268891 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Change the JNI generator to introduce a wrapper type
jni_int_wrapper which is used to prevent implicit
conversions which could be an intptr_t to int
conversion.
Change also includes fix (adding explicit case) for
a false positive which existings in the code base.
BUG=354405, 346626
Review URL: https://codereview.chromium.org/267893002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@268741 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
ReleaseByteArrayElements only releases the elements themselves, not the
jbyteArray itself. Put that in a ScopedJavaLocalRef.
BUG=none
Review URL: https://codereview.chromium.org/268213002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@268554 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=290232
Review URL: https://codereview.chromium.org/262803006
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@267684 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
jni_helper.h has been renamed to jni_weak_ref.h and all references
updated.
BUG=365021
TBR=brettw
NOTRY=true
Review URL: https://codereview.chromium.org/260873003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@267468 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Due to a platform bug in Android 4.2.1 (JB MR1),
TextView.setCompoundDrawablesRelative() is a no-op if the view has ever
been measured. As a workaround, we use setCompoundDrawables() directly.
This bug caused the favicons on the bookmarks to appear next to the
wrong bookmark (e.g. the Gmail favicon might appear next to the Twitter
bookmark). This is now fixed.
BUG=368196
Review URL: https://codereview.chromium.org/263473005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@267298 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
Also, remove an unused method.
BUG=368633
BUG=328240
Review URL: https://codereview.chromium.org/255343003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@267245 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently java traces are enabled in Java only when tracing
that is triggered from an Android intent. This flips it
around so we always notify Java to start tracing, no
matter how tracing is enabled.
The case that wasn't working in this case is the 'dev tools
tracing handler' which is how tracing is enabled for
telemetry.
BUG=None
Review URL: https://codereview.chromium.org/237853005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@265763 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=
Review URL: https://codereview.chromium.org/247143006
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@265727 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
BUG=365021
TBR=darin
Review URL: https://codereview.chromium.org/247263009
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@265715 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
All other top-level events are now labelled 'toplevel'.
BUG=None
NOTRY=true
No-try since failing tests are flakes in unrelated code.
Review URL: https://codereview.chromium.org/236203002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@263769 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
host platforms because the Android build system is intended to use
hermetic toolchains. However, using the hermetic toolchain correctly is
also difficult.
To avoid this, use a regex to strip comments instead of invoking cpp.
This regex shoud handle all the Java source cases correctly (it deals
with strings that contain //, /* or */) and should also improve
performance compared to invoking cpp. The JNI generator tests pass with
this change.
Review URL: https://codereview.chromium.org/227303003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@263650 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Due to a bug, TextView.setCompoundDrawablesRelative() is a no-op on JB
MR1 if the text view has ever been measured. This is fixed in JB MR2
onwards, but causes some ugly bugs in Chrome on JB MR1: e.g. the
favicons often don't appear on the bookmarks page.
BUG=361709
NOTRY=true
Review URL: https://codereview.chromium.org/232083005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@263137 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This class can be used to create a local reference frame. A local
reference frame ensures that some amount of local references can
be created and forces all references created in the frame to be
releases when exiting the frame.
BUG=360069
TBR=darin
Review URL: https://codereview.chromium.org/225283008
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@262273 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
This is a post-fix after landing https://codereview.chromium.org/217283005/
and downstream changes in Chrome for Android.
BUG=357655
Review URL: https://codereview.chromium.org/224613003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@261851 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Discovered following issues:
1. after each native library load, the code starts a thread
to delete old libs;
2. render process also tries to delete old libs;
3. Context.getDir(...) creates the directory if not
existing, so existence test is always successful,
and assertion may fail;
The CL fixes these issues. LibraryLoader.ensureInitialized
and LibraryLoader.loadNow have a second boolean parameter,
inBrowserProcess, used by caller to indicate whether the code should try to delete old libs.
BUG=357655
Review URL: https://codereview.chromium.org/217283005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@261382 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
The JNI generator was wrongly identifying Outer.Inner.
Raise an exception with a clearer message.
BUG=
Review URL: https://codereview.chromium.org/214543002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@260103 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
On some Android devices (e.g., Sony Xperia), package manager may
fail to extract native libraries when updating Chrome.
The change tries alleviate the situation by:
1) name libchrome with version number;
2) when failed to load library, try to extract native libraies
and load them.
BUG=311644
Review URL: https://codereview.chromium.org/200753002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@258546 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
As collateral damage, OpenContentUriForRead is now returning File instead of
a plain file descriptor.
BUG=322664
TEST=net_unittests
R=willchan@chromium.org
Review URL: https://codereview.chromium.org/189393002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@258472 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
| |
Review URL: https://codereview.chromium.org/201543002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@257921 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
| |
Review URL: https://codereview.chromium.org/196653011
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@257696 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
It's not a static field.
Review URL: https://codereview.chromium.org/195793020
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@257550 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=347558
Review URL: https://codereview.chromium.org/198093004
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@257094 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=341556
Review URL: https://codereview.chromium.org/197763003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@256755 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=341231
Review URL: https://codereview.chromium.org/189263014
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@256469 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
According to discussions, we decided to disable atrace for Chrome
because of the inaccuracy of timing in the trace due to the overhead
of atrace.
As WebView still needs atrace, move the atrace enabling code into
AwBrowserProcess.
BUG=trace-viewer:454 (https://code.google.com/p/trace-viewer/issues/detail?id=454)
Review URL: https://codereview.chromium.org/183183011
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@256186 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- android.os.MessageQueue.removeMessage is somewhat expensive.
It was calling that method for every single message, and re-inserting a new one
with slightly lower delay everytime.
Fix that by only scheduling a new message if it must happen sooner.
BUG=349059
Review URL: https://codereview.chromium.org/180783017
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@256158 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
| |
Several pieces of code were using wallclock time (System.currentTimeMillis) for
operations that were really dependent on CPU time. Switch them to using the
appropriate SystemClock.uptimeMillis.
Review URL: https://codereview.chromium.org/190773004
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@256077 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=None
Review URL: https://codereview.chromium.org/188783002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@256057 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=349925
Review URL: https://codereview.chromium.org/179793007
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@255924 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=347557
Review URL: https://codereview.chromium.org/182623003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@255902 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is only used in !NDEBUG builds, and making the variable only available
there makes it harder to break -Wunused-const-variable builds with custom
traits. Follow-up to r255159.
BUG=349521,307668
R=fischman@chromium.org
Review URL: https://codereview.chromium.org/187903007
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@255194 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
| |
BUG=349378
R=avayvod@chromium.org
Review URL: https://codereview.chromium.org/180013008
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@255078 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Renaming ChromiumTestShell* classes to ChromeShell*.
This CL depends on https://chromiumcodereview.appspot.com/184493006/
BUG=348137
NOTRY=true
TBR=pfeldman,tonyg,craigdh
Review URL: https://codereview.chromium.org/184993005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@255073 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
NOTRY=true
Review URL: https://codereview.chromium.org/182513002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@254955 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
We used TextUtils#getLayoutDirectionFromLocale(Locale) function
to get the layout direction, but Configuration#getLayoutDirection()
is more correct function for our purpose. The result is the same.
TBR=nyquist
Review URL: https://codereview.chromium.org/186873002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@254858 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
This CL fixes several issues when building files under base/ for
Android x64.
BUG=346626
Review URL: https://codereview.chromium.org/179373003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@254124 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
| |
Follow up on crrev.com/253566, the javap output format has changed.
BUG=
Review URL: https://codereview.chromium.org/176763003
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@253871 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
For our own classes, we use the "Foo.template" pre-processing trick.
For system classes, there's no easy way to expose such constants to C++.
Use the JNI generator to expose "public static final int" fields from java
as a C++ enum.
BUG=
Review URL: https://codereview.chromium.org/180993004
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@253566 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
Move the Java registration outside of RegisterNativesImpl
BUG=346342
NOTRY=true
Review URL: https://codereview.chromium.org/179263002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@253545 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
| |
BUG=341231
Review URL: https://codereview.chromium.org/167243005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@253302 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Refactor ActivityStatus to expose application level visibility over activity visbility.
- Add a listener for the visibility of the Application (are any Activities visible?)
BUG=341231
NOTRY=true
Review URL: https://codereview.chromium.org/159173002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@252679 0039d316-1c4b-4281-b951-d872f2087c98
|
|
|
|
|
|
|
|
|
|
|
| |
- Get rid of WeakContext since it ties to the Activity and is only used in one place.
- Move DeviceMotionAndOrientation onto the Application context
BUG=341231
Review URL: https://codereview.chromium.org/167803002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@252629 0039d316-1c4b-4281-b951-d872f2087c98
|