2302cc77b4005936917e0295c07921a257ff038d: Bug 1192082 - Iniialize/deinitialize JNI in nsAppShell; r=snorp
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:40 -0400 - rev 257552
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1192082 - Iniialize/deinitialize JNI in nsAppShell; r=snorp First we need to set the Gecko thread JNIEnv* in nsAndroidStartup, but after that we can initialize and deinitialize the rest of JNI, including AndroidBridge, in GeckoAppShell. This makes nsAppShell control the AndroidBridge lifetime. Over time, parts of the AndroidBridge functionality will be migrated to nsAppShell.
250e5f0da2477da1abb19ebb38a7fbdd11e95427: Bug 1192082 - De-clutter AndroidBridge init/deinit; r=snorp
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:39 -0400 - rev 257551
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1192082 - De-clutter AndroidBridge init/deinit; r=snorp Merge all the init code into AndroidBridge constructor and AndroidBridge::ConstructBridge; merge all the deinit code into AndroidBridge destructor and AndroidBridge::DeconstructBridge. In particular, the SetMainThread call is obsolete and removed.
489fd24eea7bd7832cd52fc11cb3d95665a1e60f: Bug 1192082 - Expose AndroidBridge arguments through GeckoThread; r=snorp
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:39 -0400 - rev 257550
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1192082 - Expose AndroidBridge arguments through GeckoThread; r=snorp AndroidBridge needed some arguments during its initialization. We'll provide those arguments in GeckoThread, which AndroidBridge will access.
361c14b70ac274001c0be41646f8c57f654b5bd1: Bug 1192082 - Get rid of GeckoAppShell.nativeInit; r=snorp
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:39 -0400 - rev 257549
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1192082 - Get rid of GeckoAppShell.nativeInit; r=snorp Instead of letting AndroidBridge be constructed separately, we'll let Gecko construct AndroidBridge.
e04ced85c87fb885f959a5f09abecbc132ae3cfb: Bug 1192082 - Expose GeckoThread states to C++; r=me
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:39 -0400 - rev 257548
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1192082 - Expose GeckoThread states to C++; r=me
78ae271ab59891b66c1ac01b865c5953e59f39e2: Bug 1192079 - Update generated code; r=me
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:39 -0400 - rev 257547
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1192079 - Update generated code; r=me
4ecdae08616854e21e7a30f5c66652c2c61bb2b0: Bug 1192079 - Use unqualified names when possible; r=snorp
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:39 -0400 - rev 257546
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1192079 - Use unqualified names when possible; r=snorp To reduce verbosity of the generated code, this patch makes the code generator use unqualified names when possible, e.g. use State::Ref instead of GeckoThread::State::Ref. To accomplish that, function prototypes now use the C++11 -> syntax for return types.
b06ab4073bad61b9fc714e130b27d69957c43ae0: Bug 1192079 - Support inner classes in generated JNI wrapper; r=snorp
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:39 -0400 - rev 257545
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1192079 - Support inner classes in generated JNI wrapper; r=snorp Currently, when we generate JNI wrapper for an inner class, the resulting C++ class will not actually be a nested class of the enclosing class. As a result, the class can be confusing to use. For example, wrapping Java class GeckoThread.State results in two unrelated C++ classes, GeckoThread and State, and it'd be confusing to use State by itself. This patch adds support for inner classes. We start by scanning only for top-level classes, and when processing each top-level class, we recursively scan for inner classes through JarClassIterator.getInnerClasses() and CodeGenerator.generateClasses(). For each Java inner classes, the resulting C++ class will be a nested class. For example, wrapping GeckoThread.State will produce widget::GeckoThread and widget::GeckoThread::State.
c9126c9cfc34dddf8d3d2c343a06dfe0df5059dd: Bug 1192077 - Convert AndroidBridge JNIEnv calls; r=esawin
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:39 -0400 - rev 257544
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1192077 - Convert AndroidBridge JNIEnv calls; r=esawin
c0e71e22101b50be3ee8256e02c16d187e2d7fff: Bug 1192077 - Move AndroidBridge JNIEnv calls to jni/Utils; r=esawin
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:39 -0400 - rev 257543
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1192077 - Move AndroidBridge JNIEnv calls to jni/Utils; r=esawin Calls like GetJNIForThread should now belong in jni/Utils. Moving the calls also reduce clutter in AndroidBridge.
224df670cbb67121041a6c0a262a15cfea388a36: Bug 1189995 - Move GeckoAppShell.pumpMessageLoop to GeckoThread; r=esawin
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:39 -0400 - rev 257542
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1189995 - Move GeckoAppShell.pumpMessageLoop to GeckoThread; r=esawin This method is used by Gecko to pump the Android message loop, and it's also more suited to GeckoThread than GeckoAppShell.
46f21e3aeed556a1d4119ce43b6694f80c3959bc: Bug 1189995 - Move GeckoAppShell.runGecko to GeckoThread; r=esawin
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:39 -0400 - rev 257541
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1189995 - Move GeckoAppShell.runGecko to GeckoThread; r=esawin GeckoAppShell.runGecko really should be in GeckoThread because GeckoThread already takes care of most of the preparation when running Gecko. This patch merges runGecko into GeckoThread.run, but split the argument-building code into its own method.
d3c457344c0eb1a41ccedf51e30e265c2750680b: Bug 1191083 - Merge pending events handling into mechanism for queued native calls: r=snorp
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:38 -0400 - rev 257540
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1191083 - Merge pending events handling into mechanism for queued native calls: r=snorp Right now we have a separate way of handling pending events before Gecko is loaded. We can merge that into the new mechanism for queuing native calls.
9e993120c4140d4226fe74d7a9055164ec366e83: Bug 1191083 - Add mechanism to queue native calls in GeckoThread; r=snorp
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:38 -0400 - rev 257539
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1191083 - Add mechanism to queue native calls in GeckoThread; r=snorp We cannot call native methods until Gecko is loaded. This patch adds a mechanism in GeckoThread so that other code can queue up native method calls and have those calls automatically delivered when Gecko is ready.
c0a3f57f493a59e1ec429f25d295e4fad9a99366: Bug 1191083 - Implement more GeckoThread states; r=snorp
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:38 -0400 - rev 257538
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1191083 - Implement more GeckoThread states; r=snorp Implement the MOZGLUE_READY and JNI_READY states in GeckoThread. Also change GeckoJavaSampler to use the new states instead of a separate flag.
8234563d01fcef9c365a33dac6571e78c87ee924: Bug 1191083 - Rename and expand GeckoThread.LaunchState; r=snorp
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:38 -0400 - rev 257537
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1191083 - Rename and expand GeckoThread.LaunchState; r=snorp GeckoThread.LaunchState now covers the entire GeckoThread lifetime and not just launch, so it's renamed to GeckoThread.State. More utility methods are added to check for the current state.
4684d917028485bae15b0a0d482bbf52c62be00d: Bug 1191161 - Followup to fix wrong annotation classpath; r=nalexander
Jim Chen <nchen@mozilla.com> - Thu, 13 Aug 2015 00:53:38 -0400 - rev 257536
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1191161 - Followup to fix wrong annotation classpath; r=nalexander Because we switched annoations from gecko-mozglue.jar to constants.jar, we should update the corresponding classpaths when processing annotations during code autogeneration. gecko-mozglue.jar is still needed during the javah step because gecko-browser.jar has a dependency on org.mozilla.gecko.mozglue.JNIObject.
a920bf34b5766bf74cb8c995c31f157eec5eb5fe: Bug 828844 - Add a "system-heap-allocated" memory report on Linux. r=glandium.
Nicholas Nethercote <nnethercote@mozilla.com> - Wed, 12 Aug 2015 17:44:00 -0700 - rev 257535
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 828844 - Add a "system-heap-allocated" memory report on Linux. r=glandium. This gives zero when jemalloc is enabled and non-zero when jemalloc is disabled (e.g. I got 83 MiB at start-up, which sounds plausible).
7d6edef7f2424f51be3b1539925c8caaea3effd7: Bug 1135320 - Re-enable voucher.bin generation on Win64 - r=cpearce
Edwin Flores <edwin@mozilla.com> - Thu, 13 Aug 2015 16:08:46 -0700 - rev 257534
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 1135320 - Re-enable voucher.bin generation on Win64 - r=cpearce
948b2e9d1fa282b3acd35a67e6b020b3e68510c7: Bug 992096 - Implement Sub Resource Integrity [2/2]. r=ckerschb
Francois Marier <francois@mozilla.com> - Wed, 12 Aug 2015 20:19:16 -0700 - rev 257533
Push 29221 by ryanvm@gmail.com at Thu, 13 Aug 2015 14:43:44 +0000
Bug 992096 - Implement Sub Resource Integrity [2/2]. r=ckerschb Mochitests
(0) -100000 -30000 -10000 -3000 -1000 -300 -100 -50 -20 +20 +50 +100 +300 +1000 +3000 +10000 +30000 +100000 +300000 tip