Summary:
We introduce a few optimizations:
(1) Previous diff: We defer calling any NativeAnimatedModule methods by waiting 1ms before flushing the queue, and debouncing until no flush is requested. Practically, this just means that we'll call NativeAnimatedModule methods N times at once, at the end of a render loop, instead of N times smeared throughout the render loop.
(2) Additionally, instead of calling N methods, we create multi-operation argument buffer and call a single NativeAnimatedModule API, which should essentially throttle NativeAnimatedModule API calls to once-ish per frame. On the native side, this also reduces a lot of overhead associated with scheduling work on the UI thread (we schedule 1 function to run on the UI thread and perform N operations, as opposed to scheduling N functions to run on the UI thread).
TODO:
- implement stubs for iOS
- write gating code so this can be properly tested in VR and in fb4a
Changelog: [Internal]
Reviewed By: genkikondo
Differential Revision: D36338606
fbshipit-source-id: 29ac949b53b874683128a76525586c22def3143b
Summary:
We try to remove the use test_ownership_2020 oncall, which was created to enforce oncall field in test module. This diff stack tries its best to guess who the right oncall is for these "unowned" tests.
If you don't think this belongs to you, please
1. Either re-assign to another team to your best knowledge
2. Or delete the test
Every test should have an owner!
Differential Revision: D36404812
fbshipit-source-id: daf57e860d3700928ae4518897d87247a672eae4
Summary:
Noticed that (LockFree)EventDispatcherImpl had NativeAnimatedNodesManager as a listener 100+ times, as listener registration can happen multiple times from Animated node creation. Since listener management on event dispatcher is thread-safe, we can avoid the thread hop for this.
Changelog: [Internal]
Reviewed By: genkikondo
Differential Revision: D36316102
fbshipit-source-id: f2f417b69885def87f88460d8b1e0b35b66726cb
Summary:
This diff fixed an edge case that event dispatching is failed after pre-allocation of a view and before the view is mounted.
When a cached image is loaded, we will dispatch the event to JS immediately. This is could happen after the view is created during pre-allocation phase, when the event emitter is not instantiated yet. In that case, we will see [an error](https://github.com/facebook/react-native/blob/main/ReactAndroid/src/main/java/com/facebook/react/fabric/FabricUIManager.java#L927) and the event will effectively be ignored.
To fix that we introduced a queue in this diff for those events. They will be dispatched in order when the view is mounted and the event emitter is non-null.
Changelog:
[Android][Fixed] - Fixed an edge case that event dispatching is failed after pre-allocation of a view and before the view is mounted.
Reviewed By: mullender
Differential Revision: D36331914
fbshipit-source-id: cd065b0b36978cb5f0aac793d8d16f07a48f0881
Summary:
Currently on Android, React Native can only accept font resource IDs which it internally converts into Typeface objects. This change allows ReactFontManager to be passed a Typeface at runtime to enable e.g. downloading fonts on demand.
Changelog:
[Android][Added] - Ability to pass a Typeface object to ReactFontManager in addition to a font resource ID
Reviewed By: makovkastar
Differential Revision: D36138123
fbshipit-source-id: b314e8c7fc28174b5caa017076b6eb3d4f6dbaa8
Summary:
This way we can avoid unnecessary ArrayList copies.
Changelog: [Android][Changed] Generalized the return type of ViewManagerOnDemandReactPackage.getViewManagerNames
Reviewed By: nlutsenko
Differential Revision: D36131516
fbshipit-source-id: 6a42c76cadbcce4c3720875d80062e1ee237bc2f
Summary:
Based on Google Play crash logs, crashes due to the assertions in `dispatchViewManagerCommand` is our top source of crashes:
```
com.facebook.react.uimanager.IllegalViewOperationException: Unable to execute operation dispatchViewManagerCommand: blur on view with tag: 17549, since the view does not exists
at com.facebook.react.uimanager.UIImplementation.assertViewExists(UIImplementation.java:830)
at com.facebook.react.uimanager.UIImplementation.dispatchViewManagerCommand(UIImplementation.java:713)
at com.facebook.react.uimanager.UIManagerModule.dispatchCommand(UIManagerModule.java:739)
at com.facebook.react.uimanager.UIManagerModule.dispatchViewManagerCommand(UIManagerModule.java:726)
at com.facebook.react.uimanager.ReanimatedUIManager.dispatchViewManagerCommand(ReanimatedUIManager.java:233)
at java.lang.reflect.Method.invoke(Method.java:-2)
at com.facebook.react.bridge.JavaMethodWrapper.invoke(JavaMethodWrapper.java:372)
at com.facebook.react.bridge.JavaModuleWrapper.invoke(JavaModuleWrapper.java:151)
at com.facebook.react.bridge.queue.NativeRunnable.run(NativeRunnable.java:-2)
```
There are many other places in UIImplementation that instead of throwing a fatal exception, just log a message and continues execution.
Logging a message and discarding the command should be safe for commands like "blur this view". We keep the previous behavior in debug mode, but just log and discard commands in production.
## Changelog
<!-- Help reviewers and the release process by writing your own changelog entry. For an example, see:
https://github.com/facebook/react-native/wiki/Changelog
-->
[Android] [Fixed] - Avoid crash by handling missing views in dispatchViewManagerCommand
Pull Request resolved: https://github.com/facebook/react-native/pull/33795
Test Plan: We had a tough time reproducing specific crashes, but we ran with this binary in the latest release of our app and it has eliminated this class of crashes from Google Play and Bugsnag instrumentation.
Reviewed By: cortinico
Differential Revision: D36285603
Pulled By: JoshuaGross
fbshipit-source-id: db0be74926c4cb83f07686398579236da8e586bf
Summary:
This diff adds a CompositeReactPackageTurboModuleManagerDelegate which can be used to combine 1 or more TurboModuleManagerDelegates into a single one.
This is useful when both the app and one or multiple of its libraries provides a TurboModuleManagerDelegate
Changelog:
[Internal] [Changed] - Add CompositeReactPackageTurboModuleManagerDelegate
Reviewed By: mdvacca
Differential Revision: D36187694
fbshipit-source-id: 88f1d689342345b4cb7e67186791ad5bd1a6f21b
Summary:
This Change fixes a possible null exception as the user uses AppCompat 1.4.0
as a dependency in their App and uses a `TextInput` component.
The exception may happen as `mFabricViewStateManager` is accessed during
the constructor of the superclass, and is not yet initialized.
- For reference, commit #e21f8ec34984551f87a306672160cc88e67e4793 fixes the crash which was already happening.
<!-- Explain the **motivation** for making this change. What existing problem does the pull request solve? -->
Possible null exception as `mFabricViewStateManager` is accessed during
the constructor of the superclass, white it is not yet initialized.
This commit #e21f8ec34984551f87a306672160cc88e67e4793 fixes the crash which was already happening, accessing of mFabricViewStateManager before initialization, and left another one of the call of the same nature.
## Changelog
<!-- Help reviewers and the release process by writing your own changelog entry. For an example, see:
https://github.com/facebook/react-native/wiki/Changelog
-->
Fix: accessing `mFabricViewStateManager` before initialization.
[JavaScript] [Fixed] - accessing `mFabricViewStateManager` before initialization.
Pull Request resolved: https://github.com/facebook/react-native/pull/33769
Test Plan: no external tool was used.
Reviewed By: javache
Differential Revision: D36205441
Pulled By: cortinico
fbshipit-source-id: f3fca69224c7794757514f026a5293b213986186
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/33776
The implementations of these modules is pretty much identical, and we're already shipping the fbjni version of this anyway.
Changelog: [Internal]
Reviewed By: mhorowitz
Differential Revision: D36200330
fbshipit-source-id: 135ee621e1e4c5eb9616ce7f442fc6d4b946f865
Summary:
Cherry picking https://github.com/facebook/react-native/pull/33707 to main branch
This change is extending the changes made by alespergl to reduce the file paths and command lengths of ndk build commands
Essentially we are shortening the length of the source files by using relative paths instead of absolute paths as enumerated by the wildcard expression
This commit is extending the fix by including all the new modules introduced into RN for the new architecture, including the generated modules.
We are also reverting the ndk bump as ndk23 is crashing frequently when building RN with new arch. The reduced file paths lengths ensures the ndk bump is not required for relatively short application paths.
Fix building RN with new architecture on Windows boxes by using relative paths for C++ sources
## Changelog
Fix building RN with new architecture on Windows boxes by using relative paths for C++ sources
[CATEGORY] [TYPE] - Message
Pull Request resolved: https://github.com/facebook/react-native/pull/33784
Test Plan: Verified building on windows box
Reviewed By: javache
Differential Revision: D36241928
Pulled By: cortinico
fbshipit-source-id: 1ce428a271724cbd3b00a24fe03e7d69253f169b
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/33780
Currently this just has a default ThreadPool name, which can be confusing in systraces and crash reports.
Changelog: [Internal]
Reviewed By: ryancat
Differential Revision: D36200090
fbshipit-source-id: 22918993e7c822ed721ccaf79cdcd9d2a972193d
Summary:
Noticed in D36098912 (https://github.com/facebook/react-native/commit/5de0f145eca36d3263cb8a9928add278cdfffa2b) that the definition of NativeAnimatedModule's OperationQueue.isEmpty was incorrect as it didn't account for `mPeekedOperation`, while `drainQueueIntoList` did.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D36109810
fbshipit-source-id: afd7059975ce6e5c65d1975d168a03f0881bd7c1
Summary:
The ReactAndroid cmake build assumes that the build directory is ReactAndroid/build, but this is not always the case if gradle is configured with a different build location. Instead of assuming the build location in cmake, pass it from gradle.
## Changelog
[Android] [Fixed] - Pass react build dir to cmake
Pull Request resolved: https://github.com/facebook/react-native/pull/33736
Test Plan:
Tested in an app with a custom build dir that RN now builds properly after this change.
In build.gradle add:
```groovy
allprojects {
buildDir = "${System.getProperty("user.home")}/.gradle/builds/${rootProject.name}/${project.name}"
}
```
Reviewed By: christophpurrer
Differential Revision: D36128556
Pulled By: cortinico
fbshipit-source-id: cdb8283c93ea7c384ec94cd3e24012220965b6dc
Summary:
This change caused some internal regression in release builds, hence reverting it
Changelog:
[Internal] [Changed] - Back out > Add CompositeReactPackageTurboModuleManagerDelegate
Reviewed By: alsun2001
Differential Revision: D36182014
fbshipit-source-id: e49fc897f3b1a2e9d33fcb7687244be74020c5be
Summary:
Assertion failure would lead to a crash, in real world not being able to start an activity shouldn't crash - usually a navigation won't happen as expected, user could try again.
Changelog:
[Android][Changed] - Don't assert on current activity when call startActivityForResult
Reviewed By: cortinico
Differential Revision: D35746652
fbshipit-source-id: 0b77ca5a69b2f3f3b0b969d84980ed8290ac9b1f
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/33758
This fails to compile in our CircleCI builds
```
/root/react-native/ReactAndroid/src/main/java/com/facebook/react/CompositeReactPackageTurboModuleManagerDelegate.java:52: error: incompatible types: java.util.List<java.lang.Object> cannot be converted to java.util.List<com.facebook.react.ReactPackage>
delegates.add(delegatesBuilder.build(context, Collections.emptyList()));
```
Make the empty collection generic should solve the issue
Changelog:
[Internal] [Changed] - Fix Circle CI for CompositeReactPackageTurboModuleManagerDelegate
Reviewed By: cortinico, dmitryrykun
Differential Revision: D36130573
fbshipit-source-id: 9615311798fbbd7c000838dfbc043f450c2d9e2e
Summary:
This diff adds a CompositeReactPackageTurboModuleManagerDelegate which can be used to combine 1 or more TurboModuleManagerDelegates into a single one.
This is useful when both the app and one or multiple of its libraries provides a TurboModuleManagerDelegate
Changelog:
[Internal] [Changed] - Add CompositeReactPackageTurboModuleManagerDelegate
Reviewed By: cortinico
Differential Revision: D36057036
fbshipit-source-id: bf131fa7315941b3353c0522e4d77d909b82914b
Summary: There's no need to use private feature flag for overflowInset feature. This makes future refactor on feature flags easier.
Reviewed By: javache
Differential Revision: D35866302
fbshipit-source-id: a3fde212ff40cd6d4c68832bb39b34de5a17a8d4
Summary:
Currently, we build Hermes by specifying the Cmake flag `-B ./hermes`. That means
that the output out the build is going to be placed along side the source code.
This is fine, as long as the user doesn't use the `REACT_NATIVE_OVERRIDE_HERMES_DIR`,
which is used inside the Hermes CI. In that case, the source location of Hermes can be
changed, leading to scenarios where `hermesc` can't be found.
Here I'm changing the flag to be `-B $buildDir/hermes`. Therefore the build output
will always be located within the `./ReactAndroid/hermes-engine/build` folder.
This is a more robust solution as the build output will be encapsulated within the
`build/` folder.
Changelog:i
[Internal] [Changed] - Update the hermesc output to be inside the $buildDir
Reviewed By: cipolleschi
Differential Revision: D35964402
fbshipit-source-id: aa7e0775b282897d5a99c1c46265884d19c5f289
Summary:
The `notifyNativeGestureEnded` API is added to notify user gesture ended, so that any optimization we had during handling the gesture can be restored.
It's possible that when the gesture finishes, the RootView is already unmounted from the native side. This might happen when user starts a gesture that caused leave of the RN screen, or close the app.
Changelog:
[Android][Internal] - Avoid NPE for gesture notifier
Reviewed By: javache
Differential Revision: D35902523
fbshipit-source-id: 9bb5819a53dd053290031eebaae1b8f0318ae534
Summary:
This diff updates the CMake command used for configuring the Hermes build
from `cmake` from $PATH to the `cmake` bundled with the Android SDK.
If not found, fallsback to the previous behavior.
This relaxes the requirement of having to ask our users to install CMake
in their CLIs.
Changelog:
[Internal] [Changed] - Use Android SDK version of CMake rather than an external one
Reviewed By: neildhar
Differential Revision: D35931306
fbshipit-source-id: 8d6c554e5e9040e3bd4fed5f72fbdb0eb61d745a
Summary:
Refactor of TouchTargetHelper.findTargetPathAndCoordinatesForTouch to avoid unnecessary lookup of views during the dispatching of Hover Events
changelog: [internal] internal
Reviewed By: lunaleaps, mdvacca
Differential Revision: D32296003
fbshipit-source-id: 93834c37331ad5d75645a5665a1c8c3d965765fb
Summary:
WHen porting TurboModules or adding new TurboModules, ReactModuleSpecProcessor may fail during buck build, and when the failure is caused by typeElement.getAnnotation, no useful information gets collected, making it difficult to debug.
So here I am adding a try & catch so we can get useful debugging info.
Changelog: [Internal]
Reviewed By: RSNara
Differential Revision: D35767207
fbshipit-source-id: 7e1f9dfbfd31339ab37af19c51d85085e100955a
Summary:
This diff allows to customize the Hermes version exposed by the VM with
a version provided by React Native. This will make clearer that a specific
version of Hermes was built at a specific point of the RN release process.
Changelog:
[Internal] [Changed] - Customize the Hermes version using the RN one
Reviewed By: neildhar
Differential Revision: D35820268
fbshipit-source-id: 3e76db921c12a599743fb9312f52edd06d223bad
Summary:
Pull Request resolved: https://github.com/facebook/react-native/pull/33645
With React 18, we now need to allow users on Fabric to opt-in for Concurrent Root.
This commit adds a new method that can be called on the ReactActivityDelegate
that can be used to set the `concurrentRoot` flag on the `initialProps` on the Render.
Changelog:
[Android] [Added] - Expose an API to enable Concurrent Root on Android
Reviewed By: mdvacca
Differential Revision: D35614879
fbshipit-source-id: 2de83e8115d3748c0346cdec6f31b2ab1f899478
Summary:
D34687371 (https://github.com/facebook/react-native/commit/7b5b114d578142d18bf4a7a5279b179a9ac8d958) unfortunately caused a regression with UIAutomator, where it would no longer be able to see any Views that have the ReactAccessibilityDelegate attached to them. This was because the delegate was changed to extend ExploreByTouchHelper which implements its own default AccessibilityNodeProvider, which does nothing in the case of a view without any virtual children.
This diff simply *only* uses the node provider if the view in question has virtual children, otherwise defaulting to the standard behavior from the View class.
Changelog:
[Android][Fixed] - Fixed issue where any node with an AccessibilityDelegate set (which was any node with any accessibility propoerty), was using ExploreByTouchHelper's built in AccessibilityNodeProvider, and not properly populating their AccessibilityNodeInfo's, leading to focus issues and issues with automated test services like UIAutomator.
Reviewed By: kacieb
Differential Revision: D35601320
fbshipit-source-id: 92e009c6e8b4ddcab860e2c91e6bd1a8f95359f0
Summary:
Changelog:
[Android][Fixed] - Fix Extras usage in Android implementation of Linking.sendIntent()
The implementation of sendIntent() has a bug in a way it uses extras map.
From JS layer the API sends and array of map, where each map contains exactly 2 entries: {"key" -> "name_of_extra_to_use", "value" -> value_of_extra_to_use}
However Java parsing was just picking a random pair out of this map and was sending it as extra.
Most frequently the result was "value" -> value_of_extra_to_use in Intent instead of name_of_extra_to_use -> value_of_extra_to_use
This diff fixes the problem
Reviewed By: lunaleaps
Differential Revision: D35516496
fbshipit-source-id: 7da0a1cb3b8aa30463004dbb47008c83d8e95bd1
Summary:
This diff fixes a NullPointerException when disaptching events while the SurfaceMountingManager is being destroyed
changelog: [android] android
Reviewed By: cortinico
Differential Revision: D35559550
fbshipit-source-id: c07f74493384fb1b306338ec1bc8b96f1b6f1f41
Summary: Changelog: [Internal] - Fix pointer event dispatch to also fire enter/leave for ancestors in the hit path. Compared the event order with web on the RNTester W3C pointer example
Reviewed By: appden
Differential Revision: D35403076
fbshipit-source-id: 726e45e49a901b1d97ad3e20f5898701fd1f763b
Summary:
As we introduced KGP inside `ReactAndroid` (cc ShikaSD), we now need to specify a version for it (as users will consume that build on Android New Architecture.
Currently, the version is loaded in the RN `rootProject` which is not available on user's project. I'm cleaning this up.
## Changelog
[Internal] - Do not specify a Kotlin version in the RN rootProject
Pull Request resolved: https://github.com/facebook/react-native/pull/33589
Test Plan:
Tested on a nightly version with
```
npx react-native init RNNightly --version nightly
```
Reviewed By: mdvacca
Differential Revision: D35476777
Pulled By: cortinico
fbshipit-source-id: 5a819ef5fa9a6886d7b7b683f31d59cb05a49f29
Summary:
This is a follow up to my previous diff. It applies the same customization for M1 users to the :ReactAndroid:hermes-engine build.
The two diffs are split so the first one can be cherry-picked into 0.68.1 if needed.
Changelog:
[Android] [Fixed] - Improve support for Android users on M1 machine
Reviewed By: mdvacca
Differential Revision: D35468658
fbshipit-source-id: c07822f6708f872456971378f44257f6c1e967ee
Joshua Gross
Tianyu Li
Pieter De Baets
Xin Chen
Derek Thurn
Harry Yu
Hetan Thakkar
Christoph Purrer
Mike Masam
Anandraj Govindan
Kudo Chien
Janic Duplessis
Lulu Wu
Aniket Mathur
Nicola Corti
Steven Vascellaro
Carmi Grushko
Alex Liang
Moti Zilberman
fabriziobertoglio1987
Luna Wei
Brett Lavalla
Vladyslav Stepanov
David Vacca
Marces Engel
