Summary:
I am fixing an extremely marginal case that probably impacts nothing in production, but in theory, could - see next diff in stack for the assert that is being hit.
TL;DR in marginal, complex cases with a lot of un/flattening, we can generate the following sequence of mutations:
```
UPDATE node V1 -> V2
REMOVE node V1
```
That is incorrect, and what we actually want is:
```
UPDATE node V1 -> V2
REMOVE node V2
```
While this, again, impacts /nothing/ in prod that we know of, it would be good to get this correct so that we can enable stricter asserts (see next diff).
This will also help with debugging LayoutAnimations.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D27697788
fbshipit-source-id: 47f34fe3e8107167b3df4db841d2cc14c58cb31d
Summary:
Changes in following diffs will be gated by this feature flag.
The differ in the new file is copied from the current stable implementation and will not be modified until it's deleted.
Changelog: [Internal]
Reviewed By: sammy-SC, mdvacca
Differential Revision: D27775698
fbshipit-source-id: 03d9518ffd2b1f25712386c56a38bd2b4d839fc2
Summary:
First, I make the breadcrumbs mechanism (landed just this week) more readable - I forgot to add separators between the breadcrumbs.
Second, there is a path that I am 99% sure we never hit. I've had comments to that effect for a ~year, but now I'm adding a falsey assert. If we don't hit it in prod after a few months I'll be more comfortable just deleting the branch entirely (while probably keeping the assert).
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D27697786
fbshipit-source-id: 6d74d1703b2212d069fbed510f2655ec17294458
Summary:
Move this assert so the debug logging executes first; in the case where this assert fires, we'll get a little more information (for Android, where attaching a Cxx debugger is a bit harder vs iOS).
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D27585132
fbshipit-source-id: e3f4cc3d78587744b9e73db685eda1fd6c36ca9d
Summary:
With all known remaining issues (outside of LayoutAnimations, potentially) resolved, enable this assert.
Note for future readers: this is the first time this particular assert has EVER been enabled widely, so if we hit this assert in dev, it's great signal for debugging BUT does not necessarily indicate any wide-spread problems in prod, or with any subsystems. This will simply help us become "more correct" over time.
It is possible, but not extremely likely, that cleaning up things that cause this assert will improve stability/crashes.
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D27697787
fbshipit-source-id: 28fa34eba70548f5001bbd47f41dbe3c6ff3b4c1
Summary:
Allow conversion of LayoutMetrics to DebuggableString.
We also skipped a field in comparison. It probably isn't impactful in terms of production issues, but still wasn't correct.
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D27709451
fbshipit-source-id: 987fc2de0a4562a295d6cbeffdd922cbf056b811
Summary:
Changelog: [internal]
To prevent starvation, this diff implements expiration time as a way to order tasks in priority queue. This stops higher priority tasks from preventing lower priority tasks from running. The same mechanism is implemented in [JavaScript's scheduler](https://github.com/facebook/react/blob/master/packages/scheduler/src/forks/SchedulerNoDOM.js).
Reviewed By: mdvacca
Differential Revision: D27707887
fbshipit-source-id: 3dc734c856a166ef5c17c5045ebd429565ba79f0
Summary:
Changelog: [internal]
Implement RuntimeScheduler::getShouldYield and expose it through JSI.
For now we are only returning `false`. The value is backed by atomic_bool and in the future we will be able to indicate that React should yield to native.
JavaScript implementation:
https://github.com/facebook/react/blob/master/packages/scheduler/src/forks/SchedulerNoDOM.js#L439-L441
Reviewed By: JoshuaGross
Differential Revision: D27648579
fbshipit-source-id: b9313e2efbd9daae8975357df9de803f24a35e89
Summary:
Changelog: [internal]
unstable_scheduleCallback needs to return reference to the created task so it can be cancelled later.
Reviewed By: mdvacca
Differential Revision: D27622779
fbshipit-source-id: 54160015c7f98e123d08c2e13efac4f498d3ba5e
Summary:
Changelog: [internal]
Add minimal implementation of schedule task. More features and proper scheduling will be added later.
Reviewed By: mdvacca
Differential Revision: D27622138
fbshipit-source-id: b2e4623d38e7217290a6a3c59ccc10a1c13e3a4f
Summary:
Changelog: [internal]
`glog` needs to be exported dependency because it is used in public headers.
Reviewed By: mdvacca
Differential Revision: D27617632
fbshipit-source-id: 91aa27b641286002a80a8cd5ef2e6fe6c266b452
Summary:
Changelog: [Internal]
Calls to `surfaceHandler.start()` and `setDisplayMode(DisplayMode::Visible)` in quick succession on different threads can cause a race condition between mount and commit operations.
The mountingCoordinator will try to mount an empty revision without any commits causing it to fail with:
```
TransactionTelemetry.cpp:108: function getCommitStartTime: assertion failed (commitStartTime_ != kTelemetryUndefinedTimePoint)
```
which is called from [Binding.cpp](https://www.internalfb.com/intern/diffusion/FBS/browse/master/xplat/js/react-native-github/ReactAndroid/src/main/java/com/facebook/react/fabric/jni/Binding.cpp?lines=791-791&blame=1).
This change avoids this initial commit by verifying we had at least 1 revision commited before mounting it.
Reviewed By: sammy-SC
Differential Revision: D27430174
fbshipit-source-id: d208d55f02cd218a316d6fea62d0106c2bcb4be8
Summary:
This diff adds a new variable called "DisplayMode" into SurfaceHandler.cpp and FacebookAppRouteHandler.js. The purpose of DisplayMode is for the native pre-render system to notify React that the a surface is either being "pre-rendered" or "rendered"
When the surface is being "pre-rendered" (displayMode == "SUSPENDED"), react will create and commit React Trees, but it will not execute use-effect callbacks
When the surface is being "rendered" (displayMode == "VISIBLE"), react will create and commit React Trees and it will not execute all use-effect callbacks that weren't executed during "pre-rendering"
By default surfaces are going to be rendered with displayMode == "VISIBLE".
This diff should not create any change of behavior for now, this is the infra required to integrate the new offScreen API the react team is working on for pre-rendering system
changelog: [internal] internal
Reviewed By: yungsters
Differential Revision: D27614664
fbshipit-source-id: f1f42fdf174c2ffa74174feb1873f1d5d46e7a95
Summary:
This diff extends startSurface and setSurfaceProps methods with the new parameter called displayMode
changelog: [internal] internal
Reviewed By: yungsters
Differential Revision: D27669847
fbshipit-source-id: c2ddb690ca897e46e00f07b491b91bb2bc8e847d
Summary:
This diff moves DisplayMode out of SurfaceHandler, this is necessary in order to use it from react/uimanager package
changelog: [internal] internal
Reviewed By: ShikaSD
Differential Revision: D27669846
fbshipit-source-id: 274869d8f2907b1b159f51240440acece09a746f
Summary:
This diff updates initial props when DisplayMode changes in Fabric. This method will be called during pre-rendering.
changelog: [internal] internal
Reviewed By: JoshuaGross
Differential Revision: D27607586
fbshipit-source-id: 7625943d57a786d6dfe30dd893e27704f51826d2
Summary:
This diff introduces a new fabric API called setSurfaceProps that will be used to call the new JS api "setSurfaceProps"
changelog: [internal] internal
Reviewed By: sammy-SC
Differential Revision: D27607588
fbshipit-source-id: 36a19f728af244d7e72687d9305b1c568e2b8ec6
Summary:
Introduce a new debugging mechanism for the debugger. Outside of debug mode (you must defined `DEBUG_LOGS_BREADCRUMBS` manually to enable this feature) it will have no cost or binary size.
When the debug mode is enabled, it allows you to trace the call stack to trace what the differ is doing, making logs more useful.
Motivation: tracking down a tricky bug caught by D27585136 which originates in the differ.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D27667885
fbshipit-source-id: ef75a9a1c8890f9bbe3e5b2e8a8ffcde92fb22c2
Summary:
D27316129 made it mandatory for all RCTTurboModules to have a getTurboModule: method. So, there's no need to keep the getTurboModule:initParams method in RCTTurboModuleManagerDelegate. So, to simplify the TurboModule infra, this diff gets rid of that TurboModuleManager delegate method.
Changelog: [iOS][Removed] - Delete RCTTurboModuleManagerDelegate getTurboModule:initParams
Reviewed By: fkgozali
Differential Revision: D27316873
fbshipit-source-id: c0b8449c6088bf08f17ba9a8d1c2cb644e5a242d
Summary:
## Rationale
There are two ways to associate NativeModule ObjC objects with ObjCTurboModule jsi::HostObjects:
1. Via the NativeModule ObjC class's `getTurboModule:` method.
2. Via the TurboModule manager delegate's getTurboModule:initParams: method.
There's no good reason to support both options. So, this diff stack removes 2, and make 1 mandatory for all RCTTurboModules. Not only will this simplify the infra, but it should also help eliminate a class of runtime errors in the TurboModule standalone app migration: you forget to implement the getTurboModule: method.
Changelog: [iOS][Changed] - Make RCTTurboModule getTurboModule: required
Reviewed By: PeteTheHeat
Differential Revision: D27316129
fbshipit-source-id: baccd155b8c191d0f961b316db552bdfdbeb0a97
Summary:
We're making the getTurboModule: method required for all classes that conform to RCTTurboModule.
Many of our ObjC-only and Cxx NativeModules don't implement this method. This diff implements a getTurboModule: method on all those modules that returns nullptr.
**Question:** Why is it fine to make ObjC-only NativeModules return nullptr from their getTurboModule: method?
- Because they're only accessed from ObjC, and should appear as null on the JavaScript side. Longer term, these NativeModules will also go away.
**Question:** Why is it fine to make Cxx NativeModules return nullptr from getTurboModule: method?
- Because after D27316872, the TurboModuleManager checks if the module is a CxxModule first. If it is, we do an early return, and never call the module's getTurboModule: method.
Changelog: [Internal]
Reviewed By: JoshuaGross
Differential Revision: D27316871
fbshipit-source-id: bc693f2927ab3b0de24e6e9e7699390ec0f7d729
Summary:
We're going to make RCTTurboModule getTurboModule: required in D27316129. So, RCTTurboModuleManagerDelegate getTurboModule:initParams is no longer necessary.
## Changes
1. Makes TurboModuleManager stop calling RCTTurboModuleManagerDelegate getTurboModule:initParams
2. Makes getTurboModule: have the lowest priority. So, Cxx NativeModules with a getTurboModule: method won't have their getTurboModule: method executed.
Changelog: [Internal]
Reviewed By: PeteTheHeat
Differential Revision: D27316872
fbshipit-source-id: a024e55b8e3692d7117420007dd3947ecfd5019c
Summary:
Changelog: [internal]
Add config flags for RuntimeScheduler. Even with the flags, React will not be using it. Further changes on React side will be required.
Reviewed By: mdvacca
Differential Revision: D27616916
fbshipit-source-id: 296a040c2b6dd936dd9582e937e6db75e28f31a4
Summary:
Make these logs more readable/useful and add debug print of ShadowView hashes in one place.
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D27585135
fbshipit-source-id: 5f526856d893c32015d8b480522580732fda0cc6
Summary:
Turns out that ShadowViews that have different LayoutMetrics will have the same hash. Fix that.
This helps for debugging LayoutAnimations.
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D27585133
fbshipit-source-id: f0ac50619115150339089276e34fee5ddd0270bc
Summary:
Change the `TransactionTelemetryTest` to use a mock clock or change the tests to only test if at least x amount of time has passed.
Using a mock clock is the only way to make the test deterministic and be able to assert on the sub-results of the captured phases.
Changelog: [Internal] Change to the `TelemetryTest`. Neither changes the runtime behavior nor the API.
Reviewed By: sammy-SC
Differential Revision: D27618448
fbshipit-source-id: 0cbf51b050aabb75341112ea4a43bea0115082f9
Summary:
The `TransactionalTelemetryTest`s are flaky because they use a real clock and assert on how much time has passed (with a threshold, but that's no good either).
Using the real clock in the test is the cause for the test to be flaky because it depends on the assumption that it's the sole process running, never risking to be put in the process-queue of the OS. However, that's not the case and is why the test sporadically fails if the OS decided to schedule other threads/process in the middle of the test.
This diff allows parametrising the `TransactionTelemetry` class with the clock implementation so that tests can use a Mock Clock if desired (separate diff).
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: sammy-SC
Differential Revision: D27618516
fbshipit-source-id: 5a08e115b388398ca2b05b9d5ae0fd281dfe3b04
Summary:
Changelog: [internal[
Introducing RuntimeScheduler. A coordinator of work between native and React.
Reviewed By: mdvacca
Differential Revision: D27616818
fbshipit-source-id: e90d3d9ca8907be99e61f69e62e83cece8155050
Summary:
~Scheduler and ~SurfaceHandler both contain react_native_asserts that trip whenever we reload React Native with Fabric enabled. These asserts trip because of a memory leak in Fabric. These asserts do not run in production. We should fix the memory leak before re-enabling these assertions.
Changelog: [Internal]
Reviewed By: JoshuaGross
Differential Revision: D27481220
fbshipit-source-id: 15c3d46f7efab9ed67a70714efe44b74b0acd385
Summary:
The current test failures don't include the values passed to `EXPECT` which makes it difficult to understand if the test ended earlier or later then expected.
```
Failure: Value of: (commitDuration >= 1000 - threshold) && (commitDuration <= 1000 + threshold)
Actual: false
Expected: true
```
This diff uses the gtest `EXPECT_NEAR` to get exception messages including the delta:
```
Failure: The difference between telemetryDurationToMilliseconds(telemetry.getTextMeasureTime()) and 600 is 153, which exceeds threshold, where
telemetryDurationToMilliseconds(telemetry.getTextMeasureTime()) evaluates to 753,
600 evaluates to 600, and
threshold evaluates to 70.
```
This doesn't change the test's flakiness because of how sleep is implemented.
Changelog: [Internal] Test only change
Reviewed By: sammy-SC
Differential Revision: D27595206
fbshipit-source-id: f31bdd92ecc7271c9491dda18639ea08820f5730
Summary:
Changing back the values of DisplayMode, visible is the default value, it should be 0
changelog: [internal] internal
Reviewed By: ShikaSD
Differential Revision: D27596595
fbshipit-source-id: e5a17e22dc04d380f584bbb816106ab7d3388875
Summary:
Changelog: [internal]
`ShadowNode::Unshared` is preferred over `UnsharedShadowNode`. This diff removes last uses of the alias.
Differential Revision: D27407197
fbshipit-source-id: aa1440f80dcab523d61c186f2d3ce052f314e52c
Summary:
Goals are:
1. Catch errors in parsing during dev-mode in a way that is disruptive/grabs attention, but has enough information.
2. Use react_native_assert for hitting breakpoints (less useful for Android, more for iOS), and add LOGs for when this code is used in Android (more useful for Android, less useful for iOS).
3. Return sane defaults so that prod cases don't crash, and don't return totally garbage data.
I also found a couple cases where parsing was incorrect before; see WritingDirection and TextAlignment. This could impact some layouts and RTL/LTR potentially.
Changelog: [Internal]
Reviewed By: fkgozali
Differential Revision: D27540903
fbshipit-source-id: 99e6949d97e8ef5520d008c1df3cbe408b5a43a4
Summary:
JSI callbacks are only destroyed if the callback is called. If the callback is never called, we're potentially leaking a lot of callbacks.
To mitigate this, we add a wrapper object that is owned by the std::function. Whenever the std::function is destroyed, the wrapper is destroyed and it deallocates the callback as well.
Changelog: [Internal]
Reviewed By: RSNara
Differential Revision: D27436402
fbshipit-source-id: d153640d5d7988c7fadaf2cb332ec00dadd0689a
Summary:
Changelog: [internal]
Default value for `snapToEnd` and `snapToStart` was incorrect in Fabric. This makes no practical difference as default value is never used (JavaScript always sets true or false).
Default value of `snapToEnd` is true.
https://reactnative.dev/docs/scrollview#snaptoend
Default value of `snapToStart` is true.
https://reactnative.dev/docs/scrollview#snaptostart
Reviewed By: PeteTheHeat
Differential Revision: D27505377
fbshipit-source-id: 73e88aa6ad13a851b8c401e9716fc9a650ec0098
Summary:
Changelog: [internal]
Add support for copying text. The implementation is copied over from old React Native renderer.
Reviewed By: JoshuaGross
Differential Revision: D27502474
fbshipit-source-id: 0d0df583f1adc584ca47e987f06bf78c32386fcc
Summary:
Temporarily disable due to build failures.
This is debug-only and I've never hit these asserts, so this should be safe; though it should be reenabled during active development of LA.
Changelog: [Internal]
Reviewed By: sammy-SC
Differential Revision: D27505224
fbshipit-source-id: b3d6d2184d8f226d6f82d3fe0ae6303813c8356a
Summary:
Changelog: [internal]
This method is not designed to be subclassed. Let's make it non-virtual.
Reviewed By: JoshuaGross
Differential Revision: D27464824
fbshipit-source-id: b3186d1c4cf03226005f9a3caa9c002011798faa
Summary:
Changelog: [Internal] enable support for C++ 17.
C++ 17 in React Native targets.
Short and comprehensive list of C++ features:
https://github.com/AnthonyCalandra/modern-cpp-features#c17-language-features
Reviewed By: JoshuaGross
Differential Revision: D27431145
fbshipit-source-id: e8da6fe9d70e9b7343a8caec21cdbeb043478575
Summary:
To ensure that we're eventually deleting all ShadowViews from the StubViewTree and from the mounting layer, make sure that we always queue conflicts if they're DELETE instructions, including for virtual nodes.
I was able to hit this by running some extremely complex animations. It is extremely unlikely that even a single user is hitting this in prod. Therefore, while this is the right change to make, I don't expect (for example) OOMs to go down at all.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D27492935
fbshipit-source-id: bce332feb15229af271cc6e14b8367ebcb36536b
Summary:
This is more insurance against future changes than fixing any existing bugs. In a very small number of cases it looks like this sorting isn't working, for reasons that are not easily reproducible (way less than 1 in 10000 times for me, if that) and not obvious, since we're sorting vectors in a canonical and straightforward way.
Again, this insures that logic won't change in the future.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D27492936
fbshipit-source-id: 7f47e44ac1101915e1a0433c8f0a50a3c6a0c7b3
Joshua Gross
Peter Argany
Samuel Susla
David Vacca
Andrei Shikov
Ramanpreet Nara
Micha Reiser