Summary:
We need to break up the `uimanager` module in order to solve circular dependencies problem (which future diff would have otherwise).
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: JoshuaGross
Differential Revision: D20885646
fbshipit-source-id: b8e3199c0eacc57a5be1481595cf97c84f972293
Summary:
We need to break up the `uimanager` module in order to solve circular dependencies problem (which future diff would have otherwise).
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: JoshuaGross
Differential Revision: D20885645
fbshipit-source-id: 8148bd934879802b076261ed86fa78acf0a07ed3
Summary:
We need to break up the `uimanager` module in order to solve circular dependencies problem (which future diff would have otherwise).
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: JoshuaGross
Differential Revision: D20885163
fbshipit-source-id: 08eb1ba1d408fc0948e8d0da62380786a40973af
Summary:
Yoga uses `std::array` a lot (and `std::array` is not a `std::vector`), so it's useful for printing Yoga values.
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: mdvacca
Differential Revision: D21028570
fbshipit-source-id: c6bf114d5362f085ea201ecdc5b7d59646b33ebd
Summary:
This is a debug-only feature that simply should be an assert. When it triggers in debugger and bubbles to some random exception catch block which makes it impossible to understand was exactly it happens. Making it an assert will stop debugger exactly where it happens.
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: mdvacca
Differential Revision: D21028571
fbshipit-source-id: 3df4ec0da922026bb9df61081cb71113577e06e9
Summary:
* <Image> must be a leaf node; having a proper trait will fail earlier in case of misuse (mounting something inside).
* <View> must have a `View` trait because it's for what that trait is.
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: sammy-SC
Differential Revision: D21028573
fbshipit-source-id: 457716d4661333eb2357f34316f3e495ab4fda24
Summary:
We don't use it as vitrual anymore (setLayoutMetrics is a non-virtual method already), so it does not need to be marker virtual.
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: mdvacca
Differential Revision: D21028572
fbshipit-source-id: 99f86fdd4cf2f5972034d9058d7b82bdc8680187
Summary:
Changelog:
[Internal][Yoga] YGDirection variable was initialized incorrectly by casting -1 to YGDirection. Changing it to default value of direction
Same for YGMeasureMode.
Reviewed By: pasqualeanatriello
Differential Revision: D20869042
fbshipit-source-id: 7bfe490193321baae875ef6fb49a938851950c9f
Summary:
## Summary
Please check out D21035208.
## Changes
- Update `RCTTurboModuleManagerDelegate getTurboModule:instance:jsInvoker:nativeInvoker:perfLogger` to use `RCTTurboModuleManagerDelegate getTurboModule:(const ObjCTurboModule::InitParams)`
- Update all implementations of `RCTTurboModuleManagerDelegate` in accordance with this API change
Changelog:
[iOS][Changed] - Make RCTTurboModuleManagerDelegate create TurboModules via ObjCTurboModuleManager::InitParams
Reviewed By: PeteTheHeat
Differential Revision: D21036272
fbshipit-source-id: c16002c47db26e2ba143fc1080afe9e2fe1e7816
Summary:
## Summary
Please check out D21035208.
## Changes
- `ObjCTurboModule::ObjCTurboModule` changed to accept a bag of arguments `const ObjCTurboModule::InitParams` instead of an argument list.
- TurboModule iOS codegen scripts updated to generated `ObjCTurboModule` subclasses that accept a `const ObjCTurboModule::InitParams` object in their constructor, and forward it to `ObjCTurboModule::ObjCTurboModule`.
- All manually checked in code-generated ObjC++ classes (i.e: RCTNativeSampleTurboModule, RCTTestModule, FBReactNativeSpec) are updated.
## Rationale
This way, the code-gen can remain constant while we add, remove, or modify the arguments passed to ObjCTurboModule.
## Commands run
```
function update-codegen() {
pushd ~/fbsource && js1 build oss-native-modules-specs -p ios && js1 build oss-native-modules-specs -p android && popd;
}
> update-codegen
```
Changelog:
[iOS][Changed] Update ObjCTurboModule to use ObjCTurboModule::InitParams
Reviewed By: PeteTheHeat
Differential Revision: D21036266
fbshipit-source-id: 6584b0838dca082a69e8c14c7ca50c3568b95086
Summary:
I'm trying to use JSI for a React Native custom module. I saw these existing examples where the JSI API is used in the context of a CMakeLists.txt:
https://github.com/terrysahaidak/host-object-test/blob/master/libs/android-jsi/test-jsi/src/main/cpp/CMakeLists.txthttps://github.com/ericlewis/react-native-hostobject-demo/pull/4/files#diff-834320be1b4e4016bac27c05dcd17fb9
In both cases, they manually grab the include directories and jsi.cpp from node_modules/react-native, but I also saw that node_modules/react-native/ReactCommon/jsi/jsi already has a CMakeLists.txt that appears to be intended to provide a jsi static lib, so I tried to pull this into my own CMakeLists.txt like this:
```
add_subdirectory(${RN_DIR}/ReactCommon/jsi/jsi ${CMAKE_CURRENT_BINARY_DIR}/jsi)
...
target_link_libraries(MyLib jsi)
```
Unfortunately when doing this, the consuming project still doesn't see the correct include directories. The change I'm proposing here is to use `target_include_directories` and declare that `..` is a public (to consumers) include directory for the library named `jsi`. With this change, you can do what I showed above to consume the jsi lib by just pulling in the CMakeLists.txt file into your own CMakeLists.txt file.
Changelog: [General][Fixed] Fix jsi cmake include dirs
Pull Request resolved: https://github.com/facebook/hermes/pull/207
Reviewed By: willholen
Differential Revision: D21074270
Pulled By: tmikov
fbshipit-source-id: 7d9ec3255f57a16c0b2be489dffa4540727738a1
Summary:
## Problem
Every time we want to add, remove, or change the data passed to JavaTurboModule's constructor, we have to modify the C++ TurboModule codegen. (The same is true of `ObjCTurboModule`).
**Why was this necessary?**
- `JavaTurboModule` is effectively an abstract class whose constructor is always invoked by code-generated C++ classes. These C++ code-generated class constructors accept an argument list, and manually foward each and every item in that list to `JavaTurboModule::JavaTurboModule`.
## The fix
In this diff, I introduce a struct `JavaTurboModule::InitParams`, to represent a bag of arguments:
```
class JSI_EXPORT JavaTurboModule : public TurboModule {
public:
struct InitParams {
std::string moduleName;
jni::alias_ref<JTurboModule> instance;
std::shared_ptr<CallInvoker> jsInvoker;
std::shared_ptr<CallInvoker> nativeInvoker;
};
```
All `JavaTurboModules` will be created with an instance of this `InitParams` struct, instead of a list of arguments. Our code-generated C++ `jsi::HostObject` sublcasses will simply accept `InitParams` in their constructor, and forward it to `JavaTurboModule`'s constructor. This way, the codegen remains oblivious to what arguments JavaTurboModule requires.
## Okay, but why do we need this change now?
In the future, I plan to modify the constructor for `JavaTurboModule` to accept a performance logger, and a `RuntimeExecutor`. Similar modifications are planned for ObjC. For this reason, to avoid these four codemods, and any potential other codemods that occur because we're making modifications to `JavaTurboModule` or `ObjCTurboModule`, I'm launching this codemod, and the codemods in this stack.
## Misc Fix
- Previously, we were generating the TurboModule name from the Spec filename. This is incorrect because that name represents the spec name. Now, the name will be forwarded from TurboModuleManager in the `JavaTurboModule::InitParams` struct.
## Alternative implementations
I initially considered using `ContextContainer`, but decided against it because:
1. There are no type-safety guarantees.
2. I think it's a bit overkill for this scenario. We just need an opaque bag of data, and for our purposes a simple struct does the job fine.
## Commands run
Reviewed By: fkgozali
Differential Revision: D21035208
fbshipit-source-id: 9542cafea192081bc34d337ab3a7a783083eb06c
Summary:
This pull request adds `YGNode.h` and `YGStyle.h` to the headers exported by Yoga's podspec. They are required by the new Fabric architecture of React Native.
The modulemap and its umbrella header automatically generated by Cocoapods adds all exported headers to the `modulemap`. Having YGNode and YGStyle exported through here has problems, because they are only available in environments that have C++ available, and will produce errors otherwise.
This pull request fences off the contents of those headers in an `#ifdef __cplusplus` block, so they will not cause errors when imported into environments where C++ isn't available.
I had considered adding a custom modulemap to the podspec as part of this pull request, but this way seems the least "invasive", and this way you are able to add and remove exported headers in the podspec without needing to worry about updating the umbrella header at the same time.
Changelog:
[Internal] - Yoga Podspec: Export YGNore and YGStyle headers
Pull Request resolved: https://github.com/facebook/yoga/pull/997
Reviewed By: hramos
Differential Revision: D20966075
Pulled By: mdvacca
fbshipit-source-id: 5f5caa6b639d11e660b968d681da9a4de6c0eb8e
Summary:
Make sure width/height is always passed as Undefined when measure mode is changed to Undefined.
Changelog: [Internal][Yoga] Set width and height as Undefined when we change measure mode to Undefined
Reviewed By: alickbass
Differential Revision: D20029838
fbshipit-source-id: b9931f6ddb13ffd1565889535ade5bbffbe0c304
Summary:
## Problem
For some reason, D20831545 broke the `use_frameworks!` build of RNTester.
## Building RNTester
```
pushd ~/fbsource/xplat/js/react-native-github/RNTester && USE_FRAMEWORKS=1 pod install && open RNTesterPods.xcworkspace && popd;
```
## Error
I built RNTester locally, and the error was this:
```
Undefined symbols for architecture x86_64:
"facebook::jsi::HostObject::set(facebook::jsi::Runtime&, facebook::jsi::PropNameID const&, facebook::jsi::Value const&)", referenced from:
vtable for facebook::react::ObjCTurboModule in RCTImageEditingManager.o
vtable for facebook::react::ObjCTurboModule in RCTImageLoader.o
vtable for facebook::react::ObjCTurboModule in RCTImageStoreManager.o
"facebook::jsi::HostObject::getPropertyNames(facebook::jsi::Runtime&)", referenced from:
vtable for facebook::react::ObjCTurboModule in RCTImageEditingManager.o
vtable for facebook::react::ObjCTurboModule in RCTImageLoader.o
vtable for facebook::react::ObjCTurboModule in RCTImageStoreManager.o
ld: symbol(s) not found for architecture x86_64
```
## Fix
It looked like libraries that depend on "ReactCommon/turbomodule/core" weren't linking to JSI correctly. So, I modified all such Podspecs to also depend on "React-jsi":
```
arc rfr ' s.dependency "ReactCommon/turbomodule/core", version' ' s.dependency "ReactCommon/turbomodule/core", version\n s.dependency "React-jsi", version'
```
This seemed to do the trick. In buck, we'd fix this problem using exported_dependencies. I skimmed through cocoapods, and couldn't find such a configuration option there. So, I guess this will have to do?
Changelog:
[iOS][Fixed] - Fix Cocoapods builds of RNTester
Reviewed By: fkgozali, hramos
Differential Revision: D20905465
fbshipit-source-id: 60218c8274ec165752a428f2a7a9a546607c8fec
Summary:
This diff:
1. Has ObjC NativeModules use the native `CallInvoker` to invoke JS -> native sync/async calls.
2. Integrates the native `CallInvoker` for each ObjC NativeModule with the bridge. This way, the bridge is informed of all JS -> native TurboModule method calls, and dispatches `onBatchComplete` appropriately.
Changelog:
[iOS][Fixed] Integrate ObjC TurboModules async method calls with the bridge
Reviewed By: fkgozali
Differential Revision: D20831545
fbshipit-source-id: da1cbb4ecef4cae85841ca7ef625ab8e380760cd
Summary:
## Problem:
Let `A` be the set of all ObjC NativeModules that neither provide nor reqeust a method queue.
The TurboModule system dispatches all method calls to NativeModules in `A` synchronously to the JS thread. Here is the relevant logic:
**RCTTurboModule.mm:**
Link: https://fburl.com/diffusion/nz9gqje8
```
jsi::Value performMethodInvocation(
// ...
)
{
// ...
dispatch_queue_t methodQueue = NULL;
if ([instance_ conformsToProtocol:protocol(RCTBridgeModule)] &&
[instance_ respondsToSelector:selector(methodQueue)]) {
methodQueue = [instance_ performSelector:selector(methodQueue)];
}
if (methodQueue == NULL || methodQueue == RCTJSThread) {
// This is the default mode of execution: on JS thread.
block();
} else if (methodQueue == dispatch_get_main_queue()) {
```
**Why does this end up happening?**
1. NativeModules that request a method queue have `synthesize methodQueue = _methodQueue` in their `implementation` section. This generates a `methodQueue` getter for the NativeModule, and also creates an ivar to back that getter. The TurboModule system generates a `dispatch_queue_t` and uses ObjC's KVC API to write to the ivar. So in the above logic, for NativeModules that provide a method queue, methodQueue will neither be `NULL` nor `RCTJSThread`, so we don't dispatch synchronously to the JS thread.
2. NativeModules that provide a method queue will return something that is not `NULL` or something that is `RCTJSThread`. If they return `NULL`, the infra will throw an error early. If they return `RCTJSThread`, we'll dispatch synchronously to the JS thread, as we should (...wait. For async NativeModule methods that dispatch to `RCTJSThread`, should we dispatch asynchronously to the JS thread, via jsInvoker? **Edit:** Nope: https://fburl.com/diffusion/ivt9b40s.). In all other cases, we dispatch to appropriately to the respective method queue.
3. For NativeModules that neither provide nor request a method queue (i.e: NativeModules in `A`), they don't implement the `methodQueue` selector. Therefore, we dispatch synchronously to the JS thread.
## The fix (Part 1):
The first step towards fixing this problem is to generate `dispatch_queue_t`s for NativeModules in `A`.
That's what this diff accomplishes.
Changelog:
[iOS][Fixed] - Create method queue for NativeModules that don't provide nor request one.
Reviewed By: fkgozali
Differential Revision: D20821054
fbshipit-source-id: 17a73550ad96766c5c7e719e28e1cc879e36465c
Summary:
`RCTTurboModuleManager` will create a native `CallInvoker` for each ObjC NativeModule. This `CallInvoker` will be used to dispatch calls from JS to native. Before passing the native `CallInvoker` to the `ObjCTurboModule`, it'll first use `RCTCxxBridge decorateNativeCallInvoker` to get a bridge-aware decorated native `CallInvoker`. That way, the bridge remains informed about async TurboModule method calls that took place since the last time it was flushed. This ensures that we don't end up dispatching `onBatchComplete` any less with TurboModules on than we do with TurboModules off.
Changelog: [Internal]
Reviewed By: fkgozali
Differential Revision: D20831546
fbshipit-source-id: b2eb4e0097e0dabf8c4bd8fdc4c850a0858af699
Summary:
We'll be using a native CallInvoker to dispatch sync and async method calls to ObjC NativeModules. This native CallInvoker will hold a reference to the ObjC NativeModule's method queue.
**Why is the native CallInvoker required for ObjC NativeModules?**
In the case where the ObjC NativeModule neither provides nor requests a method queue, we must create a method queue for it. When we go to invoke a method from JS, for these NativeModules specifically, there is no way to access this method queue. A native CallInvoker is a convenient abstraction that holds on to that method queue. For async calls, we'll just call `CallInvoker::invokeAsync`, and for sync calls, we'll just call `CallInvoker::invokeSync`.
**Why do we need sync call support for native `CallInvoker`?**
In ObjC, sync NativeModule method calls block the JS thread, then execute synchronously on the NativeModule's method queue, and then unblock the JS thread. This is what'll be implemented by `CallInvoker::invokeSync`.
Changelog: [Internal]
Reviewed By: fkgozali
Differential Revision: D20829955
fbshipit-source-id: efb9d5408a1ade81069a943c865f232d4d10acfe
Summary:
Now, instead of accepting a `std::function` that schedules work, and returning a `CallInvoker`, `Instance::getDecoratedNativeCallInvoker` will accept a `CallInvoker` that schedules work, and return a decorated `CallInvoker`.
I think this change will help with readability. It also clarifies that the bridge is adding additional behaviour to the native `CallInvoker`.
Changelog: [Internal]
Reviewed By: mdvacca
Differential Revision: D20826885
fbshipit-source-id: a2c5681d10a4544ee3d2a0d1f1cbd386ef06d0e6
Summary:
Might be worthwhile to just kill this method instead, since we're having all NativeModules provide their TurboModule jsi::HostObjects. But I'll leave that decision to a later time.
Changelog: [Internal]
Reviewed By: fkgozali
Differential Revision: D20809201
fbshipit-source-id: ee73d4b5454a76460832a54f9b864841e5b2b9c0
Summary:
This is necessary to integrate TurboModule async method dispatch with the bridge's `onBatchComplete` event. See D20717931 for more details.
This diff is similar to D20480971.
**Note:** This stack doesn't really make any functional changes, since the native CallInvoker is `nullptr` right now.
Changelog:
[Internal]
Reviewed By: fkgozali
Differential Revision: D20809199
fbshipit-source-id: bf465a3a51bdddb8b56d1e696ca510fdf071f9ec
Summary:
This is the first of three PRs related to enabling multi-bundle support in React Native. More details, motivation and reasoning behind it can be found in RFC [here](https://github.com/react-native-community/discussions-and-proposals/issues/152).
Logic responsible for installing globals was pulled out from `loadApplicationScript` to `initializeRuntime` since it should be ran only once, what was left was renamed to `loadBundle`.
It's based on dratwas work from [here](https://github.com/callstack/react-native/tree/feat/multibundle/split-load-application), but applied to current `master` to avoid rebasing 3-months old branch and issues that come with that.
## Changelog
[Internal] [Changed] - split `loadApplicationScript` into `initializeRuntime` and `loadBundle` to enable multi-bundle support in the future
Pull Request resolved: https://github.com/facebook/react-native/pull/27844
Test Plan: Initialized new RN app with CLI, set RN to build from source and verified the still app builds and runs OK using code from this branch.
Reviewed By: rickhanlonii
Differential Revision: D19888605
Pulled By: ejanzer
fbshipit-source-id: 24ace48ffe8978796591fe7c6cf53a61b127cce6
Summary:
While resolving the flexible items we calculate totalFlexShrinkScaledFactors which uses the flexBasis or initial width of node (Not min-width).
At a later stage during distribution of space we are subtracting value from this which also takes care of min-width.
For example
If node has flexShrink 1 and width 100 and min-width 301 then totalFlexShrinkScaledFactors will become -1*100 = -100
but later we are subtracting -1 * 301 (min-width) = -301 which is ambiguous and causing layout inconsistencies with how web behaves.
Fixed this by only using the flexBasis or width for these calculations.
Changelog:
[Internal][Yoga] Fix layout issue when flexShrink and min-width are used together
Reviewed By: pasqualeanatriello
Differential Revision: D20219419
fbshipit-source-id: 948fbc06ca541d4ad307c88c8a2df65d157778b1
Summary:
If name is passed in as part of RuntimeConfig, that is included
in the description, too.
Changelog: [Internal]
Reviewed By: dulinriley
Differential Revision: D20716320
fbshipit-source-id: f2fba6df32f496090dee787d8b7f55a6a4dd8ed8
Summary:
## Context
For now, assume TurboModules doesn't exist.
**What happens when we call an async NativeModule method?**
Everytime JS calls an async NativeModule method, we don't immediately execute it. The legacy infra pushes the call into some queue managed by `MessageQueue.js`. This queue is "flushed" or "emptied" by the following events:
- **Flushed:** A C++ -> JS call. NativeModule async methods can called with an `onSuccess` and/or `onFail` callback(s). Calling `NativeToJsBridge::invokeCallback` to invoke one of these callbacks is one way for ObjC++/C++/Java to call into JS. Another way is via JSModule method calls, which are initiated by `NativeToJsBridge::callFunction`.
- **Flushed:** When `JSIExecutor::flush` is called. Since TurboModules don't exist, this only happens when we call `JSIExecutor::loadApplicationScript`.
- **Emptied:** When more than 5 ms have passed, and the queue hasn't been flushed/emptied, on the next async NativeModule method call, we add to the queue. Afterwards, we empty it, and invoke all the NativeModule method calls.
**So, what's the difference between flushed and emptied?**
> Note: These are two terms I just made up, but the distinction is important.
If the queue was "flushed", and it contained at least one NativeModule method call, `JsToNativeBridge` dispatches the `onBatchComplete` event. On Android, the UIManager module is the only module that listens to this event. This `onBatchComplete` event doesn't fire if the queue was "emptied".
**Why does any of this matter?**
1. TurboModules exist.
2. We need the TurboModules infra to have `JsToNativeBridge` dispatch `onBatchComplete`, which depends on:
- **Problem 1:** The queue being flushed on calls into JS from Java/C++/ObjC++.
- **Problem 2:** There being queued up NativeModule async method calls when the queue is flushed.
In D14656466, fkgozali fixed Problem 1 by making every C++/Java/Obj -> JS call from TurboModules also execute `JSIExecutor::flush()`. This means that, with TurboModules, we flush the NativeModule async method call queue as often as we do without TurboModules. So far, so good. However, we still have one big problem: As we convert more NativeModules to TurboModules, the average size of the queue of NativeModule method calls will become smaller and smaller, because more NativeModule method calls will be TurboModule method calls. This queue will more often be empty than not. Therefore, we'll end up dispatching the `onBatchComplete` event less often with TurboModules enabled. So, somehow, when we're about to flush the NativeModule method call queue, we need `JsToNativeBridge` to understand that we've executed TurboModule method calls in the batch. These calls would have normally been queued, which would have led the queue size to be non-zero. So if, during a batch, some TurboModule async method calls were executed, `JsToNativeBridge` should dispatch `onBatchComplete`.
**So, what does this diff do?**
1. Make `Instance` responsible for creating the JS `CallInvoker`.
2. Make `NativeToJsBridge` responsible for creating the native `CallInvoker`. `Instance` calls into `NativeToJsBridge` to get the native `CallInvoker`.
3. Hook up `CatalystInstanceImpl`, the Android bridge, with the new JS `CallInvoker`, and the new native `CallInvoker`. This fixes `onBatchComplete` on Android. iOS work is pending.
Changelog:
[Android][Fixed] - Ensure `onBatchComplete` is dispatched correctly with TurboModules
Reviewed By: mdvacca
Differential Revision: D20717931
fbshipit-source-id: bc3ccbd6c135b7f084edbc6ddb4d1e3c0c7e0875
Summary:
Changelog: [Internal]
View with `ShadowColor` was getting flattened and therefore views didn't have shadow property set.
This is fixed by promoting ShadowColor so in case it is set, it forms stacking context.
Reviewed By: shergin
Differential Revision: D20792201
fbshipit-source-id: 1033ac00e32047ffbb14e61b7c26348c578d132d
Summary:
The differ was still producing correct, but not minimal, instruction sets in some cases due to an optimization that was buggy.
This affected cases where 2+ nodes were inserted at the beginning of a list. It would trigger the old behavior where all nodes after the first would be removed, deleted, then reinserted.
See the test case (which was failing and now passed) and P128190998 (the 3->4 transition) for samples.
Changelog: [Internal] Fabric differ
Reviewed By: mdvacca
Differential Revision: D20785729
fbshipit-source-id: 2fea6a816753066abb358ed7bb51003140cd5fc4
Summary:
This diff implements the instruction pointer save/restore trick Tzvetan came up with; allowing us to observe and modify the IP from outside the interpreter loop with negligible overhead.
From Tzvetan's internal post on the subject:
> [Today] the interpreter IP is just a local variable in the interpreter function, so there is no way to get to its value from outside the function. It lives in a register and we don't want to make it a Runtime field since the overhead [of accessing it via memory in the interpeter loop] would kill us.
> However, if you really think about it, it only lives in a register while the interpreter function is running. For the rest of the time, it is spilled by the C++ compiler onto the stack.
So, precisely when we need it, it is actually stored in memory. The only problem is, we don't know where! Admittedly, that is an annoying problem, but it feels like it should be solvable.
> What if, instead of relying on the compiler to spill the IP register, we manually spill it ourselves, to a known location? It works. Example: https://godbolt.org/z/ftSDnp
This diff implements this approach across the whole interpreter loop: whenever we call out of the loop we capture/publish the IP and restore it again immediately after the external call returns. This means we can now see the IP outside the interpret loop and even change it. This is effectively "for free" as the compiler now skips spilling/restoring the IP behind the scenes.
The immediate benefit of this is knowing the current IP allows us to have more accurate stack-traces during execution. In future this may enabled tricks like changing the IP before returning to the interpreter loop, allowing things outside the interpreter to affect program flow without adding logic to the interpreter loop.
Reviewed By: tmikov
Differential Revision: D20151091
fbshipit-source-id: 3814382639800208d8985a32ede31ba8f7ff7c80
Summary:
Changelog: [Internal]
`orderIndex_` was only being assigned for `ViewShadowNode`, not for other `ShadowNodes` that are later represented on the screen.
Reviewed By: shergin
Differential Revision: D20746477
fbshipit-source-id: c04c2cfea14b9141d22bc3d9e9bb4c0c59925754
Summary:
Changelog: [internal]
`ModalHostViewShadowNode` didn't have `RootNodeKit` trait, therefore `getRelativeLayoutMetrics` was including nodes in ancestors that it shouldn't have.
Reviewed By: shergin
Differential Revision: D20735801
fbshipit-source-id: 6b81e3b174c2f82e530abc2bca2da8bebc2270b0
Summary:
Here we implement a bunch of helper methods that allow customizing the behavior of a RuntimeExecutor "on-demand" on the caller side. We will use it in the next diff(s).
Changelog: [Internal] Fabric-specific internal change.
Reviewed By: PeteTheHeat
Differential Revision: D20551411
fbshipit-source-id: 51d3cd02b69753110c0e1155347c6e52eb882c7d
Summary:
Changelog: [Internal]
Current implementation of `measure` doesn't take transform into account..
So if you had a view which has width and height 100 and had `Scale(0.5, 0.5, 1)` (this will shrink view by half). Calling `getRelativeLayoutMetrics` would report its size being `{100, 100}`.
This applies if view's parent has transformation as well, because transformation is applied to all subviews of the view as well.
Reviewed By: mdvacca
Differential Revision: D20621590
fbshipit-source-id: 2cf902a0494291c821ecada56f810c5e6620db5a
Summary:
Changelog: [Internal]
Until now `Rect operator*(Rect const &rect, Transform const &transform)` supported only scaling. Now it supports translation and rotation as well.
Reviewed By: shergin
Differential Revision: D20622876
fbshipit-source-id: 1b65393bd3fd6fd9a8941903e0f2681a10097e4a
Summary:
`LayoutInspectingPolicy` has two flags, `includeTransform` and `includeScrollViewContentOffset`.
`includeScrollViewContentOffset` seems to be redundant for two reasons.
# 1st
From looking at callers, they have always the same value.
I looked at all call sites, and they are either always both set to true or both set to false.
# 2nd
The way we include scroll view content offset, is through transformation, so setting `includeTransform` to true and `includeScrollViewContentOffset` to false will include content offset anyway. In order to make both flags work, we would need to introduce further changes to `getRelativeLayoutMetrics`. But since the flag isn't used anyway, I think it is better to get rid of it for now. If we need it in the future, we could re-introduce it.
Reviewed By: shergin
Differential Revision: D20622256
fbshipit-source-id: fb6156c66b752319ea928239fa723ff90688b0a0
Summary:
Two things:
1) I reimplement Valentin's idea in D19965405, so that TinyMaps can be iterated over, with a couple of bugfixes (calling front() or back() on an empty vector will crash).
2) I now use TinyMap instead of better::map in the "optimized" diffing algorithm.
3) `erase` now actually removes elements from the vector, but only when more than half of elements have been erased.
4) If you repeatedly erase elements at the beginning of the vector, they will no longer be iterated over. This is a specific optimization for our heaviest TinyMap use-cases.
These amount to some small but hopefully somewhat meaningful perf improvements.
Changelog: [Internal] Fabric perf
Reviewed By: shergin
Differential Revision: D20718719
fbshipit-source-id: 91f4b2e2e0f6387ae484e43d5b0095103087baa6
Summary:
This unit test is to verify that the new diffing algorithm generates a "minimal" instruction set, with regards to removes and inserts ("moves").
These unit tests are here to verify the expected behavior in this new algorithm, but these tests may be modified or deleted in the future if we decide we want to change this behavior.
Changelog: [Internal] fabric unit test
Reviewed By: mdvacca
Differential Revision: D20706592
fbshipit-source-id: 5f9991498e0d788ecbf88d938bfe6d3f0f27af40
Summary:
An evolution of D20633188 but more performant.
There are three optimized paths before the slow path.
The first optimized path tries to pair identical nodes from old/new tree, and generate Update mutations, until we hit nodes that are different (indicating either a remove or an insert). This already existed.
The next two optimizations, introduced by Tim in his JS pseudocode, were inspired by ReactJS's diffing algorithm. They work in cases where the rest of the nodes are (1) all removals/deletes or (2) all creates+inserts.
Finally, if those final two optimized paths can't run, it's because there is a mix of delete+remove, create+insert, and "move" operations, mixed at the beginning, middle, and/or end of the list.
This has slightly better average/best-case complexity as the previous implementation.
In particularly pathological cases where all nodes are arbitrarily reordered, or reversed, for instance (ABCDE->EDCBA) the algorithm has the same complexity as the previous algorithm (quadratic).
For now iOS is pinned to the older differ
Changelog: [Internal] Experiment to optimize diffing algorithm in Fabric
Reviewed By: shergin
Differential Revision: D20684094
fbshipit-source-id: d29fba95a0328156c023e1c87804f23770ee1d91
Summary:
`std::this_thread::sleep_for` is not really precise and will attempt to sleep for "at least" that much time, but may sleep much longer depending on what CPUs are doing and scheduling policies.
To get this to pass on my machine, I had to substantially increase the thresholds.
Changelog: [Internal]
Reviewed By: shergin
Differential Revision: D20689571
fbshipit-source-id: f159420d24a95da2b5d95d49ed7882e783291e98
Summary:
State Reconciliation has been running 50/50 for a while and all metrics look stable. This is necessary for providing a good experience so we should ship to everyone unconditionally.
Changelog: [Internal] Fabric diffing reconciliation process improvement
Reviewed By: mdvacca
Differential Revision: D20715694
fbshipit-source-id: 25b2635ecc29b67e2911679c9db66bc84d37dec1
Valentin Shergin
Sidharth Guglani
Ramanpreet Nara
Ryan Tremblay
Rick Hanlon
empyrical
Samuel Susla
maciej simka
Marc Horowitz
Joshua Gross
Jacob Bower
Emilis Baliukonis
