I'm unable to run my app with ASAN enabled when targeting a physical iOS device. Simulator targets do work. With Xcode 12 and an iPad mini 4 running iOS 14 beta 1 I get the following error during app launch ==750==ERROR: AddressSanitizer failed to allocate 0xffffffffff9fc000 (-6307840) bytes at address 2db624000 (errno: 22) ==750==ReserveShadowMemoryRange failed while trying to map 0xffffffffff9fc000 bytes. Perhaps you're using ulimit -v With Xcode 11.5 and an iPad Air 2 running OS 12.4.1 the error is ==2177==Unable to find a memory range after restricting VM. ==2177==AddressSanitizer CHECK failed: /BuildRoot/Library/Caches/com.apple.xbs/Sources/clangcompilerrt/clang-1103.0.32.62/compiler-rt/lib/asan/asanmac.cc:92 "((0 && "cannot place shadow after restricting vm")) != (0)" (0x0, 0x0) <empty stack>==2177==AddressSanitizer CHECK failed: /BuildRoot/Library/Caches/com.apple.xbs/Sources/clangcompilerrt/clang-1103.0.32.62/compiler-rt/lib/asan/../sanitizercommon/sanitizermallocmac.inc:143 "((!asaninitisrunning)) != (0)" (0x0, 0x0) warning: could not execute support code to read Objective-C class data in the process. This may reduce the quality of type information available. AddressSanitizer report breakpoint hit. Use 'thread info -s' to get extended information about the report. (lldb) thread info -s thread #1: tid = 0x1076c2, 0x000000011531e984 libclangrt.asaniosdynamic.dylib`__asan::AsanDie() My coworker is able to use ASAN with the same App using iPad Pro 10.5, iPadOS 13.5.1, Xcode 11.5 Are there any configuration changes I need to make to be able to use ASAN on my devices?

I am developing SDK and swizzling viewDidAppear. I have a customer who implements a custom TabBar Navigation where VC's are added to the hierarchy on the first load, and then, he changes the opacity to the currently displayed tab so the next time the user sees the tab - viewDidAppear isn't called, so my code isn't called. I'm attaching a sample project which reproduces that. Is there any way to trigger ViewDidAppear intentionally? If yes, what can be the side effect of doing that? Do I have any other alternative in this case? @main struct DemoCustomTabViewApp: App { @UIApplicationDelegateAdaptor(AppDelegate.self) var appDelegate var body: some Scene { WindowGroup { TabBarRouterView() } } } import UIKit // MARK: - TabBarItem (unchanged) enum TabBarItem: Identifiable, CaseIterable { case home, search, profile var id: Self { self } var title: String { switch self { case .home: return "Home" case .search: return "Search" case .profile: return "Profile" } } var icon: String { switch self { case .home: return "house" case .search: return "magnifyingglass" case .profile: return "person" } } } // MARK: - NavigationControllerView struct NavigationControllerView: UIViewControllerRepresentable { var rootViewController: UIViewController func makeUIViewController(context: Context) -> UINavigationController { let navigationController = UINavigationController(rootViewController: rootViewController) return navigationController } func updateUIViewController(_ uiViewController: UINavigationController, context: Context) {} } // MARK: - TabBarRouterViewModel class TabBarRouterViewModel: ObservableObject { @Published var currentTab: TabBarItem = .home @Published var cachedViews: [TabBarItem: AnyView] = [:] let tabs: [TabBarItem] = TabBarItem.allCases func switchTab(to tab: TabBarItem) { currentTab = tab } func createView(for tab: TabBarItem) -> AnyView { if let cachedView = cachedViews[tab] { return cachedView } let rootViewController: UIViewController switch tab { case .home: rootViewController = UIHostingController(rootView: Text("Home View")) case .search: rootViewController = UIHostingController(rootView: Text("Search View")) case .profile: rootViewController = UIHostingController(rootView: Text("Profile View")) } let navigationView = NavigationControllerView(rootViewController: rootViewController) let anyView = AnyView(navigationView) cachedViews[tab] = anyView return anyView } } // MARK: - CustomTabBarView (unchanged) struct CustomTabBarView: View { let tabs: [TabBarItem] @Binding var selectedTab: TabBarItem let onTap: (TabBarItem) -> Void var body: some View { HStack { ForEach(tabs) { tab in Spacer() VStack { Image(systemName: tab.icon) .font(.system(size: 24)) Text(tab.title) .font(.caption) } .foregroundColor(selectedTab == tab ? .blue : .gray) .onTapGesture { onTap(tab) } Spacer() } } .frame(height: 60) .background(Color.white) .shadow(radius: 2) } } // MARK: - TabBarRouterView struct TabBarRouterView: View { @StateObject private var viewModel = TabBarRouterViewModel() var body: some View { VStack(spacing: .zero) { contentView CustomTabBarView( tabs: viewModel.tabs, selectedTab: $viewModel.currentTab, onTap: viewModel.switchTab ) } .edgesIgnoringSafeArea(.bottom) } private var contentView: some View { ZStack { ForEach(viewModel.tabs) { tab in viewModel.createView(for: tab) .opacity(viewModel.currentTab == tab ? 1.0 : 0.0) } } } }

I'm investigating an odd problem that surfaced with macOS Sequoia related to CoreGraphics (CGColorSpaceCreateWithICCData()). I couldn't find a reliable way to track the lifetime (retain/releases) of a CGColorSpaceRef (any pointers appreciated) so I was hoping to use objc_setAssociatedObject() to attach an instance of my own class that would be deallocated when the "owning" CGColorSpaceRef is deallocated. In this way I could set a breakpoint on my own code, presumably invoked when the CGColorSpaceRef is itself going away. While objc_setAssociatedObject() does seem to work, the results don't seem deterministic. Is it because objc_setAssociatedObject() simply won't work reliably with CF types?

I have existing code in production that links against the mach-o library, and uses the following code: NSData *data = [NSData dataWithBytes: getsectiondata(&_mh_execute_header, "__TEXT", "__somePlist_plist", &len) length:len]; In order for this to compile with Sequoia Beta 4 and Xcode 16 beta 4, I have to replace _mh_execute_header with _mh_dylib_header. If I don't, the compiler raises the following error: ld: Undefined symbols: __mh_execute_header, referenced from: -[MyClass init] in MyClass.o clang: error: linker command failed with exit code 1 (use -v to see invocation) Any idea why the linking behavior might have changed? Should I file a bug? (I realize this is a C issue, not Objective-C - but didn't find a tag for that) Thanks!

Does anyone know if any Apple Open Source repository would contain the source code to NSStringFromSelector? TSAN has pointed out an issue in my code that stems from assuming that the value returned by NSStringFromSelector is immutable, and that the function itself is re-entrant. I’d like to check both those assumptions against the actual sources... assuming they haven't changed in macOS Sequoia to begin with.

I think I found a bug in the Objective-C++ compiler, linker, or runtime. Here’s the scenario: We have a macOS app written in Swift. To control hardware from a 3rd party manufacturer, we use a couple SDKs provided by the manufacturer. The SDKs use dynamically loaded libraries and the interface is defined in C++ headers. To bridge between our Swift code and the C++ APIs we have a private Cocoapod that wraps the 3rd party interface with Objective-C++ classes. The two SDKs each provide an interface for discovering attached devices using a callback class that the programmer provides. By accident we named both callback implementations DiscoveryCallback, but this was not a compiler error because neither class was publicly declared, and each was defined in the .mm file where it was used. However, the problem we’re seeing is this: We want to discover Videohub devices, so we register a new instance of DiscoveryCallback (defined in the same .mm file as this code) with the Videohub SDK. A Videohub device is connected and the SDK calls a method on our callback. Surprise! The callback we registered in step 1 was actually the one intended for Decklink devices, defined in a completely different .mm file. This violates all sorts of assumptions and our app quickly crashes. The funny thing is, the two implementations of DiscoveryCallback have completely different method names. The Videohub SDK is supposed to be calling NewVideohubDevice, yet somehow it successfully calls DeckLinkDeviceArrived on an instance of a class it shouldn’t even know about. So the compiler has checked that our intended DiscoveryCallback matches the protocol that the SDK expects, but at runtime the compiled code instantiates a completely different implementation of DiscoveryCallback and somehow doesn’t immediately fail; we still call a method on it that doesn’t even share a name with the intended target. I imagine at this point the method names are long forgotten and are just pointers in a table. I don’t know if this is a bug in the compiler, the Objective-C++ runtime, or if this is just “working as designed” undefined behavior that I should have avoided by not giving two private classes the same name. I know it’s possible to use a private API simply by redeclaring it in my own code, and this seems related to that, but I feel like the compiler or linker should have warned me that I had two implementations of the same class, or if that is not an error, then the runtime should have instantiated the class that was privately defined in the same source file where it was used. Obviously I can’t share our entire project; I’d like to provide some sample code that replicates the issue, but I don’t have time to do that right now. I’m posting this to see if other developers have had a similar experience.

if we define a property in the following way: @property (atomic) NSString *latestObject; Can we assume that the read write to that value is thread safe? i.e the value will be correct. Or it is better to write our own setter/getter with Locks?

I have created an app for iOS using React Native. I've gotten the app to a point where it works on the iPhone simulator via the Expo App. It even works when creating a build of the app to distribute using TestFlight. Now, I want to add Google Ads into the app. However, I've learned that Google Ads won't work via an Expo build. Instead one has to create a development build to test the ads in the app and see how they look. I've also finally been able to compile a successful development build using eas (expo application services). When I install the app on the simulator and try to open it however, the app crashes almost immediately and I don't know why. I have tried deciphering the symbolicated crash report, but I have not been able to figure out what could be causing the error. It looks like it has something to do with the user interface. Below is the thread where the app crashes: Thread 0 Crashed:: Dispatch queue: com.apple.main-thread 0 libsystem_kernel.dylib 0x11324a14a __pthread_kill + 10 1 libsystem_pthread.dylib 0x1132abebd pthread_kill + 262 2 libsystem_c.dylib 0x7ff80016dd1c abort + 133 3 libc++abi.dylib 0x7ff8002c6d12 abort_message + 241 4 libc++abi.dylib 0x7ff8002b951a demangling_terminate_handler() + 266 5 libobjc.A.dylib 0x7ff800061fba _objc_terminate() + 96 6 libc++abi.dylib 0x7ff8002c616b std::__terminate(void (*)()) + 6 7 libc++abi.dylib 0x7ff8002c6126 std::terminate() + 54 8 libdispatch.dylib 0x7ff8001796ec _dispatch_client_callout + 28 9 libdispatch.dylib 0x7ff80017d1e2 _dispatch_block_invoke_direct + 508 10 FrontBoardServices 0x7ff807a8b3a7 __FBSSERIALQUEUE_IS_CALLING_OUT_TO_A_BLOCK__ + 30 11 FrontBoardServices 0x7ff807a8b281 -[FBSMainRunLoopSerialQueue _targetQueue_performNextIfPossible] + 188 12 FrontBoardServices 0x7ff807a8b3cf -[FBSMainRunLoopSerialQueue _performNextFromRunLoopSource] + 19 13 CoreFoundation 0x7ff800429ff3 __CFRUNLOOP_IS_CALLING_OUT_TO_A_SOURCE0_PERFORM_FUNCTION__ + 17 14 CoreFoundation 0x7ff800429f35 __CFRunLoopDoSource0 + 157 15 CoreFoundation 0x7ff800429732 __CFRunLoopDoSources0 + 215 16 CoreFoundation 0x7ff800423e67 __CFRunLoopRun + 919 17 CoreFoundation 0x7ff8004236ed CFRunLoopRunSpecific + 557 18 GraphicsServices 0x7ff8103ba08f GSEventRunModal + 137 19 UIKitCore 0x7ff805cdf6ee -[UIApplication _run] + 972 20 UIKitCore 0x7ff805ce416e UIApplicationMain + 123 21 LeftOff 0x10f870380 main + 96 22 dyld_sim 0x112d3a3e0 start_sim + 10 23 dyld 0x113d57366 start + 1942 I have been trying to solve this for some days now and could really use some help. I have tried changing versions of dependencies in my app too, with a focus on those that manage the UI. For example, I have changed packages such as react-native-ranimated, react-navigation/stack, react-native-gesture-handler, and react-native-screens to name a few. Nothing has worked so far. Please help.

My application is currently crashing in production in iPhone devices running iOS 15.x. The app will work as normal while running developer mode, enterprise builds and Test Flight builds. Unfortunately, I am unable to reproduce this issue even app is downloaded from the app store. I observed crash is happening at main.m. Can someone help me to reproduce this issue? How to read this crash log file? Please find the crash log details attached. 2022-03-04_14-15-23.6227_+0900-a1899fd803a4ed683466702be71b9d224b4cd5bc.crash

I have a public swift function with the below declaration : public func InternalMain (_ pNumOfArgs : Int32, _ pCmdlineArgs : UnsafeMutablePointer<UnsafeMutablePointer<Int8>?>) -> Int32 {..} I need this function as public because I need to invoke in from a different library. But the above declaration produces an error in my generated swift header i.e '-swift.h' file because 'UnsafeMutablePointer<UnsafeMutablePointer?>' type cannot be represented in the swift generated header. Can someone help how do I get past this. This is a class independent function and I tried using the '@nonobj' to prevent this from getting in the generated header but it still gives an error.

DatePicker always get timeFormat from device setting, but i need to show time in datePicker based on My app setting not device setting. any solution for this problem I tried to set locale "us_POSIX" locale for 12 hour and "us_GB" for 24 hour format this way is work in gregorian calendar , but in japanese calendar not showing year properly like (picker showing " 6 " instead of " Reiwa 6 " )

In swift if using custom dateFormat like "dd-mm-yyyy" and device follows japanese calendar now get dateForm dateFormatter , expected "30-05-Reiwa6". but dateFormatter return string is "30=05-0006"

I used a custom time format in the DateTimePicker, such as 'hh:mm a' for 12-hour format. Although the device's time format is set to 24-hour mode, my sample app displays time in 24-hour format like '23:10' instead of the desired 12-hour format like '11:10 PM'. I've already set the locale for the 12-hour format. Has anyone else encountered this issue, and what could be the solution?

Hi. We have weird trouble related to the multi-threading. Does anyone have ideas how to resolve or avoid? Sample Code // sample.hpp class Sample { public: Sample(); ~Sample(); void Init(); void Dealloc(); void OnEvent(); private: std::mutex sample_mutex_; int counter = 1; } // sample.cpp #include "sample.hpp" Sample::Sample() {} Sample::~Sample() {} void Sample::Init() { std::async([]() { std::this_thread::sleep_for(std::chrono::seconds(2)); this->OnEvent(); }); } void Sample::Dealloc() { std::lock_guard<std::mutex> lock(sample_mutex_); counter = 0; } void Sample::OnEvent { // called from another thread std::lock_guard<std::mutex> lock(sample_mutex_); counter += 1; } // obj_sample.h @interface ObjSample : NSObject - (id _Nonnull)init; @property(nonatomic, readonly) std::shared_ptr<Sample> sample; @end // obj_sample.mm @implementation ObjSample - (id _Nonnull)init{ if (self = [super init]) { _sample = std::make_shared<Sample>(); _sample->Init(); } return self; } - (void)dealloc { sample->Dealloc(); } @end What happens the deadlock happens. according to the debug navigator with the Xcode, we figured out 2 facts below. OnEvent does not end although it looks completed. void Sample::OnEvent { // called from another thread std::lock_guard<std::mutex> lock(sample_mutex_); counter += 1; } // <= Thread 35: stop here Sample::Dealloc is run on the same thread of OnEvent. // debug navigator Thread 35 2 Sample::Dealloc <- this is weird. 3 Sample::OnEvent We guess they causes the deadlock. probability less than 10% Environment MacOS: 14.3.1(Apple M1) Xcode: 15.3 iOS Simulator: 17.0.1

Intermittent Failures Occur During macOS Build Execution on Bamboo. Below is an Extract from the Failure Logs. Note: Please note that within our larger Bamboo build environment, several Xcode projects are constructed. Failures occur intermittently across various Xcode projects each time the Bamboo build is initiated. Therefore, these failures are not specific to any particular Xcode project. It's important to emphasize that the Xcode project mentioned, "myProject.xcodeproj," is used purely as a placeholder or example in this context. which 'xcodebuild' VALID_ARCHS="x86_64 arm64" ARCHS="x86_64" -project “”myProject.xcodeproj"" -alltargets -configuration "Release" -UseModernBuildSystem=NO build 08-Feb-2024 09:53:00 Command line invocation: build 08-Feb-2024 09:53:00 /Applications/Xcode.app/Contents/Developer/usr/bin/xcodebuild "VALID_ARCHS=x86_64 arm64" ARCHS=x86_64 -project myProject.xcodeproj -alltargets -configuration Release -UseModernBuildSystem=NO build 08-Feb-2024 09:53:00 build 08-Feb-2024 09:53:00 User defaults from command line: build 08-Feb-2024 09:53:00 IDEPackageSupportUseBuiltinSCM = YES build 08-Feb-2024 09:53:00 UseModernBuildSystem = NO build 08-Feb-2024 09:53:00 build 08-Feb-2024 09:53:00 Build settings from command line: build 08-Feb-2024 09:53:00 ARCHS = x86_64 build 08-Feb-2024 09:53:00 VALID_ARCHS = x86_64 arm64 build 08-Feb-2024 09:53:00 error 08-Feb-2024 09:53:01 objc[39076]: thread is already initializing this class! error 08-Feb-2024 09:53:01 make[6]: *** [compile] Abort trap: 6

I'm working on an FFI for working with ObjectiveC/Foundation/Metal. AFAIU, as many APIs create and autorelease objects, I need to ensure that an NSAutoreleasePool is active when calling into these APIs. So I've created a wrapper that basically mimics @autoreleasepool by creating and initializing an NSAutoreleasePool, running some code, and then draining the pool. So far so good; I'm using this functionality around most of my entry points into the ObjectiveC world. There's two issues I don't understand though. The first is that NSAutoreleasePool initialization seems to require an active autorelease pool, which seems like a chicken-and-egg problem. Running with OBJC_DEBUG_MISSING_POOLS=YES and breaking on objc_autoreleaseNoPool, I see: objc[30336]: MISSING POOLS: (0x1e2e89c40) Object 0x6000007e0050 of class __NSCFString autoreleased with no pool in place - just leaking - break on objc_autoreleaseNoPool() to debug (lldb) bt objc_autoreleaseNoPool at /usr/lib/libobjc.A.dylib (unknown line) _ZN19AutoreleasePoolPage17autoreleaseNoPageEP11objc_object at /usr/lib/libobjc.A.dylib (unknown line) _ZN19AutoreleasePoolPage4pushEv at /usr/lib/libobjc.A.dylib (unknown line) _CFAutoreleasePoolPush at /System/Library/Frameworks/CoreFoundation.framework/Versions/A/CoreFoundation (unknown line) -[NSAutoreleasePool init] at /System/Library/Frameworks/Foundation.framework/Versions/C/Foundation (unknown line) macro expansion at /Users/tim/Julia/pkg/ObjectiveC/src/syntax.jl:163 [inlined] NSAutoreleasePool at /Users/tim/Julia/pkg/ObjectiveC/src/foundation.jl:427 The second, related problem is that some APIs that are called by Metal from a background thread I don't control: (lldb) bt * thread #11, queue = 'com.Metal.CompletionQueueDispatch', stop reason = breakpoint 1.1 * frame #0: 0x000000018c21b9e8 libobjc.A.dylib`objc_autoreleaseNoPool frame #1: 0x000000018c1eb99c libobjc.A.dylib`AutoreleasePoolPage::autoreleaseNoPage(objc_object*) + 252 frame #2: 0x000000018c21c9ec libobjc.A.dylib`AutoreleasePoolPage::push() + 76 frame #3: 0x00000001aaef9694 IOGPU`-[IOGPUMetalBuffer dealloc] + 104 frame #4: 0x00000001f83554e4 AGXMetalG15X_B0`-[AGXBuffer dealloc] + 44 frame #5: 0x00000001f837eb98 AGXMetalG15X_B0`-[AGXG15XFamilyBuffer dealloc] + 76 frame #6: 0x000000019687c1c4 Metal`MTLResourceListChunkFreeEntries(MTLResourceListChunk*) + 64 frame #7: 0x000000019674e2b0 Metal`-[MTLResourceList releaseAllObjectsAndReset] + 72 frame #8: 0x00000001aaefbd10 IOGPU`IOGPUMetalCommandBufferStorageReset + 36 frame #9: 0x00000001aaefbcac IOGPU`IOGPUMetalCommandBufferStorageDealloc + 76 frame #10: 0x00000001aaefa130 IOGPU`-[IOGPUMetalCommandBuffer didCompleteWithStartTime:endTime:error:] + 240 frame #11: 0x000000019674dce4 Metal`-[_MTLCommandQueue commandBufferDidComplete:startTime:completionTime:error:] + 108 frame #12: 0x00000001aaf03b54 IOGPU`IOGPUNotificationQueueDispatchAvailableCompletionNotifications + 128 frame #13: 0x00000001aaf03c60 IOGPU`__IOGPUNotificationQueueSetDispatchQueue_block_invoke + 64 frame #14: 0x000000018c4049d0 libdispatch.dylib`_dispatch_client_callout4 + 20 frame #15: 0x000000018c420c5c libdispatch.dylib`_dispatch_mach_msg_invoke + 468 frame #16: 0x000000018c40bd28 libdispatch.dylib`_dispatch_lane_serial_drain + 368 frame #17: 0x000000018c421998 libdispatch.dylib`_dispatch_mach_invoke + 444 frame #18: 0x000000018c40bd28 libdispatch.dylib`_dispatch_lane_serial_drain + 368 frame #19: 0x000000018c40ca08 libdispatch.dylib`_dispatch_lane_invoke + 432 frame #20: 0x000000018c40bd28 libdispatch.dylib`_dispatch_lane_serial_drain + 368 frame #21: 0x000000018c40c9d4 libdispatch.dylib`_dispatch_lane_invoke + 380 frame #22: 0x000000018c41761c libdispatch.dylib`_dispatch_root_queue_drain_deferred_wlh + 288 frame #23: 0x000000018c416e90 libdispatch.dylib`_dispatch_workloop_worker_thread + 404 frame #24: 0x000000018c5b2114 libsystem_pthread.dylib`_pthread_wqthread + 288 (lldb) c Process 26902 resuming objc[26902]: MISSING POOLS: (0x1e2e89c40) Object 0x14c8a2400 of class AGXG15SDevice autoreleased with no pool in place - just leaking - break on objc_autoreleaseNoPool() to debug Again, I'm not sure how I'm supposed to run this under an autorelease pool.

I have a class category declared and compiled in a mac os static lib: #import <Foundation/Foundation.h> @interface NSNumber(MyExtension) -(NSString *)CallMe; @end then this staticLib is added to a mac console app in XCode and used: #import <Foundation/Foundation.h> #import "MacStaticLib.h" int main(int argc, const char * argv[]) { @autoreleasepool { NSNumber *one = [NSNumber numberWithInt:12]; NSLog(@"%@",[one CallMe]); } return 0; } Compiles fine but fails in runtime, unrecognized selector 'CallMe' send to instance... what's wrong or what I missed? Thanks!!

Hello. iOS 15 and its version of Swift had an issue with applying class level annotations, such as @available to the members of its class. When we attempt to use our Objective-C library to get the class, if there are things inside the class not available on iOS 15, then the call to objc_getClass crashes with EXC_BAD_ACCESS. This appears to have been fixed on iOS 16 and newer, however, as we still support iOS 15, is there a better workaround than having to annotate every needed member of the class with @available, instead of being able to use the class level annotation?

I refer to Google's ESF project code to handle NOTIFY events, but after my notifyQueue is set to QOS_CLASS_BACKGROUND, the process memory gets larger when a lot of events occur. Is there any way to fix this without affecting performance The code is as follows： notifyQueue = dispatch_queue_create("notify",dispatch_queue_attr_make_with_qos_class(DISPATCH_QUEUE_CONCURRENT_WITH_AUTORELEASE_POOL,QOS_CLASS_BACKGROUND, 0)); if (msg->action_type == ES_ACTION_TYPE_NOTIFY) { es_message_t *copied_msg = [self copy_message:msg]; if (!copied_msg) { return; } dispatch_async(self->_notifyQueue, ^{ @autoreleasepool { [self handle:copied_msg]; [self free_message:copied_msg]; } }); }

i’ve seen evidence that UIViewController has logic to prevent its dealloc/deinit methods from running on a background thread. what seems to occur is that, if the last strong reference to a UIVC is zeroed off the main thread, then the VC is logically marked as ‘deallocating’, but actual invocation of dealloc/deinit is enqueued to the main queue. during the window of time between the beginning and end of this asynchronous deallocation, some strange issues can occur. in particular, if the deallocating VC is a parent view controller, then its children can still access it via their parent property. despite this property being marked as weak, a non-nil parent VC reference will be returned. if a weak reference is attempted to be stored to the parent, you get a runtime crash immediately to the effect of: Cannot form weak reference to instance (0x1234) of class <some-UIVC-sublass>... surprisingly, if you load the reference via the objc runtime's objc_loadWeak method, you'll get nil, but no crash. unsurprisingly, if a strong reference to the deallocating parent is stored and persists past its dealloc invocation, you’ll generally end up with a segmentation violation if the reference is accessed. i imagine the UIVC source is quite complex and there are probably good reasons to try and ensure deallocation only ever occurs on the main thread, but it seems surprising that simply passing view controller variables across threads could lead to exposing unsafe references like this. is this behavior expected? assuming not, i've preemptively filed feedback FB13478946 regarding this issue. attached is some sample code that can reliably reproduce the unexpected behavior. UIKitAsyncDeallocCrashTests.swift