Processes & Concurrency

RSS for tag

Discover how the operating system manages multiple applications and processes simultaneously, ensuring smooth multitasking performance.

Concurrency Documentation

Posts under Processes & Concurrency subtopic

Post

Replies

Boosts

Views

Activity

BGAppRefreshTask Canceled Immediately by dasd in Network Extension
Dear Apple Support Team, My app, io.cylonix.sase, has a BGAppRefreshTask (io.cylonix.sase.ios.refresh) that is canceled by dasd ~9ms after submission from a Network Extension. Please help identify the cause and suggest a solution. App Details: App ID: io.cylonix.sase iOS Version: 17.1.1 (iPhone Xs Max) Network Extension: saseWgNetworkExtension with packet-tunnel-provider entitlement Use Case: VPN app; Network Extension records file receipts in shared group UserDefaults and schedules BGAppRefreshTask to wake the main app. App Usage: High (frequently used) System State: Sufficient resources (not low on battery or memory) Issue: The task is submitted but canceled immediately with priority 10. It has never run, so rate-limiting is not an issue. ` debug 22:09:37.952749-0700 dasd Best binding found for evaluator 0x16d541720: <private> debug 22:09:37.954483-0700 dasd Invoking selector backgroundTaskSchedulerPermittedIdentifiersWithContext:tableID:unitID:unitBytes: on <LSApplicationRecord 0x724844650> default 22:09:37.955563-0700 dasd CANCELED: bgRefresh-io.cylonix.sase.ios.refresh:ABDAFA at priority 10 <private>!
6
0
107
May ’25
Helper app is sandboxed (entitlement + runtime check), but `URLsForDirectory:` returns user home (`/Users//`) instead of container path — why?
Problem summary I have a macOS helper app that is launched from a sandboxed main app. The helper: has com.apple.security.app-sandbox = true and com.apple.security.inherit = true in its entitlements, is signed and embedded inside the main app bundle (placed next to the main executable in Contents/MacOS), reports entitlement_check = 1 (code signature contains sandbox entitlement, implemented via SecStaticCode… check), sandbox_check(getpid(), NULL, 0) returns 1 (runtime sandbox enforcement present), APP_SANDBOX_CONTAINER_ID environment variable is not present (0). Despite that, Cocoa APIs return non-container home paths: NSHomeDirectory() returns /Users/<me>/ (the real home). [[NSFileManager defaultManager] URLsForDirectory:inDomains:] and URLForDirectory:inDomain:appropriateForURL:create:error: return paths rooted at /Users/<me>/ (not under ~/Library/Containers/<app_id>/Data/...) — i.e. they look like non-sandboxed locations. However, one important exception: URLForDirectory:... for NSItemReplacementDirectory (temporary/replacement items) does return a path under the helper's container (example: ~/Library/Containers/<app_id>/Data/tmp/TemporaryItems/NSIRD_<helper_name>_hfc1bZ). This proves the sandbox is active for some FileManager APIs, yet standard directory lookups (Application Support, Documents, Caches, and NSHomeDirectory()) are not being redirected to the container. What I expect The helper (which inherits the sandbox and is clearly sandboxed) should get container-scoped paths from Cocoa’s FileManager APIs (Application Support, Documents, Caches), i.e. paths under the helper’s container: /Users/<me>/Library/Containers/<app_id>/Data/.... What I tried / diagnostics already gathered Entitlements & code signature codesign -d --entitlements :- /path/to/Helper.app/Contents/MacOS/Helper # shows com.apple.security.app-sandbox = true and com.apple.security.inherit = true Runtime checks (Objective-C++ inside helper): extern "C" int sandbox_check(pid_t pid, const char *op, int flags); NSLog(@"entitlement_check = %d", entitlement_check()); // SecStaticCode check NSLog(@"env_variable_check = %d", (getenv("APP_SANDBOX_CONTAINER_ID") != NULL)); NSLog(@"runtime_sandbox_check = %d", sandbox_check(getpid(), nullptr, 0)); NSLog(@"NSHomeDirectory = %s", NSHomeDirectory()); NSArray *urls = [[NSFileManager defaultManager] URLsForDirectory:NSApplicationSupportDirectory inDomains:NSUserDomainMask]; NSLog(@"URLsForDirectory: %@", urls); Observed output: entitlement_check = 1 env_variable_check = 0 runtime_sandbox_check = 1 NSHomeDirectory = /Users/<me> URLsForDirectory: ( "file:///Users/<me>/Library/Application%20Support/..." ) Temporary/replacement directory (evidence sandbox active for some APIs): NSURL *tmpReplacement = [[NSFileManager defaultManager] URLForDirectory:NSItemReplacementDirectory inDomain:NSUserDomainMask appropriateForURL:nil create:YES error:&err]; NSLog(@"NSItemReplacementDirectory: %@", tmpReplacement.path); Observed output (example): /Users/<me>/Library/Containers/<app_id>/Data/tmp/TemporaryItems/NSIRD_<helper_name>_hfc1bZ Other facts Calls to NSHomeDirectory() and URLsForDirectory: are made after main() to avoid "too early" initialization problems. Helper is placed in Contents/MacOS (not Contents/Library/LoginItems). Helper is a non-GUI helper binary launched by the main app (not an XPC service). macOS version: Sequoia 15.6 Questions Why do NSHomeDirectory() and URLsForDirectory: return the real /Users/<me>/... paths in a helper process that is clearly sandboxed (entitlement + runtime enforcement), while NSItemReplacementDirectory returns a container-scoped temporary path? Is this behavior related to how the helper is packaged or launched (e.g., placement in Contents/MacOS vs Contents/Library/LoginItems, or whether it is launched with posix_spawn/fork+exec vs other APIs)? Are there additional entitlements or packaging rules required for a helper that inherits sandbox to have Cocoa directory APIs redirected to the container (for Application Support, Documents, Caches)? *Thanks in advance — I can add any requested logs
6
0
86
2w
Coordination of Video Capture and Audio Engine Start in iOS Development
Question: When implementing simultaneous video capture and audio processing in an iOS app, does the order of starting these components matter, or can they be initiated in any sequence? I have an actor responsible for initiating video capture using the setCaptureMode function. In this actor, I also call startAudioEngine to begin the audio engine and register a resultObserver. While the audio engine starts successfully, I notice that the resultObserver is not invoked when startAudioEngine is called synchronously. However, it works correctly when I wrap the call in a Task. Could you please explain why the synchronous call to startAudioEngine might be blocking the invocation of the resultObserver? What would be the best practice for ensuring both components work effectively together? Additionally, if I were to avoid using Task, what approach would be required? Lastly, is the startAudioEngine effective from the start time of the video capture (00:00)? Platform: Xcode 16, Swift 6, iOS 18 References: Classifying Sounds in an Audio Stream – In my case, the analyzeAudio() method is not invoked. Setting Up a Capture Session – Here, the focus is on video capture. Classifying Sounds in an Audio File Code Snippet: (For further details. setVideoCaptureMode() surfaces the problem.) // ensures all operations happen off of the `@MainActor`. actor CaptureService { ... nonisolated private let resultsObserver1 = ResultsObserver1() ... private func setUpSession() throws { .. } ... setVideoCaptureMode() throws { captureSession.beginConfiguration() defer { captureSession.commitConfiguration() } /* -- Works fine (analyseAudio is printed) Task { self.resultsObserver1.startAudioEngine() } */ self.resultsObserver1.startAudioEngine() // Does not work - analyzeAudio not printed captureSession.sessionPreset = .high try addOutput(movieCapture.output) if isHDRVideoEnabled { setHDRVideoEnabled(true) } updateCaptureCapabilities() }
5
0
891
Oct ’24
What are the reasons for an application to be launched from the background?
Our application has seen a surge in the volume of background launches starting from April and May, and we want to know under what circumstances the application can be launched from the background. First, here's how I determined background launches: we analyze user logs and append UIApplication.appState to each line of log, finding that every log from the start to the end of user sessions has an appState of UIApplicationStateBackground. By checking the "ActivePrewarm" in main() and printing the launch options from application:didFinishLaunchingWithOptions:, we found several scenarios for background launches: launchOptions has a value with the key UIApplicationLaunchOptionsRemoteNotificationKey. launchOptions has no value and there is no "ActivePrewarm." launchOptions has no value but has "ActivePrewarm." I would like to know: Under what circumstances will notifications trigger a background launch (I cannot replicate this locally)? Under what circumstances does an application launch in the background and trigger application:didFinishLaunchingWithOptions: but without any launch options? I hope informations below can provide some insights. Regarding "ActivePrewarm," I've read various questions and answers in the Apple Developer Forums, such as this thread, which states that "ActivePrewarm" does not trigger application:didFinishLaunchingWithOptions: but occurs due to certain behaviors in the application. I would like to know what behaviors may cause this background launch, as there is no information in the launch options, or how I can identify what behaviors triggered it. Specifically, based on that same thread, I've tried to gather more information using runningboardd, and I've currently identified two special cases: When I restart my phone and unlock it after a short period, there is information: <RBSDomainAttribute| domain:"com.apple.dasd" name:"DYLDLaunch" sourceEnvironment:"(null)"> ]> Every day, at intervals of a few hours, there is information: <RBSDomainAttribute| domain:"com.apple.dasd" name:"DYLDLaunch" sourceEnvironment:"(null)"> ]> Then, the following similar information follows: 12:15:56.047625+0800 runningboardd Executing launch request for app<{my_bundle_id}((null))> (DAS Prewarm launch) 12:15:56.050311+0800 runningboardd Creating and launching job for: app<{my_bundle_id}((null))> 12:15:56.050333+0800 runningboardd _mutateContextIfNeeded called for {my_bundle_id} 12:15:56.080560+0800 runningboardd app<{my_bundle_id}((null))>: -[RBPersonaManager personaForIdentity:context:personaUID:personaUniqueString:] required 0.000954 ms (wallclock); resolved to {1000, 39E408CF-2E67-4DB0-BF73-CFC5792285CD} 12:15:56.080632+0800 runningboardd 'app<{my_bundle_id}(39E408CF-2E67-4DB0-BF73-CFC5792285CD)>' Skipping container path lookup because containerization was prevented (<RBSLaunchContext: 0xcd8cc9180>) 12:15:56.080939+0800 runningboardd 'app<{my_bundle_id}(39E408CF-2E67-4DB0-BF73-CFC5792285CD)>' Constructed job description: <dictionary: 0xcd8aa2a00> { count = 19, transaction: 0, voucher = 0x0, contents = *** } 12:15:56.084839+0800 runningboardd [app<{my_bundle_id}((null))>:1649] Memory Limits: active 4096 inactive 4096 <private> 12:15:56.084861+0800 runningboardd [app<{my_bundle_id}((null))>:1649] This process will be managed. 12:15:56.084882+0800 runningboardd Now tracking process: [app<{my_bundle_id}((null))>:1649] 12:15:56.084928+0800 runningboardd Calculated state for app<{my_bundle_id}((null))>: running-active (role: Background) (endowments: (null)) 12:15:56.086762+0800 runningboardd Using default underlying assertion for app: [app<{my_bundle_id}((null))>:1649] 12:15:56.086977+0800 runningboardd Acquiring assertion targeting [app<{my_bundle_id}((null))>:1649] from originator [app<{my_bundle_id}((null))>:1649] with description <RBSAssertionDescriptor| "RB Underlying Assertion" ID:33-33-23101 target:1649 attributes:[ <RBSDomainAttribute| domain:"com.apple.underlying" name:"defaultUnderlyingAppAssertion" sourceEnvironment:"(null)">, <RBSAcquisitionCompletionAttribute| policy:AfterApplication> ]> 12:15:56.087203+0800 runningboardd Assertion 33-33-23101 (target:[app<{my_bundle_id}((null))>:1649]) will be created as active 12:15:56.087946+0800 runningboardd [app<{my_bundle_id}((null))>:1649] reported to RB as running 12:15:56.088053+0800 runningboardd Calculated state for app<{my_bundle_id}((null))>: running-active (role: Background) (endowments: (null)) 12:15:56.088114+0800 runningboardd [app<{my_bundle_id}((null))>:1649] Set jetsam priority to 0 [0] flag[1] 12:15:56.088136+0800 runningboardd [app<{my_bundle_id}((null))>:1649] Resuming task. 12:15:56.088211+0800 runningboardd [app<{my_bundle_id}((null))>:1649] Set darwin role to: Background 12:15:56.088449+0800 runningboardd [app<{my_bundle_id}((null))>:1649] set Memory Limits to Hard Inactive (4096) 12:15:56.089314+0800 runningboardd Successfully acquired underlying assertion for [app<{my_bundle_id}((null))>:1649] 12:15:56.589755+0800 runningboardd Invalidating assertion 33-76-23100 (target:app<{my_bundle_id}((null))>) from originator [osservice<com.apple.dasd>:76] 12:15:56.590332+0800 runningboardd Removed last relative-start-date-defining assertion for process app<{my_bundle_id}((null))> 12:15:56.593760+0800 runningboardd [app<{my_bundle_id}((null))>:1649] Suspending task. 12:15:56.594120+0800 runningboardd Calculated state for app<{my_bundle_id}((null))>: running-suspended (role: None) (endowments: (null)) From these logs, I understand that the system is accelerating the launch speed of the application. But the time interval between these two logs below is very short, which suggests that the prewarm is executed just before main, and then the process is suspended. Is this understanding correct? 12:15:56.089314+0800 runningboardd Successfully acquired underlying assertion ... 12:15:56.589755+0800 runningboardd Invalidating assertion ... Regarding "DAS DYLD3 Closure Generation," I speculate that after a user restarts their phone, the system uses DYLD3 to prepare closures for frequently used applications, allowing for faster application launches. Is this assumption correct?
5
0
669
Nov ’24
Async/Await and updating state
When using conformance to ObservableObject and then doing async work in a Task, you will get a warning courtesy of Combine if you then update an @Published or @State var from anywhere but the main thread. However, if you are using @Observable there is no such warning. Also, Thread.current is unavailable in asynchronous contexts, so says the warning. And I have read that in a sense you simply aren't concerned with what thread an async task is on. So for me, that begs a question. Is the lack of a warning, which when using Combine is rather important as ignoring it could lead to crashes, a pretty major bug that Apple seemingly should have addressed long ago? Or is it just not an issue to update state from another thread, because Xcode is doing that work for us behind the scenes too, just as it manages what thread the async task is running on when we don't specify? I see a lot of posts about this from around the initial release of Async/Await talking about using await MainActor.run {} at the point the state variable is updated, usually also complaining about the lack of a warning. But ow years later there is still no warning and I have to wonder if this is actually a non issue. On some ways similar to the fact that many of the early posts I have seen related to @Observable have examples of an @Observable ViewModel instantiated in the view as an @State variable, but in fact this is not needed as that is addressed behind the scenes for all properties of an @Observable type. At least, that is my understanding now, but I am learning Swift coming from a PowerShell background so I question my understanding a lot.
5
0
1.5k
Dec ’24
BGProcessingTask File Upload Limits
I have BGProcessingTask & BGAppRefreshTask working fine. The main purpose of my use of BGProcessingTask is to upload a file to AWS S3 using multipart/form-data. I have found that any file above about 2.5MB times out after running almost four minutes. If I run the same RESTful api using curl or Postman, I can upload a 25MB file in 3 seconds or less. I have tried to deliberately set .earliestBeginDate to 01:00 or 02:00 local time on the iPhone, but that does not seem to help. I use the delegate (yes, I am writing in Objective C) - URLSession:task:didSendBodyData:totalBytesSent:totalBytesExpectedToSend: and find that the iOS system uploads about 140kB every 15 seconds or so. I am looking for recommendations or insight into how I might enable uploads of 25MB files. I would be happy it I could do just one a day for my use case. I provide code on how I set up the NSURLSession and NSURLSessionDownloadTask, as it is my guess that if there is something that needs to be modified it is there. I have to believe there is a solution for this since I read in many posts here and in StackOverflow how developers are using this functionality for uploading many, many files. NSURLSessionConfiguration *sConf = [NSURLSessionConfiguration backgroundSessionConfigurationWithIdentifier:bkto.taskIdentifier]; sConf.URLCache = [NSURLCache sharedURLCache]; sConf.waitsForConnectivity = YES; sConf.allowsCellularAccess = NO; sConf.networkServiceType = NSURLNetworkServiceTypeVideot; sConf.multipathServiceType = NSURLSessionMultipathServiceTypeNone; sConf.discretionary = YES; sConf.timeoutIntervalForResource = kONEHOURINTERVAL; sConf.timeoutIntervalForRequest = kONEMINUTEINTERVAL; sConf.allowsExpensiveNetworkAccess = NO ; sConf.allowsConstrainedNetworkAccess = NO; sConf.sessionSendsLaunchEvents = YES; myURLSession = [NSURLSession sessionWithConfiguration:sConf delegate:self delegateQueue:nil]; And then later in the code... NSMutableURLRequest *request = [NSMutableURLRequest requestWithURL:[NSURL URLWithString:pth]]; request.HTTPMethod = kHTTPPOST; request.HTTPBody = [NSData my body data]; request.timeoutInterval = 60; [request setValue:@"*/*" forHTTPHeaderField:@"Accept"]; [request setValue:@"en-us,en" forHTTPHeaderField:@"Accept-Language"]; [request setValue:@"gzip, deflate, br" forHTTPHeaderField:@"Accept-Encoding"]; [request setValue:@"ISO-8859-1,utf-8" forHTTPHeaderField:@"Accept-Charset"]; [request setValue:@"600" forHTTPHeaderField:@"Keep-Alive"]; [request setValue:@"keep-alive" forHTTPHeaderField:@"Connection"]; NSString *contType = [NSString stringWithFormat:@"multipart/form-data; boundary=%@",bnd]; [request setValue:contType forHTTPHeaderField:@"Content-Type"]; [request addValue:[NSString stringWithFormat:@"%lu",(unsigned long)myData.length] forHTTPHeaderField:@"Content-Length"]; and here are a few lines from my logs to show the infrequent multi-part uploads of only small chunks of data by the iOS system: -[BKSessionManager URLSession:task:didSendBodyData:totalBytesSent:totalBytesExpectedToSend:]: bytesSent = 393,216 -[BKSessionManager URLSession:task:didSendBodyData:totalBytesSent:totalBytesExpectedToSend:]: totalBytesSent = 393,216 -[BKSessionManager URLSession:task:didSendBodyData:totalBytesSent:totalBytesExpectedToSend:]: task = BackgroundDownloadTask <76A81A80-4703-4686-8742-A0048EB65108>.<2>, time Fri Mar 7 16:25:27 2025 -[BKSessionManager URLSession:task:didSendBodyData:totalBytesSent:totalBytesExpectedToSend:]: bytesSent = 131,072 -[BKSessionManager URLSession:task:didSendBodyData:totalBytesSent:totalBytesExpectedToSend:]: totalBytesSent = 524,288 -[BKSessionManager URLSession:task:didSendBodyData:totalBytesSent:totalBytesExpectedToSend:]: task = BackgroundDownloadTask <76A81A80-4703-4686-8742-A0048EB65108>.<2>, time Fri Mar 7 16:25:42 2025 -[BKSessionManager URLSession:task:didSendBodyData:totalBytesSent:totalBytesExpectedToSend:]: bytesSent = 131,072 -[BKSessionManager URLSession:task:didSendBodyData:totalBytesSent:totalBytesExpectedToSend:]: totalBytesSent = 655,360 -[BKSessionManager URLSession:task:didSendBodyData:totalBytesSent:totalBytesExpectedToSend:]: task = BackgroundDownloadTask <76A81A80-4703-4686-8742-A0048EB65108>.<2>, time Fri Mar 7 16:25:56 2025 -[BKSessionManager URLSession:task:didSendBodyData:totalBytesSent:totalBytesExpectedToSend:]: bytesSent = 131,072 -[BKSessionManager URLSession:task:didSendBodyData:totalBytesSent:totalBytesExpectedToSend:]: totalBytesSent = 786,432
5
0
100
Mar ’25
Re-enrolling a LaunchDaemon, does it require user auth?
I am building an app that uses the SMAppService to register a LaunchDaemon that is bundled with my .app. I've got a priming flow created which walks the user through approving the service so that it will start on login. However, I need to also be able to upgrade this background service if the user updates the app. To do this, I think I need to call unregisterAndReturnError and then registerAndReturnError. From my testing, this seems to work correctly, but I have a concern. Will the user ever be prompted to re-authorize the LaunchDaemon that I am registering? If so, under what circumstances will that happen, and what does it look like (so that I can guide the user through it)?
5
0
186
May ’25
Bundled app error with Electron app on Mac App Store
Hi, I built an Electron app that uses puppeteer-cluster to open a bundled version of Chrome. Everything works before packaging/signing with electron builder. Transporter does not report any issues and the app opens in TestFlight. the Chrome.app is signed separately before running builder hardenedRuntime = false However, a permission error occurs when cluster attempts to launch Chrome: Error: Unable to launch browser, error message: Failed to launch the browser process! [0601/152740.225314:ERROR:bootstrap.cc(65)] bootstrap_check_in org.chromium.crashpad.child_port_handshake.9915.63117.BUEXLMXFWPLCEONM: Permission denied (1100) [0601/152740.226091:ERROR:file_io.cc(94)] ReadExactly: expected 4, observed 0 [0601/152740.229808:ERROR:bootstrap.cc(65)] bootstrap_check_in org.chromium.crashpad.child_port_handshake.9913.63115.VVKELOQUCWUYPFMQ: Permission denied (1100) [0601/152740.230244:ERROR:file_io.cc(94)] ReadExactly: expected 4, observed 0 [9911:45571:0601/152740.506968:ERROR:named_platform_channel_mac.cc(44)] bootstrap_check_in com.google.chrome.for.testing.apps.52995c87946bbcc94fc9a27df1478a13: Permission denied (1100) [9911:62467:0601/152740.507564:FATAL:mach_port_rendezvous.cc(281)] Check failed: kr == KERN_SUCCESS. bootstrap_check_in com.google.chrome.for.testing.MachPortRendezvousServer.9911: Permission denied (1100) at Cluster.<anonymous> (/Applications/MyApp.app/Contents/Resources/app.asar/node_modules/puppeteer-cluster/dist/Cluster.js:119:23) at Generator.throw (<anonymous>) at rejected (/Applications/MyApp.app/Contents/Resources/app.asar/node_modules/puppeteer-cluster/dist/Cluster.js:6:65) at process.processTicksAndRejections (node:internal/process/task_queues:105:5) I'm wondering if it's an issue with entitlements, or something more. The entitlements.mas.plist (aside from identifiers): com.apple.security.app-sandbox com.apple.security.cs.allow-jit com.apple.security.cs.allow-unsigned-executable-memory com.apple.security.cs.allow-dyld-environment-variables com.apple.security.network.client com.apple.security.network.server com.apple.security.files.user-selected.read-write com.apple.security.cs.disable-executable-page-protection com.apple.security.files.user-selected.executable I've spent many hours searching for a solution. Any help or insight would be greatly appreciated.
5
0
118
Jun ’25
How to capture stderr from a macOS GUI app (UIKit/AppKit) when launched via open or Finder?
For a macOS GUI application (with a UIKit or AppKit entry point), I want to reliably capture diagnostic logs sent to stderr — especially useful when the app is launched from a terminal script or runs in the background, and traditional GUI elements (like alert dialogs) may not be viable. This is to log startup failures or even success messages for later inspection. However, when the app is launched via open MyApp.app, stderr redirection like open MyApp.app 2> log.txt does not capture any output — the file is created, but remains empty. The only way I can capture stderr reliably is by bypassing the bundle and directly launching the binary inside with ./MyApp.app/Contents/MacOS/MyApp 2> ~/log.txt This logs as expected, but is not the user-friendly or typical way to launch apps on macOS. Double-clicking the app in Finder also does not show any stderr output. Is there any recommended or supported way to redirect or access stderr output when launching a .app bundle via open, or any best practice for logging critical failures from a GUI app when terminal output isn't visible?
5
0
136
Jun ’25
BGProcessingTaskRequest executing even after force-quit from App Switcher
Hello, I have a question regarding the behavior of BGProcessingTaskRequest when the app is force-quit by the user via the App Switcher. Based on common understanding and various discussions — including the following Apple Developer Forum threads: Waking up an iOS app after app is … | Apple Developer Forums Will BGAppRefreshTaskRequest will … | Apple Developer Forums Background fetch after app is forc… | Apple Developer Forums …it is widely understood that iOS prevents background execution (such as background fetch, push notifications, or BGTaskScheduler) after a user force-quits an app via the App Switcher. However, in my app, I have observed that a scheduled BGProcessingTaskRequest still executes even after the app has been explicitly terminated via App Switcher. The task is scheduled using submit(_:error:), and it is clearly running some time after the app has been closed by the user. That said, the task does run, but it appears to operate under tighter constraints — for example, it may be allowed to run for a shorter duration, and network requests appear to be more restricted compared to when the app is not force-quit. My questions are: Are there any documented or undocumented exceptions that allow this kind of behavior after force-quit? Could this be a bug or a behavior change in recent iOS versions? (I am observing this on iOS 18.3, 18.4, and 18.5) Any insights, experiences, or clarifications from Apple engineers or fellow developers would be greatly appreciated. Thank you!
5
0
123
Jul ’25
Unsandboxed XPCService launches sandboxed child processes
I have an application, it has main process and some child processes. As we want those child processes to have their own minimum sandbox privilege, not inheriting from parent process, we plan to use XPCService which uses a NSTask to launch those child processes, so those child processes can have its own sandbox privilege. We plan to deliver the application to Mac App Store, so process mode is: the sandboxed main process builds connections to the unsandboxed XPCService, the unsandboxed XPCService launch those sandboxed child processes. Can this process mode pass the Mac App Store rules? I see, there is a rule that all processes must be sandboxed, including XPCService. But I tested locally, the Application downloaded from Mac apple store also launches unsandboxed XPCService, like OneDrive. Do you have any suggestions for my application scenario, sandboxed child processes having its own privilege not inheriting from parent?
4
0
1k
Nov ’24
Schedule A Daily Task in MacOS Sandbox App
I am new to building apps for MacOS using SwiftUI but built apps for iOS currently in the store. I built an events app that stores a bunch of dates. The issue I have is that after X amount of time, the app needs to generate more events. In iOS I would use Background task to handle this, runs once daily etc. After much research I am pointed to using a LaunchAgents with an Embedded Helper Tool https://developer.apple.com/documentation/Xcode/embedding-a-helper-tool-in-a-sandboxed-app I am following this post: https://developer.apple.com/forums/thread/721737?answerId=739716022#739716022 I am stuck on setting up the plist file and clicking the button to try to add the launch item in simulator I get the following error: did not register, error: Error Domain=SMAppServiceErrorDomain Code=1 "Operation not permitted" UserInfo={NSLocalizedFailureReason=Operation not permitted} If this is the incorrect approach please let me know as I am stuck. Thanks.
4
0
582
Oct ’24
SMAppService re-register after app upgrade
I was experimenting with Service Management API and Xcode project from https://developer.apple.com/documentation/servicemanagement/updating-your-app-package-installer-to-use-the-new-service-management-api and faced some issues with the API. I replaced agent with XPC service and tried to re-register it. Use case is a new app package installation with a newer service binary. In order to get the running service restarted with the new binary it's required to unregister old version and register new one. Otherwise the old version would be still running after app upgrade. The problem is that register fails with "Operation not permitted" error after running unregister which seems to work fine. Experiments with some delays (500ms) between unregister and register seem to help but it's a not a good solution to work around the problem. I'm using open func unregister() async throws with description: The completion handler will be invoked after the running process has been killed if successful or will be invoked whenever an error occurs. After the completion handler has been invoked it is safe to re-register the service. Sample output with no 500ms sleep between unregister and register calls: /Library/Application\ Support/YourDeveloperName/SMAppServiceSampleCode.app/Contents/MacOS/SMAppServiceSampleCode unregister &amp;&amp; /Library/Application\ Support/YourDeveloperName/SMAppServiceSampleCode.app/Contents/MacOS/SMAppServiceSampleCode register Successfully unregistered LaunchDaemon(com.xpc.example.service.plist) Unable to register LaunchDaemon(com.xpc.example.service.plist): Error Domain=SMAppServiceErrorDomain Code=1 "Operation not permitted" UserInfo={NSLocalizedFailureReason=Operation not permitted} In fact it doesn't seem to be safe to re-register. Any explanation would much appreciated! ===================================================== Side issue #2: I tried to add a similar helper executable as in the original project with register/unregister and put it inside the same app bundle but at a different location like Contents/Helpers/ folder instead of Contents/MacOS. And it always fails with this error: Error Domain=SMAppServiceErrorDomain Code=3 "Codesigning failure loading plist: com.okta.service.osquery code: -67028" UserInfo={NSLocalizedFailureReason=Codesigning failure loading plist: com.okta.service.osquery code: -67028} When I moved the helper binary to Contents/MacOS/ folder along with the main app executable it starts working fine again. Other folders like Resources/XPCServices also don't work. Is it a hard requirement for an executable to be located inside main Contents/MacOS folder in order to be able to call SMAppService register/unregister APIs? I haven't found any documentation regarding this requirement. Thanks, Pavel
4
0
908
Nov ’24
Launchd ran job quits early
Hey, I have a user script that I have set up to run daily with launchd, the launchd logs indicate that the script is started but it's almost immediately exited with status 0. The odd thing is that it doesn't say that the script was killed or stopped in anyway, even though that is what seems to happen. 2024-12-04 14:15:04.970214 (gui/501/com.me.test) <Notice>: internal event: WILL_SPAWN, code = 0 2024-12-04 14:15:04.970320 (gui/501/com.me.test) <Notice>: service state: spawn scheduled 2024-12-04 14:15:04.970325 (gui/501/com.me.test) <Notice>: service state: spawning 2024-12-04 14:15:04.970431 (gui/501/com.me.test) <Notice>: launching: xpc event 2024-12-04 14:15:04.972067 (gui/501/com.me.test [26446]) <Notice>: xpcproxy spawned with pid 26446 2024-12-04 14:15:04.972096 (gui/501/com.me.test [26446]) <Notice>: internal event: SPAWNED, code = 0 2024-12-04 14:15:04.972107 (gui/501/com.me.test [26446]) <Notice>: service state: xpcproxy 2024-12-04 14:15:04.972160 (gui/501/com.me.test [26446]) <Notice>: internal event: SOURCE_ATTACH, code = 0 2024-12-04 14:15:04.998157 (gui/501/com.me.test [26446]) <Notice>: service state: running 2024-12-04 14:15:04.998180 (gui/501/com.me.test [26446]) <Notice>: internal event: INIT, code = 0 2024-12-04 14:15:04.998199 (gui/501/com.me.test [26446]) <Notice>: Successfully spawned shell_wrapper.sh[26446] because xpc event 2024-12-04 14:15:05.217482 (gui/501/com.me.test [26446]) <Notice>: exited due to exit(0), ran for 246ms 2024-12-04 14:15:05.217494 (gui/501/com.me.test [26446]) <Notice>: service state: exited 2024-12-04 14:15:05.217502 (gui/501/com.me.test [26446]) <Notice>: internal event: EXITED, code = 0 2024-12-04 14:15:05.217506 (gui/501 [100003]) <Notice>: service inactive: com.me.test 2024-12-04 14:15:05.217523 (gui/501/com.me.test [26446]) <Notice>: service state: not running In the script itself I log at the end to indicate that it finished successfully, when ran with launchd I never reach this point however. The script also do two network requests and make a pause for between 5-10 seconds with time.sleep(paus) (python script). However the launchd log indicates that the script finished in 246ms. Not sure what is going on, the script itself does not require any privileges above a normal user. I have tried running the script directly, and it works as expected. I have also tried evoking the script from launchd explicitly with: launchctl kickstart which starts the job but gives the same result as the scheduled job, except the log now says: `launching: non-ipc demand' instead. I have also tried to remove the networking requests and replace them by reading data from a file without any sleep, no difference in outcome though. I have generated the launchd plist file using Lingon.app and it's placed in ~/Library/LaunchAgents Not sure what more information I can provide, but need some suggestion on how I can proceed help solving this.
4
0
491
Dec ’24
MacOS Application as a daemon or in non-interaction mode
We are building a 'server' application that can either run as a daemon or can run in background without showing any GUI. Basically, the end user can either configure this to run as a daemon so that it can be tied to the user's session or will launch the process which user will start and quit as needed. I wanted to understand what is the recommended mechanism for such an application from Apple - Should this application be built as a macOS Bundle ? Apple documentation also says that we should not daemonize the process by calling fork. Hence if we create a unix-style executable, will I not need to fork to make it run in a detached state when I launch the executable via double-click ? [Reference Link] Is it fine to have an application on macOS which is a bundle but does not show any UI when launched by double click on the app-icon or via 'open'? While we have been able to achieve this by using NSApplicationMain and not showing the UI, was wondering if using CFRunLoop is best for this case as it is a non-gui application. If we can get the right documentation link or recommendations on how we should build such an application which can run in a non-gui mode and also in a daemonized manner, it will help us. Should the application be always built as a macos bundle or should it be a unix-style executable to support both the cases - by the same application/product and how should we look at the distribution of such applications.
4
1
475
Mar ’25
XPC activity doesn’t fire while main app is closed
Hi, I have a sandboxed app with a bundled sandboxed XPC service. When it’s launched, the XPC service registers a repeating XPC activity with the system. The activity’s handler block does get called regularly like I’d expect, but it stops being called once the main app terminates. What’s the recommended way to fix this issue? Could I have a bundled XPC service double as a launch agent, or would that cause other problems?
4
0
112
May ’25
Incorrect behaviour of task_info() syscall after an unrelated dlclose() call
For some reason, after invoking an unrelated dlclose() call to unload any .dylib that had previously been loaded via dlopen(..., RTLD_NOW), the subsequent call to task_info(mach_task_self(), TASK_DYLD_INFO, ...) syscall returns unexpected structure in dyld_uuid_info image_infos-&gt;uuidArray, that, while it seems to represent an array of struct dyld_uuid_info elements, there is only 1 such element (dyld_all_image_infos *infos-&gt;uuidArrayCount == 1) and the app crashes when trying to access dyld_uuid_info image-&gt;imageLoadAddress-&gt;magic, as image-&gt;imageLoadAddress doesn't seem to represent a valid struct mach_header structure address (although it looks like a normal pointer within the process address space. What does it point to?). This reproduces on macOS 15.4.1 (24E263) Could you please confirm that this is a bug in the specified OS build, or point to incorrect usage of the task_info() API? Attaching the C++ file that reproduces the issue to this email message It needs to be built on macOS 15.4.1 (24E263) via Xcode or just a command line clang++ compiler. It may crash or return garbage, depending on memory layout, but on this macOS build it doesn’t return a correct feedfacf magic number for the struct mach_header structure. Thank you Feedback Assistant reference: FB18431345 //On `macOS 15.4.1 (24E263)` create a C++ application (for example, in Xcode), with the following contents. Note, that this application should crash on this macOS build. It will not crash, however, if you either: //1. Comment out `dlclose()` call //2. Change the order of the `performDlOpenDlClose()` and `performTaskInfoSyscall()` functions calls (first performTaskInfoSyscall() then performDlOpenDlClose()). #include &lt;iostream&gt; #include &lt;dlfcn.h&gt; #include &lt;mach/mach.h&gt; #include &lt;mach-o/dyld_images.h&gt; #include &lt;mach-o/loader.h&gt; void performDlOpenDlClose() { printf("dlopen/dlclose function\n"); printf("Note: please adjust the path below to any real dylib on your system, if the path below doesn't exist!\n"); std::string path = "/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/lib/libswiftDemangle.dylib"; printf("Dylib to open: %s\n", path.c_str()); void* handle = ::dlopen(path.c_str(), RTLD_NOW); if(handle) { ::dlclose(handle); } else { printf("Error: %s\n", dlerror()); } } void performTaskInfoSyscall() { printf("Making a task_info() syscall\n"); printf("\033[34mSource File: %s\033[0m\n", __FILE__); task_t task = mach_task_self(); struct task_dyld_info dyld_info; mach_msg_type_number_t size = TASK_DYLD_INFO_COUNT; kern_return_t kr = task_info(task, TASK_DYLD_INFO, (task_info_t)&amp;dyld_info, &amp;size); if (kr != KERN_SUCCESS) { fprintf(stderr, "task_info failed: %s\n", mach_error_string(kr)); } const struct dyld_all_image_infos* infos = (const struct dyld_all_image_infos*)dyld_info.all_image_info_addr; printf("version: %d, infos-&gt;infoArrayCount: %d\n", infos-&gt;version, infos-&gt;infoArrayCount); for(uint32_t i=0; i&lt;infos-&gt;infoArrayCount; i++) { dyld_image_info image = infos-&gt;infoArray[i]; const struct mach_header* header = image.imageLoadAddress; printf("%d ", i); printf("%p ", (void*)image.imageLoadAddress); printf("(%x) ", header-&gt;magic); printf("%s\n", image.imageFilePath); fflush(stdout); } printf("\n\n"); printf("infos-&gt;uuidArrayCount: %lu\n", infos-&gt;uuidArrayCount); for(uint32_t i=0; i&lt;infos-&gt;uuidArrayCount; i++) { dyld_uuid_info image = infos-&gt;uuidArray[i]; const struct mach_header* header = image.imageLoadAddress; printf("%d ", i); printf("%p ", (void*)image.imageLoadAddress); printf("(%x)\n", header-&gt;magic); fflush(stdout); } printf("task_info() syscall result processing is completed\n\n"); } int main(int argc, const char * argv[]) { performDlOpenDlClose(); performTaskInfoSyscall(); return 0; }
4
0
140
Jun ’25
Using raise in GCD can cause timing issues with the signal mechanism.
when we use raise in GCD, the signal handler is executed asynchronously, whereas in pthread, it is executed synchronously as expected. example: #include &lt;Foundation/Foundation.h&gt; #include &lt;pthread/pthread.h&gt; static void HandleSignal(int sigNum, siginfo_t* signalInfo, void* userContext) { printf("handle signal %d\n", sigNum); printf("begin sleep\n"); sleep(3); printf("end sleep\n"); } void InstallSignal(void) { static const int g_fatalSignals[] = { SIGABRT, SIGBUS, SIGFPE, SIGILL, SIGPIPE, SIGSEGV, SIGSYS, SIGTRAP, }; int fatalSignalsCount = sizeof(g_fatalSignals) / sizeof(int); struct sigaction action = {{0}}; action.sa_flags = SA_SIGINFO | SA_ONSTACK; #if defined(__LP64__) action.sa_flags |= SA_64REGSET; #endif sigemptyset(&amp;action.sa_mask); action.sa_sigaction = &amp;HandleSignal; struct sigaction pre_sa; for(int i = 0; i &lt; fatalSignalsCount; i++) { int sigResult = sigaction(g_fatalSignals[i], &amp;action, &amp;pre_sa); } } void* RaiseAbort(void *userdata) { raise(SIGABRT); printf("signal handler has finished\n"); return NULL; } int main(int argc, const char * argv[]) { InstallSignal(); dispatch_async(dispatch_get_global_queue(0, 0), ^{ raise(SIGABRT); // abort(); // abort() is ok RaiseAbort(nullptr); }); // pthread is ok // pthread_t tid; // int ret = pthread_create(&amp;tid, NULL, RaiseAbort, NULL); // if (ret != 0) { // fprintf(stderr, "create thread failed\n"); // return EXIT_FAILURE; // } [[NSRunLoop mainRunLoop] run]; return 0; } console log: signal handler has finished handle signal 6 begin sleep end sleep
4
0
176
3w
BGContinuedProcessingTask code pauses when device is locked
I have been experimenting with the BGContinuedProcessingTask API recently (and published sample code for it https://github.com/infinitepower18/BGContinuedProcessingTaskDemo) I have noticed that if I lock the phone, the code that runs as part of the task stops executing. My sample code simply updates the progress each second until it gets to 100, so it should be completed in 1 minute 40 seconds. However, after locking the phone and checking the lock screen a few seconds later the progress indicator was in the same position as before I locked it. If I leave the phone locked for several minutes and check the lock screen the live activity says "Task failed". I haven't seen anything in the documentation regarding execution of tasks while the phone is locked. So I'm a bit confused if I encountered an iOS bug here?
4
0
158
4w