Posts under App & System Services topic

Post

Replies

Boosts

Views

Activity

Supported third-party integration for system Apple Account sign-in on macOS
We are assessing an independent macOS client for user-owned physical Macs. The requirement is to initiate and complete a real system Apple Account sign-in, including system account persistence and service activation. This is not Sign in with Apple for signing in to our own application. Our current test machine runs macOS 14.6.1 (23G93). We understand that observations on that version do not establish the supported API surface on current macOS releases. Could Apple clarify: Is there a documented third-party API or Apple-provided broker that supports this workflow? If so, what are its supported OS versions, caller requirements, and completion/result contract? If this requires an approved capability or a specific program, is there a published eligibility and application channel? We are not assuming that ordinary Developer Program membership grants private system-account privileges. If the only supported approach is a handoff to System Settings, can you confirm that boundary and identify any documented completion/status mechanism? A UI-only handoff would be a separate integration and would not fulfill our independent-client requirement. We are asking about supported integration options, not implementation details of internal APIs. We do not have a separate approved backend or a developer-program team at present, and would like to establish feasibility before purchasing membership or attempting account writes. No credentials, account identifiers, tokens, device materials, host addresses, or binary dumps are included.
0
0
16
8h
Supported non-admin service identity for a macOS LaunchDaemon
I'm developing a private macOS application on an Apple-silicon Mac mini running macOS Golden Gate 27.2. I'm evaluating a system LaunchDaemon running under a dedicated, non-administrator identity that cannot log in interactively. A separate desktop application would communicate with the daemon through authenticated local IPC. Before implementing this architecture, I would like to establish the officially supported approach. My main question: What is the Apple-supported method for creating and maintaining a dedicated, non-administrator, non-interactive service identity for a system LaunchDaemon on macOS 27.2? In particular: Is there a documented procedure for creating, restricting, disabling, recovering and eventually removing this identity? Are dscl, sysadminctl or another account-management mechanism supported for this purpose? Can a LaunchDaemon run under this identity without an interactive user login, including when registered through SMAppService? Are there any secure-token, FileVault, volume-ownership or keychain restrictions that would affect this design? If a dedicated non-login service identity is not a supported architecture, what alternative does Apple recommend? I'd particularly appreciate links to official Apple documentation or guidance from Apple engineers. Thank you.
10
0
558
8h
Sandbox restriction blocks Keychain access during Live Activity rendering: supported design for encrypted updates?
Hi, I'm building a Live Activity prototype using fictitious data. I want incoming updates to stay encrypted until they reach my iPhone, without putting the decryption key in ActivityAttributes, ContentState, or archived views. My goal is timely updates without an alert for each change. I'm using Xcode 27.2 and the iPhoneOS 27.2 SDK. My iPhone was running iOS 27.2.1 when I last checked. I haven't recorded its exact OS build or the OS version and build for every earlier test. My app and WidgetKit extension have the same shared Keychain access group in their signed entitlements. My app creates a 32-byte test key as a generic password with kSecAttrAccessibleAfterFirstUnlockThisDeviceOnly, with synchronization disabled. The extension only reads the key. A failed lookup never causes the key to be replaced or deleted. When I try to read the key while rendering the Live Activity, the first, metadata-only SecItemCopyMatching call returns errSecNotAvailable (-25291). It fails before I request any secret bytes. I reproduced this after stopping and restarting the activity with my iPhone unlocked. I get the same result with a non-interactive LAContext and with kSecUseAuthenticationUIFail without a supplied context. I used the latter only for comparison, since it is deprecated. My query specifies the class, fixed service and account, and shared access group. It also sets kSecUseDataProtectionKeychain. For the metadata lookup, I use kSecAttrSynchronizableAny, kSecReturnAttributes: true, and kSecMatchLimitAll, without requesting data or item references. I only request the key bytes after finding exactly one item and validating its policy, so that second step is never reached here. For comparison, I added a regular widget to the same signed WidgetKit extension. Its timeline provider successfully validates the metadata using the same read-only query and client, while the Live Activity still returns errSecNotAvailable. I captured these results separately; I haven't established whether they ran simultaneously or in different processes. I have only confirmed metadata access in the regular widget, not access to the key bytes. I also tried decrypting in my foreground app and then passing the fictitious summary to ActivityKit through a local update. That renders correctly on my iPhone. It doesn't test remote delivery or background decryption, or fix the Keychain failure during rendering. I repeated the test with debug messages enabled in Console. The Live Activity in my diagnostic build still shows the same Keychain failure. The capture from the extension now includes the connection error, not just the generic securityd message: activating connection: ... name=com.apple.securityd failed to do a bootstrap look-up: xpc_error=[159: Unknown error: 159] invalidated after a failed init got event: Connection invalid Failed to talk to securityd after 4 attempts. error:[-25291] Error Domain=com.apple.security.xpc Code=3 Connection invalid - Connection init failed at lookup with error 159 - Sandbox restrictionThis excerpt omits connection addresses and account fields. The same -25291 error and sandbox-restriction reason appear five times in this capture. It establishes the reported reason for the connection failure in this test; I haven't established whether this is an intended restriction, an OS issue, or a configuration issue. I also don't infer the actual retry count from the generic message. I'd appreciate guidance on these questions: Given the explicit sandbox-restriction error for com.apple.securityd, is non-interactive Keychain access intentionally unavailable during Live Activity rendering, or should I investigate an OS or configuration issue? Displaying live data with Live Activities documents a separate sandbox and restrictions on network and location access, but I haven't found its Keychain access contract. How should I interpret the different result from my regular widget's timeline provider? For encrypted content-state updates, is there a supported execution point where my app can authenticate and decrypt each update before handing a local render value to ActivityKit? I found the documented app wake-up for push-to-start, but not an equivalent for ordinary updates. If neither approach is supported, how can I deliver timely, alert-free encrypted Live Activity updates while keeping the key solely in Keychain and out of archived state and views? How does ActivityKit protect static ActivityAttributes at rest? Can those attributes be included in backups or replicated to an Apple Watch or Mac, separately from rendered views? Does any of that data pass through Apple servers, and is there a public API to opt out of replication? Would a per-activity session key in locally created static attributes be a supported design, or does Apple advise against storing secrets there? I haven't implemented this alternative. I would keep the long-term pairing key in Keychain, exclude the session key from every APNs payload, and start the activity locally rather than through push-to-start. I'd like to understand the storage and replication guarantees before considering it further. I'm looking for a supported design without weakening Keychain protection, allowing authentication UI during rendering, or suppressing an ordinary notification to bypass its requirements. I can prepare a minimal reproducer with fictitious data and no tokens, real identifiers, or conversation content. Thanks for your help.
3
0
105
9h
macOS: pre-established stop mechanism when a VM supervisor stops progressing
I am evaluating a bounded workload host using Virtualization.framework on Apple silicon. This is a question about supported architecture and API contracts, not a report of a reproduced bug. No particular macOS release has yet been qualified for this design; please identify the releases and prerequisites to which an answer applies. I have read the discussions about cleanup after the VM-owning application terminates: https://developer.apple.com/forums/thread/848542 https://developer.apple.com/forums/thread/850709 I understand that this cleanup is different from an externally observable, time-bounded teardown guarantee. I am not asking again for a replacement application's handle to a VM whose owner has already died. The unresolved case is different: A host application owns a VM, but its serial VM-control queue is no longer progressing. A separate supervisor is intended to initiate the fallback stop action against that application. An external stop trigger occurs. The supervisor is then lost or stops progressing before it enters the signal operation; the VM-owning application remains alive. Loss of the supervisor role does not necessarily mean that its process has exited, so a process-exit notification cannot be assumed. Complete host or kernel failure is not part of this scenario. Is there a supported public mechanism that can be established before workload preparation and retain authority to act on the same target application after this loss, without requiring another action by the supervisor or progress on the VM-control queue? For any such mechanism, I would particularly appreciate clarification of: How its authority is bound to the intended target during incomplete setup, including process creation that has begun but has not returned a usable target identity. Assuming an already-running watchdog leaves this setup boundary unanswered. Which actor performs the remaining action when the supervisor cannot, and what lifecycle, privilege and progress assumptions that actor requires. Whether any documented progress bound applies when a required operation does not return. The design requirement is completion of the whole prescribed stop/escalation attempt, including required failure-handling steps, within 10 seconds of the same first external trigger. Starting a timer or entering a signal call is not equivalent to completing that attempt. This is not a request for unconditional successful process or VM termination. A supported narrower mechanism, or an explicit statement that macOS provides no such contract for this case, would also be useful. I am looking for public interfaces and their documented limits, not private helper-process identities or kernel implementation details. Thank you.
5
0
91
9h
StoreKit 2 transaction verification fails during App Review with localized error 「資訊對裝置無效」; how should our app handle this?
We are implementing in-app purchases using StoreKit 2. The issue we need guidance on is how our app should correctly handle a StoreKit 2 transaction that becomes .unverified during Apple App Review. In our local development testing, purchases work correctly. When we test the same IAP products locally, the transaction returned from Product.purchase() is .verified, and the purchase flow completes successfully. However, during Apple App Review, the reviewer appears to encounter a failure where the StoreKit 2 transaction is returned as .unverified on the client side. The original localized error message shown is in Traditional Chinese: 「資訊對裝置無效」 Our English translation of this message is: “The information is not valid for this device.” Because the message is localized, we are not sure what the exact original English StoreKit error string would be. This failure happens in the client app before our backend receipt or server-side validation flow. We are not using verifyReceipt for this step. The failure occurs at the StoreKit 2 VerificationResult level when checking the transaction returned by StoreKit. Our current implementation only accepts .verified transactions. If StoreKit 2 returns .unverified, we treat the transaction as failed and do not unlock the purchased content. We believe this is the recommended secure behavior, but we would like to confirm the correct handling, especially because this issue is appearing during App Review while our local development purchases succeed. Additional context: • Distribution/review context: App Review / IAP review • API: StoreKit 2 • Failure point: VerificationResult returned as .unverified • Localized error message: 「資訊對裝置無效」 • Rough English translation: “The information is not valid for this device.” • Local development testing: purchases return .verified and succeed • The issue occurs before backend validation • Our app currently rejects .unverified transactions and does not unlock content Questions: What are the known causes of StoreKit 2 returning .unverified with the localized error message 「資訊對裝置無效」 during App Review? Is our current behavior correct: only unlock content for .verified transactions and reject .unverified transactions? If App Review receives .unverified, what should the app do from Apple’s recommended perspective? Should we show a retry/error message, ask the reviewer to retry, refresh entitlements, or take another action? Is there any App Review environment, device configuration, Apple ID configuration, or StoreKit account state that can cause this device-related verification error? Since local development purchases succeed but App Review fails, what diagnostics or logs should we collect and provide to identify the root cause? Should we ever bypass StoreKit 2 transaction verification in App Review/TestFlight, or should .unverified always be treated as a failed transaction? We would appreciate guidance on the correct code-level handling so that our app remains secure while also passing App Review reliably. Thank you.
2
0
560
9h
Full Disk Access on macOS 27.3
I'm looking into the changes being implemented and I find this way of handling by Apple on short notice problematic. Now our customers (those that don't have MDM) will have to jump through hoops to enable FDA for our product, a security product that also uses Endpoint Security. We already have the entitlement for endpoint security, which requires FDA in order to work, so why make it so difficult instead of just allowing FDA for apps that have this entitlement? Why only backup applications are exempted?
2
0
87
10h
Access to related files outside Sandbox with NSFileCoordinator and NSFilePresenter does not work anymore in macOS 27.3 beta 2
In macOS 27.3 beta 2, it is no longer possible to use NSFileCoordinator with NSFilePresenter instances to open a SQLite file with related WAL and SHM files when a sandboxed app has access to the primary SQLite file. In macOS 27.3 beta 1 and previous macOS releases, the Related File Access API works as expected. This happens regardless where the files are located, so it seems to be unrelated to recent changes in Privacy & Security, although it could be an unexpected side effect of these changes. Instead of correctly registering NSFilePresenter instances, errors like the following are logged in the XCode console: NSFileSandboxingRequestRelatedItemExtension: Failed to issue extension for /Users/testuser/Desktop/Sample/Example.sqlite-wal because: Error Domain=NSPOSIXErrorDomain Code=1 "Operation not permitted" (The related item is protected by a different storage class than the primary item) UserInfo={NSDebugDescription=The related item is protected by a different storage class than the primary item} +[NSFileCoordinator addFilePresenter:] could not get a sandbox extension. primaryPresentedItemURL: file:///Users/testuser/Desktop/Sample/Example.sqlite, presentedItemURL: file:///Users/testuser/Desktop/Sample/Example.sqlite-walThe mentioned Related File Access API is documented here. I have filed a Feedback, which includes a sample project: FB25625743 My question is: Is this a bug, or new intended behavior, or is there something else changed that should be dealed with when using this API in macOS 27.3 beta 2?
1
0
59
10h
Can macOS verify a worker’s effective sandbox policy before it runs?
I’m building a local app on an Apple silicon Mac. Before a worker or any of its child processes can run, I need to verify the worker’s effective filesystem and network restrictions, including inherited restrictions. I need to allow that same pending launch only if the policy matches, and deny it if verification fails or times out. I reviewed Apple’s App Sandbox guidance and the Endpoint Security documentation for exec authorization and entitlement metadata. I haven’t found documentation showing that these expose the complete effective policy for the pending launch. Is there a supported public macOS API that provides this guarantee? If so, which API and macOS versions support it? If not, is this outside what the public APIs can verify?
1
0
184
10h
Subject: Critical APFS Data Loss Bug & Error 1110 Loop during official Finder Update on iPhone 14 (A15)
Hello, I am requesting an engineering escalation regarding a critical bug in Apple's APFS update pipeline that trapped an iPhone 14 (A15, 128 GB) in recovery mode during an official Finder update from iOS 16.6 to iOS 18 (27.2.1, 24B156) on macOS 14.5. The device hardware is 100% healthy (Apple MRI diagnostics all green), the passcode is known, and 101,175 MB (341,967 files) of user data remain intact on the flash storage. However, the device cannot boot or complete an update due to a corruption caused by Apple's own restore tools. --- ### What Happened (Verified in Logs): 1. **Initial Full Storage State:** The device had ~105 GB utilized and was updated via Finder using "Update" (not Restore). 2. **Destructive fsck Behavior:** In pass 3, the restore ramdisk ran `/sbin/fsck_apfs -dy /dev/disk2`. On the nearly full container, fsck erroneously flagged 8,182 live omap mappings as "orphans" and deallocated them, causing extent-ref b-tree node `0x10bb3` to be zeroed out (TRIMmed by the flash controller). 3. **Split Failure:** In pass 4, the iOS 18 System volume installed and sealed, but `split_data_volume` failed in 4 seconds when `apfs_migrator` encountered block `0x10bb3` (`obj_read` checksum failure returning EILSEQ / error 92). 4. **False Error 1110 Loop:** Every subsequent attempt now halts immediately at `space_checks` with Error 1110. --- ### Reverse Engineering Proof: Disassembly of `restored_update` (arm64e, VA 0x10001afcc / 0x10001b0e8) confirms a major error-handling bug: When `splitter_dump` encounters the unreadable block `0x10bb3`, it fails with `EILSEQ` (92). Instead of handling this error, `restored_update` executes: `mov x19, #-1; mov x0, x19; ret` The caller receives `-1 bytes of free space`, logs: "Data volume requires splitting. Splitter will leave -1 bytes of free space after completion. Current free space on device: 6620069888 bytes. Insufficient space on device to run splitter" and aborts with **CHECKPOINT FAILURE:(FAILURE:1110)**. The device is NOT out of space (it has 6.3 GB free); it is blocked solely because `restored_update` misinterprets an EILSEQ filesystem error as negative free space. --- ### Current State & Request: - The encryption keys and SEP/gigalocker are intact (`gigalocker: ONLINE`). - In the restore ramdisk, `/dev/disk2s2` mounts cleanly as read-write on `/mnt2`. - Zero APFS snapshots remain on the volume, meaning the extent-ref tree is theoretically fully reconstructible from the valid fsroot inode tree. I am requesting assistance from CoreOS / APFS engineering: 1. Is there an engineering update ramdisk or build that can bypass the `splitter_dump` EILSEQ check, or rebuild the extent-ref tree, so the update can finalize? 2. Alternatively, can an engineering liaison facilitate a non-destructive extraction given that the volume mounts read-write and the user passcode is known? I can immediately provide full terminal transcripts, Finder updater logs (`iPhoneUpdater-*.log`), and disassembly details. Thank you, Samuel Journo
1
0
97
11h
macOS APFS: com.apple.macl and UF_TRACKED appear after creating a hard link — safe file replacement and rollback?
I'm investigating an unexpected metadata change during a controlled file deployment on macOS. Environment and observed state Two regular configuration files are used by a local service. Before replacement, the first file had: SHA-256: verified. File flags: 0. Extended attributes: none. Hard-link count: 1. Owner/group: 501/20. Permissions: 0644. ACL: empty. A backup procedure created a hard link to preserve the original inode before replacing the active pathname with an independently created file. Immediately after creating and verifying the hard link, the original file was observed with: The same SHA-256 and inode. Flags: 64 (UF_TRACKED). Extended attribute: com.apple.macl. Hard-link count: 2. No file content was replaced. The procedure stopped because the metadata no longer matched the original baseline. Additional observations Other backup methods also encountered unexpected metadata changes during destination creation. Tests using renamex_np with RENAME_SWAP successfully exchanged and restored artificial files with flags 0 and empty extended attributes. A test with artificial files already containing com.apple.provenance also succeeded. We could not reproduce the specific UF_TRACKED + com.apple.macl + two-hard-link state in an isolated fixture using ordinary documented operations. The operational original and its second hard link currently share the same inode. No attempt has been made to remove the protected attribute or change its flags. Questions for Apple filesystem/security specialists Under what documented circumstances can creating a hard link or related Finder/TCC document-tracking behavior cause UF_TRACKED and com.apple.macl to appear on an existing file? Are these metadata changes expected for a file accessed through a user-authorized Documents location? What supported mechanism can preserve the current original inode and provide a verifiable rollback when replacing its active pathname? Is renamex_np(..., RENAME_SWAP) appropriate for this situation, and what special considerations apply to the existing second hard link and protected extended attributes? What verification criteria would you recommend before permitting a production replacement? I am looking for a documented, reversible procedure that preserves macOS security protections. I do not want to disable SIP, forcibly remove com.apple.macl, or bypass TCC. I can provide sanitized metadata comparisons and a minimal reproduction if needed.
1
0
138
11h
watchOS — Continuous Core Motion monitoring in the background for independent fall detection
I'm developing an independent fall-detection application for Apple Watch using Core Motion (CMMotionManager) at 20 Hz. The goal is to identify potential falls using accelerometer, gyroscope and orientation data, then notify nominated family contacts. The application is deliberately independent of Apple's built-in Fall Detection and Emergency SOS. Test environment Physical Apple Watch running watchOS 27.2 Xcode 27.2 Core Motion device-motion callbacks on a dedicated serial OperationQueue Motion processing on the MainActor Tests performed without the Xcode debugger attached Observed results With the Watch display awake: 2,300 processed samples Longest sensor timestamp gap: 0.05 seconds Longest callback arrival gap: 0.66 seconds Callback interruptions over one second: 0 With the Watch dimmed/backgrounded: 3,229 processed samples Longest sensor timestamp gap: 0.30 seconds Longest callback arrival gap: 29.11 seconds Callback interruptions over one second: 6 Maximum sample delivery lag: 29.065 seconds The motion timestamps appear nearly continuous, but callback delivery becomes substantially delayed after the app enters the background. The monitoring flag remains true, and no explicit call to stop motion updates occurs when the app becomes inactive. Questions Is there an Apple-supported background execution mechanism or entitlement allowing a third-party watchOS app to continuously process accelerometer and gyroscope data for independent fall detection, without requiring an active workout session? Would CMBatchedSensorManager provide any useful background delivery capability for this application, or would its callbacks remain subject to normal watchOS suspension? Is there a recommended architecture for detecting a potential fall and initiating a timely family notification while the Watch app is backgrounded, without relying on Apple's built-in Fall Detection or Emergency SOS? We understand that workout sessions and extended runtime sessions have specific intended purposes. We want to implement this correctly rather than use an unrelated background mode to prevent suspension. Any guidance from Apple engineers or developers experienced with watchOS background Core Motion processing would be greatly appreciated.
2
0
98
12h
NEURLFilterManager: inconsistent per-domain blocking (some sites never block, one flaps) - confirmed PIR database is correct
Hello, Some context: DropBet is an app that helps people with gambling addiction stay off gambling sites and apps, using NEURLFilterManager/PIR to block a large list of known gambling domains system-wide. For a lot of our users this filter is a core part of their recovery > reliability really matters here. Since fixing an unrelated HPKE encryption bug (separate thread, now closed), the filter works end-to-end for the first time, but blocking is inconsistent per domain in a way I can't explain from my side: winamax.fr and the Winamax app: blocked reliably, every single time > clean "connection failed" style block, confirmed to load fine with protection off. unibet.fr and betclic.fr: never blocked, consistently > confirmed even after a full device reboot (rules out stale caching), and confirmed identical across both Safari and Chrome on iOS tested side-by-side (so not browser-specific). olybet.fr: flaps > the page reloads every 2-3 seconds with protection on, also identical across both browsers. Confirmed this only happens with the filter enabled (tested toggling protection off/on against the same page). unibet.fr auto-redirects into the native Unibet app via a Universal Link before the page loads, unlike betclic.fr which stays in the browser > different code path, possibly relevant. To rule out any issue on our end, I wrote a standalone test that queries our deployed PIR database shards directly - bypassing OHTTP, the app, and the device entirely - for all of the domains above plus their www. variants. All of them correctly return the blocked value, no exceptions, including unibet.fr and betclic.fr which never block on the actual device. So the database, the PIR protocol implementation, and the server are confirmed correct for every domain I've tested. This points at something in how NEURLFilterManager/ciphermld applies a correctly-received PIR verdict on-device - possibly related to query batching, an evaluation-key edge case, or Universal Link handling bypassing the filter before a web request is even made. Is there anything you can share about how the client applies results once decrypted, or known issues around any of this?
7
0
339
20h
StoreKit introductoryOfferEligibility: safe retry after JWS expiration
How should a server safely retire or renew an unused JWS for Product.PurchaseOption.introductoryOfferEligibility(compactJWS:) after cancellation or interruption? Our application allows a first-purchase discount once per application account across payment platforms. We bind the signed eligibility request to a verified Apple app transaction, product, and purchase attempt, persist the exact JWS, and reuse it for the same attempt. Once a paid purchase is verified, we consume the application's eligibility. We do not automatically issue another nonce after cancellation, a signature error, or an empty transaction history. Apple's JWS documentation describes a one-time nonce and expiration calculated from iat, and says not to include exp: https://developer.apple.com/documentation/storekit/generating-jws-to-sign-app-store-requests I could not find the lifetime or a supported retirement/renewal procedure for the introductory-offer-eligibility audience. Keeping an unusable authorization forever would prevent the customer from trying another checkout; replacing it without retiring the original might leave two redeemable authorizations. What we observed Environment: Xcode 27.2 (27B52), macOS 26.6.2, physical iPhone 16 with iOS 27.0, Apple Sandbox. Verify the Apple-signed AppTransaction and issue one eligibility JWS. Start the purchase using that JWS, then cancel the payment sheet (userCancelled). Terminate and relaunch the app without resetting purchase history or replacing the authorization. Retry the same product and Apple account using the byte-for-byte identical JWS. Complete the Sandbox purchase and verify the Apple-signed transaction on an isolated server. The successful verification happened about 164 seconds after issuance. This is one successful cancellation/retry observation, not a reproduced expiration failure or proof of a general validity/nonce-consumption rule. It also did not verify our intended introductory price. The purchase call uses the persisted value: let result = try await product.purchase(options: [ .introductoryOfferEligibility(compactJWS: existingJWS) ])Supported lifecycle questions What is the validity period derived from iat, including clock-skew allowances and any Sandbox/production differences? Which API or documented error can establish that a previously issued authorization is definitively unusable, rather than merely absent from transaction history? When is the nonce consumed, and what happens after userCancelled, interruption, pending approval, expiration, or an uncertain result? Is invalidOfferSignature specific enough to distinguish expiration from consumption, account mismatch, or a bad signature? Is it supported to keep the same nonce and all bindings, update only iat, and re-sign? If so, is one-time use enforced across every version and for what retention period? If renewal requires a new nonce, how can the prior authorization be invalidated first? Is there another recommended recovery flow that prevents duplicate redemption without indefinitely blocking checkout? DTS directed us to ask this lifecycle question on the forums. A sanitized minimal Xcode project was already supplied privately to DTS; it demonstrates the observed cancel/relaunch/same-JWS retry, not an expiration failure. We would appreciate a supported procedure or documentation reference.
1
0
76
20h
Supported publication of HID Report Reference descriptor 0x2908 on iOS
Supported publication of HID Report Reference descriptor 0x2908 on iOS We request clarification of the supported Core Bluetooth API contract for one specific call: publishing the Bluetooth Low Energy HID Report Reference descriptor (0x2908) using CBMutableDescriptor on iOS. Our iPhone application uses CBPeripheralManager to publish a Human Interface Device over GATT Profile (HOGP) service for an iPad. Its Report characteristics use two-byte Report Reference values containing the report ID and report type. The relevant call is: let reportReference = CBMutableDescriptor( type: CBUUID(string: "00002908-0000-1000-8000-00805F9B34FB"), value: Data([0x01, 0x01]) )The initializer documentation lists only CBUUIDCharacteristicUserDescriptionString and CBUUIDCharacteristicFormatString as supported types. The iPhoneOS 27.2 SDK CBDescriptor.h also limits supported mutable descriptors to those types. This appears to exclude Report Reference. Is constructing and publishing 0x2908 in this way officially supported on iOS? If yes, please identify the supported OS versions, value format, restrictions, and an authoritative document or sample. Does spelling the assigned UUID in its full Bluetooth-base 128-bit form change the support contract? If it is unsupported, please confirm that runtime acceptance should not be relied on, and whether a documentation correction or API enhancement request is appropriate. Existing physical-device records from the full application include remote descriptor reads and notification subscriptions through an independent Central application. The owner also reports successful input, automatic reconnection and a 60-minute session after initiating a connection with nRF Connect. These are observations, not evidence of a supported API contract or a verified bonding sequence. They are not results from the attached reduced sample. The attached ReportReferenceProbe.swift isolates construction and an add-service request. It does not advertise or send input and is not a complete HOGP implementation. It has been compiled to an object file against the iPhoneOS SDK but has not been linked, installed or run on a device. No runtime error or success is claimed for it. A didAdd error would not, by itself, identify the descriptor as the rejected attribute because other required HID attributes are omitted. A successful callback would not prove official support or remote visibility. Please answer the API-support question independently of our observed interoperability. We are not asking for private APIs or an App Review approval prediction. Attachments: ReportReferenceProbe.swift, README.en.md, ENVIRONMENT.en.md. References: https://developer.apple.com/documentation/corebluetooth/cbmutabledescriptor https://developer.apple.com/documentation/corebluetooth/cbmutabledescriptor/init(type:value:)
1
0
33
22h
Healthkit forces 30 days of data or all data
I feel like this should be a bug report but at the very least it is a bad design decision. Confirmed by this documentation Authorizing access to health data Apple is now only allowing the backfill of either 30 days or all data. This is ridiculous as this amounts to all or nothing. We used to sync 90 days worth of data as that was enough to establish trends from the start, 30 days wasn't enough. Now our only option is to backfill years worth of data that is going to bloat our database unacceptably. Could somebody please explain to me the logic behind this? We should absolutely have the ability to choose the backfill window that meets our requirements. My app has been doing the initial backfill for over an hour now and I am storing 100Ks of rows of data I don't want.
1
0
99
23h
APNs delivery when the device has both cellular and Wi-Fi: which connection gets a pending push?
We're investigating alert pushes (priority 10, mutable-content, short expiration) that occasionally arrive 2-4 minutes late, usually around a cellular -> Wi-Fi handover. The provider gets HTTP 200 immediately. What we've observed with the APNs logging profile: apsd logs mention "dual channel is enabled" and "Connected on N interfaces", and per-interface states for WWAN and NonCellular. When we silently break the device's push connection (no RST), the device notices via a keep-alive timeout, reconnects (falling back to 443 when 5223 is blocked), and pending pushes arrive within ~0.3s of the new connection. In the field, a push sent later can arrive first: push 1 is held ~2-3 minutes while push 2, sent ~60s after it, arrives over Wi-Fi in ~2s. Questions: Can a device hold APNs connections over cellular and Wi-Fi at the same time, and if so, how does APNs choose which one to deliver a push on? If a push was sent on one connection and isn't acknowledged, does APNs keep retrying on that same connection until it times out, or does it move the push to another live connection for the same device? What is the retry/backoff behaviour for an unacknowledged push (intervals, any jitter, the timeout before it's re-queued), and does that schedule reset when the device opens a new connection? On the device side, how does apsd back off between reconnect attempts after consecutive failures, and can anything (for example network changes) cut that wait short? Anything a provider can do to reduce this (for example apns-expiration choices), or is this expected best-effort behaviour? Separate from the questions: I've been genuinely impressed by how resilient the push connection is. Even when I silently killed the connection, the device noticed, fell back to another port, and APNs delivered the pending pushes within a fraction of a second of reconnecting. It's an amazing piece of engineering and I wish I got to see behind the curtain.
1
0
50
1d
Can an iOS CarPlay app reliably identify the active head unit to enable features on approved units only?
We’re developing, an iOS navigation app with CarPlay support, for a managed vehicle fleet. We want to enable certain in-app features only when the application is running on an approved physical CarPlay head unit. The allowlist needs to distinguish individual units, including several units of the same make and model. The same iPhone may be used in both approved and unapproved vehicles. This application already uses CPTemplateApplicationSceneDelegate to detect when its CarPlay scene connects or disconnects. The scene’s persistentIdentifier is scoped to the UIKit scene session. We’re looking for a stable identity that we can reliably associate with the head unit hosting the current CarPlay session. Is there a supported iOS API or Apple-approved entitlement that lets a third-party navigation app identify the currently active physical CarPlay head unit? Ideally, the identifier would persist across reconnects and be available over both wired and wireless CarPlay. If this requires manufacturer cooperation, is an iAP2 protocol exposed through ExternalAccessory and EASession the recommended approach? Can the app reliably determine that the EAAccessory it is communicating with is the same head unit hosting the current CarPlay session, or would the manufacturer need to provide that binding in firmware? Can an existing third-party navigation app use this integration with an OEM partner, or is it reserved for apps published by the vehicle manufacturer? Which entitlement or Apple program would apply? The use case is to gate specific features inside this application while keeping the standard CarPlay experience available. We can request the required entitlements and work with the hardware vendor if needed. Any guidance on the supported integration path or the appropriate Apple team to contact would be appreciated.
1
0
54
1d
How does the driverkit serve as a bridge between hardware and the system?
I want to use DriverKit to develop a USBDriver, which serves as a bridge between USB devices and the system. All messages between USB devices and the system will be forwarded through the USBDriver. Can anyone give me some tips or suggestions? What API should I use? I couldn't find anything like this in the documentation or sample code. class MyUSBDriver: public IOUserClient { public: virtual bool init() override; virtual kern_return_t Start(IOService * provider) override; virtual kern_return_t Stop(IOService * provider) override; virtual void free() override; virtual kern_return_t GetRegistryEntryID(uint64_t * registryEntryID) override; virtual kern_return_t NewUserClient(uint32_t type, IOUserClient** userClient) override; virtual kern_return_t ExternalMethod(uint64_t selector, IOUserClientMethodArguments* arguments, const IOUserClientMethodDispatch* dispatch, OSObject* target, void* reference) override; }; I am now able to retrieve the device descriptor in the Start method IOUSBHostDevice *device = OSDynamicCast(IOUSBHostDevice, provider); if (device) { const IOUSBDeviceDescriptor *deviceDescriptor = device->CopyDeviceDescriptor(); if (deviceDescriptor) { uint16_t idVendor = deviceDescriptor->idVendor; uint16_t idProduct = deviceDescriptor->idProduct; uint8_t iSerialNumber = deviceDescriptor->iSerialNumber; IOUSBHostFreeDescriptor(deviceDescriptor); } }
1
0
915
1d
StoreKit Sandbox returns 0 products after TN3186 checks — catalog association issue?
We are testing two auto-renewable subscriptions for a new iOS app. In Apple Sandbox, StoreKit 2 Product.products(for:) always returns an empty array. Legacy SKProductsRequest independently reports both identifiers as invalid. Environment: Physical iPhone, iOS 26.5 Xcode development build with automatic signing Storefront.current = USA (143441) canMakePayments = true No StoreKit Configuration file in the Run scheme Direct native StoreKit 2 diagnostics, without a third-party purchase SDK TN3186 checks completed: Explicit Bundle ID matches the App Store Connect app, installed app, and provisioning profile In-App Purchase is enabled for the App ID Developer membership, Paid Apps Agreement, banking, and tax are active One subscription group with monthly and yearly subscriptions is visible Pricing, localizations, and availability are configured Neither subscription is Rejected or Developer Action Needed The scheme is using real Sandbox data, not local StoreKit testing Results: StoreKit 2 batch query: rawCount = 0 SKProductsRequest: valid = []; both identifiers are invalid StoreKit 2 individual monthly query: 0 StoreKit 2 individual yearly query: 0 No exception; failure occurs during product discovery Related reports: https://developer.apple.com/forums/thread/839313 — TN3186 verified and USA storefront valid, but still 0 products; no replies https://developer.apple.com/forums/thread/841722 — Apple Commerce mentioned Rejected or Developer Action Needed states; ours have neither https://developer.apple.com/forums/thread/820656 — Apple DTS confirmed first subscriptions do not require review before Sandbox testing Apple Developer Support referred us to TN3186, and we verified every applicable item. The individual subscription detail route in App Store Connect also intermittently remains on a loading indicator although the group and both records are visible. Could an App Store Commerce or StoreKit engineer check whether the app-to-IAP catalog association is populated correctly on Apple's side, or advise an official escalation path to have the Sandbox catalog reprocessed? Exact account, app, product, device, email, and support-case identifiers are omitted from this public post and can be supplied privately through Apple Support or Feedback Assistant.
1
0
511
1d
Supported third-party integration for system Apple Account sign-in on macOS
We are assessing an independent macOS client for user-owned physical Macs. The requirement is to initiate and complete a real system Apple Account sign-in, including system account persistence and service activation. This is not Sign in with Apple for signing in to our own application. Our current test machine runs macOS 14.6.1 (23G93). We understand that observations on that version do not establish the supported API surface on current macOS releases. Could Apple clarify: Is there a documented third-party API or Apple-provided broker that supports this workflow? If so, what are its supported OS versions, caller requirements, and completion/result contract? If this requires an approved capability or a specific program, is there a published eligibility and application channel? We are not assuming that ordinary Developer Program membership grants private system-account privileges. If the only supported approach is a handoff to System Settings, can you confirm that boundary and identify any documented completion/status mechanism? A UI-only handoff would be a separate integration and would not fulfill our independent-client requirement. We are asking about supported integration options, not implementation details of internal APIs. We do not have a separate approved backend or a developer-program team at present, and would like to establish feasibility before purchasing membership or attempting account writes. No credentials, account identifiers, tokens, device materials, host addresses, or binary dumps are included.
Replies
0
Boosts
0
Views
16
Activity
8h
Supported non-admin service identity for a macOS LaunchDaemon
I'm developing a private macOS application on an Apple-silicon Mac mini running macOS Golden Gate 27.2. I'm evaluating a system LaunchDaemon running under a dedicated, non-administrator identity that cannot log in interactively. A separate desktop application would communicate with the daemon through authenticated local IPC. Before implementing this architecture, I would like to establish the officially supported approach. My main question: What is the Apple-supported method for creating and maintaining a dedicated, non-administrator, non-interactive service identity for a system LaunchDaemon on macOS 27.2? In particular: Is there a documented procedure for creating, restricting, disabling, recovering and eventually removing this identity? Are dscl, sysadminctl or another account-management mechanism supported for this purpose? Can a LaunchDaemon run under this identity without an interactive user login, including when registered through SMAppService? Are there any secure-token, FileVault, volume-ownership or keychain restrictions that would affect this design? If a dedicated non-login service identity is not a supported architecture, what alternative does Apple recommend? I'd particularly appreciate links to official Apple documentation or guidance from Apple engineers. Thank you.
Replies
10
Boosts
0
Views
558
Activity
8h
Sandbox restriction blocks Keychain access during Live Activity rendering: supported design for encrypted updates?
Hi, I'm building a Live Activity prototype using fictitious data. I want incoming updates to stay encrypted until they reach my iPhone, without putting the decryption key in ActivityAttributes, ContentState, or archived views. My goal is timely updates without an alert for each change. I'm using Xcode 27.2 and the iPhoneOS 27.2 SDK. My iPhone was running iOS 27.2.1 when I last checked. I haven't recorded its exact OS build or the OS version and build for every earlier test. My app and WidgetKit extension have the same shared Keychain access group in their signed entitlements. My app creates a 32-byte test key as a generic password with kSecAttrAccessibleAfterFirstUnlockThisDeviceOnly, with synchronization disabled. The extension only reads the key. A failed lookup never causes the key to be replaced or deleted. When I try to read the key while rendering the Live Activity, the first, metadata-only SecItemCopyMatching call returns errSecNotAvailable (-25291). It fails before I request any secret bytes. I reproduced this after stopping and restarting the activity with my iPhone unlocked. I get the same result with a non-interactive LAContext and with kSecUseAuthenticationUIFail without a supplied context. I used the latter only for comparison, since it is deprecated. My query specifies the class, fixed service and account, and shared access group. It also sets kSecUseDataProtectionKeychain. For the metadata lookup, I use kSecAttrSynchronizableAny, kSecReturnAttributes: true, and kSecMatchLimitAll, without requesting data or item references. I only request the key bytes after finding exactly one item and validating its policy, so that second step is never reached here. For comparison, I added a regular widget to the same signed WidgetKit extension. Its timeline provider successfully validates the metadata using the same read-only query and client, while the Live Activity still returns errSecNotAvailable. I captured these results separately; I haven't established whether they ran simultaneously or in different processes. I have only confirmed metadata access in the regular widget, not access to the key bytes. I also tried decrypting in my foreground app and then passing the fictitious summary to ActivityKit through a local update. That renders correctly on my iPhone. It doesn't test remote delivery or background decryption, or fix the Keychain failure during rendering. I repeated the test with debug messages enabled in Console. The Live Activity in my diagnostic build still shows the same Keychain failure. The capture from the extension now includes the connection error, not just the generic securityd message: activating connection: ... name=com.apple.securityd failed to do a bootstrap look-up: xpc_error=[159: Unknown error: 159] invalidated after a failed init got event: Connection invalid Failed to talk to securityd after 4 attempts. error:[-25291] Error Domain=com.apple.security.xpc Code=3 Connection invalid - Connection init failed at lookup with error 159 - Sandbox restrictionThis excerpt omits connection addresses and account fields. The same -25291 error and sandbox-restriction reason appear five times in this capture. It establishes the reported reason for the connection failure in this test; I haven't established whether this is an intended restriction, an OS issue, or a configuration issue. I also don't infer the actual retry count from the generic message. I'd appreciate guidance on these questions: Given the explicit sandbox-restriction error for com.apple.securityd, is non-interactive Keychain access intentionally unavailable during Live Activity rendering, or should I investigate an OS or configuration issue? Displaying live data with Live Activities documents a separate sandbox and restrictions on network and location access, but I haven't found its Keychain access contract. How should I interpret the different result from my regular widget's timeline provider? For encrypted content-state updates, is there a supported execution point where my app can authenticate and decrypt each update before handing a local render value to ActivityKit? I found the documented app wake-up for push-to-start, but not an equivalent for ordinary updates. If neither approach is supported, how can I deliver timely, alert-free encrypted Live Activity updates while keeping the key solely in Keychain and out of archived state and views? How does ActivityKit protect static ActivityAttributes at rest? Can those attributes be included in backups or replicated to an Apple Watch or Mac, separately from rendered views? Does any of that data pass through Apple servers, and is there a public API to opt out of replication? Would a per-activity session key in locally created static attributes be a supported design, or does Apple advise against storing secrets there? I haven't implemented this alternative. I would keep the long-term pairing key in Keychain, exclude the session key from every APNs payload, and start the activity locally rather than through push-to-start. I'd like to understand the storage and replication guarantees before considering it further. I'm looking for a supported design without weakening Keychain protection, allowing authentication UI during rendering, or suppressing an ordinary notification to bypass its requirements. I can prepare a minimal reproducer with fictitious data and no tokens, real identifiers, or conversation content. Thanks for your help.
Replies
3
Boosts
0
Views
105
Activity
9h
macOS: pre-established stop mechanism when a VM supervisor stops progressing
I am evaluating a bounded workload host using Virtualization.framework on Apple silicon. This is a question about supported architecture and API contracts, not a report of a reproduced bug. No particular macOS release has yet been qualified for this design; please identify the releases and prerequisites to which an answer applies. I have read the discussions about cleanup after the VM-owning application terminates: https://developer.apple.com/forums/thread/848542 https://developer.apple.com/forums/thread/850709 I understand that this cleanup is different from an externally observable, time-bounded teardown guarantee. I am not asking again for a replacement application's handle to a VM whose owner has already died. The unresolved case is different: A host application owns a VM, but its serial VM-control queue is no longer progressing. A separate supervisor is intended to initiate the fallback stop action against that application. An external stop trigger occurs. The supervisor is then lost or stops progressing before it enters the signal operation; the VM-owning application remains alive. Loss of the supervisor role does not necessarily mean that its process has exited, so a process-exit notification cannot be assumed. Complete host or kernel failure is not part of this scenario. Is there a supported public mechanism that can be established before workload preparation and retain authority to act on the same target application after this loss, without requiring another action by the supervisor or progress on the VM-control queue? For any such mechanism, I would particularly appreciate clarification of: How its authority is bound to the intended target during incomplete setup, including process creation that has begun but has not returned a usable target identity. Assuming an already-running watchdog leaves this setup boundary unanswered. Which actor performs the remaining action when the supervisor cannot, and what lifecycle, privilege and progress assumptions that actor requires. Whether any documented progress bound applies when a required operation does not return. The design requirement is completion of the whole prescribed stop/escalation attempt, including required failure-handling steps, within 10 seconds of the same first external trigger. Starting a timer or entering a signal call is not equivalent to completing that attempt. This is not a request for unconditional successful process or VM termination. A supported narrower mechanism, or an explicit statement that macOS provides no such contract for this case, would also be useful. I am looking for public interfaces and their documented limits, not private helper-process identities or kernel implementation details. Thank you.
Replies
5
Boosts
0
Views
91
Activity
9h
StoreKit 2 transaction verification fails during App Review with localized error 「資訊對裝置無效」; how should our app handle this?
We are implementing in-app purchases using StoreKit 2. The issue we need guidance on is how our app should correctly handle a StoreKit 2 transaction that becomes .unverified during Apple App Review. In our local development testing, purchases work correctly. When we test the same IAP products locally, the transaction returned from Product.purchase() is .verified, and the purchase flow completes successfully. However, during Apple App Review, the reviewer appears to encounter a failure where the StoreKit 2 transaction is returned as .unverified on the client side. The original localized error message shown is in Traditional Chinese: 「資訊對裝置無效」 Our English translation of this message is: “The information is not valid for this device.” Because the message is localized, we are not sure what the exact original English StoreKit error string would be. This failure happens in the client app before our backend receipt or server-side validation flow. We are not using verifyReceipt for this step. The failure occurs at the StoreKit 2 VerificationResult level when checking the transaction returned by StoreKit. Our current implementation only accepts .verified transactions. If StoreKit 2 returns .unverified, we treat the transaction as failed and do not unlock the purchased content. We believe this is the recommended secure behavior, but we would like to confirm the correct handling, especially because this issue is appearing during App Review while our local development purchases succeed. Additional context: • Distribution/review context: App Review / IAP review • API: StoreKit 2 • Failure point: VerificationResult returned as .unverified • Localized error message: 「資訊對裝置無效」 • Rough English translation: “The information is not valid for this device.” • Local development testing: purchases return .verified and succeed • The issue occurs before backend validation • Our app currently rejects .unverified transactions and does not unlock content Questions: What are the known causes of StoreKit 2 returning .unverified with the localized error message 「資訊對裝置無效」 during App Review? Is our current behavior correct: only unlock content for .verified transactions and reject .unverified transactions? If App Review receives .unverified, what should the app do from Apple’s recommended perspective? Should we show a retry/error message, ask the reviewer to retry, refresh entitlements, or take another action? Is there any App Review environment, device configuration, Apple ID configuration, or StoreKit account state that can cause this device-related verification error? Since local development purchases succeed but App Review fails, what diagnostics or logs should we collect and provide to identify the root cause? Should we ever bypass StoreKit 2 transaction verification in App Review/TestFlight, or should .unverified always be treated as a failed transaction? We would appreciate guidance on the correct code-level handling so that our app remains secure while also passing App Review reliably. Thank you.
Replies
2
Boosts
0
Views
560
Activity
9h
Full Disk Access on macOS 27.3
I'm looking into the changes being implemented and I find this way of handling by Apple on short notice problematic. Now our customers (those that don't have MDM) will have to jump through hoops to enable FDA for our product, a security product that also uses Endpoint Security. We already have the entitlement for endpoint security, which requires FDA in order to work, so why make it so difficult instead of just allowing FDA for apps that have this entitlement? Why only backup applications are exempted?
Replies
2
Boosts
0
Views
87
Activity
10h
Access to related files outside Sandbox with NSFileCoordinator and NSFilePresenter does not work anymore in macOS 27.3 beta 2
In macOS 27.3 beta 2, it is no longer possible to use NSFileCoordinator with NSFilePresenter instances to open a SQLite file with related WAL and SHM files when a sandboxed app has access to the primary SQLite file. In macOS 27.3 beta 1 and previous macOS releases, the Related File Access API works as expected. This happens regardless where the files are located, so it seems to be unrelated to recent changes in Privacy & Security, although it could be an unexpected side effect of these changes. Instead of correctly registering NSFilePresenter instances, errors like the following are logged in the XCode console: NSFileSandboxingRequestRelatedItemExtension: Failed to issue extension for /Users/testuser/Desktop/Sample/Example.sqlite-wal because: Error Domain=NSPOSIXErrorDomain Code=1 "Operation not permitted" (The related item is protected by a different storage class than the primary item) UserInfo={NSDebugDescription=The related item is protected by a different storage class than the primary item} +[NSFileCoordinator addFilePresenter:] could not get a sandbox extension. primaryPresentedItemURL: file:///Users/testuser/Desktop/Sample/Example.sqlite, presentedItemURL: file:///Users/testuser/Desktop/Sample/Example.sqlite-walThe mentioned Related File Access API is documented here. I have filed a Feedback, which includes a sample project: FB25625743 My question is: Is this a bug, or new intended behavior, or is there something else changed that should be dealed with when using this API in macOS 27.3 beta 2?
Replies
1
Boosts
0
Views
59
Activity
10h
Can macOS verify a worker’s effective sandbox policy before it runs?
I’m building a local app on an Apple silicon Mac. Before a worker or any of its child processes can run, I need to verify the worker’s effective filesystem and network restrictions, including inherited restrictions. I need to allow that same pending launch only if the policy matches, and deny it if verification fails or times out. I reviewed Apple’s App Sandbox guidance and the Endpoint Security documentation for exec authorization and entitlement metadata. I haven’t found documentation showing that these expose the complete effective policy for the pending launch. Is there a supported public macOS API that provides this guarantee? If so, which API and macOS versions support it? If not, is this outside what the public APIs can verify?
Replies
1
Boosts
0
Views
184
Activity
10h
Subject: Critical APFS Data Loss Bug & Error 1110 Loop during official Finder Update on iPhone 14 (A15)
Hello, I am requesting an engineering escalation regarding a critical bug in Apple's APFS update pipeline that trapped an iPhone 14 (A15, 128 GB) in recovery mode during an official Finder update from iOS 16.6 to iOS 18 (27.2.1, 24B156) on macOS 14.5. The device hardware is 100% healthy (Apple MRI diagnostics all green), the passcode is known, and 101,175 MB (341,967 files) of user data remain intact on the flash storage. However, the device cannot boot or complete an update due to a corruption caused by Apple's own restore tools. --- ### What Happened (Verified in Logs): 1. **Initial Full Storage State:** The device had ~105 GB utilized and was updated via Finder using "Update" (not Restore). 2. **Destructive fsck Behavior:** In pass 3, the restore ramdisk ran `/sbin/fsck_apfs -dy /dev/disk2`. On the nearly full container, fsck erroneously flagged 8,182 live omap mappings as "orphans" and deallocated them, causing extent-ref b-tree node `0x10bb3` to be zeroed out (TRIMmed by the flash controller). 3. **Split Failure:** In pass 4, the iOS 18 System volume installed and sealed, but `split_data_volume` failed in 4 seconds when `apfs_migrator` encountered block `0x10bb3` (`obj_read` checksum failure returning EILSEQ / error 92). 4. **False Error 1110 Loop:** Every subsequent attempt now halts immediately at `space_checks` with Error 1110. --- ### Reverse Engineering Proof: Disassembly of `restored_update` (arm64e, VA 0x10001afcc / 0x10001b0e8) confirms a major error-handling bug: When `splitter_dump` encounters the unreadable block `0x10bb3`, it fails with `EILSEQ` (92). Instead of handling this error, `restored_update` executes: `mov x19, #-1; mov x0, x19; ret` The caller receives `-1 bytes of free space`, logs: "Data volume requires splitting. Splitter will leave -1 bytes of free space after completion. Current free space on device: 6620069888 bytes. Insufficient space on device to run splitter" and aborts with **CHECKPOINT FAILURE:(FAILURE:1110)**. The device is NOT out of space (it has 6.3 GB free); it is blocked solely because `restored_update` misinterprets an EILSEQ filesystem error as negative free space. --- ### Current State & Request: - The encryption keys and SEP/gigalocker are intact (`gigalocker: ONLINE`). - In the restore ramdisk, `/dev/disk2s2` mounts cleanly as read-write on `/mnt2`. - Zero APFS snapshots remain on the volume, meaning the extent-ref tree is theoretically fully reconstructible from the valid fsroot inode tree. I am requesting assistance from CoreOS / APFS engineering: 1. Is there an engineering update ramdisk or build that can bypass the `splitter_dump` EILSEQ check, or rebuild the extent-ref tree, so the update can finalize? 2. Alternatively, can an engineering liaison facilitate a non-destructive extraction given that the volume mounts read-write and the user passcode is known? I can immediately provide full terminal transcripts, Finder updater logs (`iPhoneUpdater-*.log`), and disassembly details. Thank you, Samuel Journo
Replies
1
Boosts
0
Views
97
Activity
11h
macOS APFS: com.apple.macl and UF_TRACKED appear after creating a hard link — safe file replacement and rollback?
I'm investigating an unexpected metadata change during a controlled file deployment on macOS. Environment and observed state Two regular configuration files are used by a local service. Before replacement, the first file had: SHA-256: verified. File flags: 0. Extended attributes: none. Hard-link count: 1. Owner/group: 501/20. Permissions: 0644. ACL: empty. A backup procedure created a hard link to preserve the original inode before replacing the active pathname with an independently created file. Immediately after creating and verifying the hard link, the original file was observed with: The same SHA-256 and inode. Flags: 64 (UF_TRACKED). Extended attribute: com.apple.macl. Hard-link count: 2. No file content was replaced. The procedure stopped because the metadata no longer matched the original baseline. Additional observations Other backup methods also encountered unexpected metadata changes during destination creation. Tests using renamex_np with RENAME_SWAP successfully exchanged and restored artificial files with flags 0 and empty extended attributes. A test with artificial files already containing com.apple.provenance also succeeded. We could not reproduce the specific UF_TRACKED + com.apple.macl + two-hard-link state in an isolated fixture using ordinary documented operations. The operational original and its second hard link currently share the same inode. No attempt has been made to remove the protected attribute or change its flags. Questions for Apple filesystem/security specialists Under what documented circumstances can creating a hard link or related Finder/TCC document-tracking behavior cause UF_TRACKED and com.apple.macl to appear on an existing file? Are these metadata changes expected for a file accessed through a user-authorized Documents location? What supported mechanism can preserve the current original inode and provide a verifiable rollback when replacing its active pathname? Is renamex_np(..., RENAME_SWAP) appropriate for this situation, and what special considerations apply to the existing second hard link and protected extended attributes? What verification criteria would you recommend before permitting a production replacement? I am looking for a documented, reversible procedure that preserves macOS security protections. I do not want to disable SIP, forcibly remove com.apple.macl, or bypass TCC. I can provide sanitized metadata comparisons and a minimal reproduction if needed.
Replies
1
Boosts
0
Views
138
Activity
11h
watchOS — Continuous Core Motion monitoring in the background for independent fall detection
I'm developing an independent fall-detection application for Apple Watch using Core Motion (CMMotionManager) at 20 Hz. The goal is to identify potential falls using accelerometer, gyroscope and orientation data, then notify nominated family contacts. The application is deliberately independent of Apple's built-in Fall Detection and Emergency SOS. Test environment Physical Apple Watch running watchOS 27.2 Xcode 27.2 Core Motion device-motion callbacks on a dedicated serial OperationQueue Motion processing on the MainActor Tests performed without the Xcode debugger attached Observed results With the Watch display awake: 2,300 processed samples Longest sensor timestamp gap: 0.05 seconds Longest callback arrival gap: 0.66 seconds Callback interruptions over one second: 0 With the Watch dimmed/backgrounded: 3,229 processed samples Longest sensor timestamp gap: 0.30 seconds Longest callback arrival gap: 29.11 seconds Callback interruptions over one second: 6 Maximum sample delivery lag: 29.065 seconds The motion timestamps appear nearly continuous, but callback delivery becomes substantially delayed after the app enters the background. The monitoring flag remains true, and no explicit call to stop motion updates occurs when the app becomes inactive. Questions Is there an Apple-supported background execution mechanism or entitlement allowing a third-party watchOS app to continuously process accelerometer and gyroscope data for independent fall detection, without requiring an active workout session? Would CMBatchedSensorManager provide any useful background delivery capability for this application, or would its callbacks remain subject to normal watchOS suspension? Is there a recommended architecture for detecting a potential fall and initiating a timely family notification while the Watch app is backgrounded, without relying on Apple's built-in Fall Detection or Emergency SOS? We understand that workout sessions and extended runtime sessions have specific intended purposes. We want to implement this correctly rather than use an unrelated background mode to prevent suspension. Any guidance from Apple engineers or developers experienced with watchOS background Core Motion processing would be greatly appreciated.
Replies
2
Boosts
0
Views
98
Activity
12h
NEURLFilterManager: inconsistent per-domain blocking (some sites never block, one flaps) - confirmed PIR database is correct
Hello, Some context: DropBet is an app that helps people with gambling addiction stay off gambling sites and apps, using NEURLFilterManager/PIR to block a large list of known gambling domains system-wide. For a lot of our users this filter is a core part of their recovery > reliability really matters here. Since fixing an unrelated HPKE encryption bug (separate thread, now closed), the filter works end-to-end for the first time, but blocking is inconsistent per domain in a way I can't explain from my side: winamax.fr and the Winamax app: blocked reliably, every single time > clean "connection failed" style block, confirmed to load fine with protection off. unibet.fr and betclic.fr: never blocked, consistently > confirmed even after a full device reboot (rules out stale caching), and confirmed identical across both Safari and Chrome on iOS tested side-by-side (so not browser-specific). olybet.fr: flaps > the page reloads every 2-3 seconds with protection on, also identical across both browsers. Confirmed this only happens with the filter enabled (tested toggling protection off/on against the same page). unibet.fr auto-redirects into the native Unibet app via a Universal Link before the page loads, unlike betclic.fr which stays in the browser > different code path, possibly relevant. To rule out any issue on our end, I wrote a standalone test that queries our deployed PIR database shards directly - bypassing OHTTP, the app, and the device entirely - for all of the domains above plus their www. variants. All of them correctly return the blocked value, no exceptions, including unibet.fr and betclic.fr which never block on the actual device. So the database, the PIR protocol implementation, and the server are confirmed correct for every domain I've tested. This points at something in how NEURLFilterManager/ciphermld applies a correctly-received PIR verdict on-device - possibly related to query batching, an evaluation-key edge case, or Universal Link handling bypassing the filter before a web request is even made. Is there anything you can share about how the client applies results once decrypted, or known issues around any of this?
Replies
7
Boosts
0
Views
339
Activity
20h
StoreKit introductoryOfferEligibility: safe retry after JWS expiration
How should a server safely retire or renew an unused JWS for Product.PurchaseOption.introductoryOfferEligibility(compactJWS:) after cancellation or interruption? Our application allows a first-purchase discount once per application account across payment platforms. We bind the signed eligibility request to a verified Apple app transaction, product, and purchase attempt, persist the exact JWS, and reuse it for the same attempt. Once a paid purchase is verified, we consume the application's eligibility. We do not automatically issue another nonce after cancellation, a signature error, or an empty transaction history. Apple's JWS documentation describes a one-time nonce and expiration calculated from iat, and says not to include exp: https://developer.apple.com/documentation/storekit/generating-jws-to-sign-app-store-requests I could not find the lifetime or a supported retirement/renewal procedure for the introductory-offer-eligibility audience. Keeping an unusable authorization forever would prevent the customer from trying another checkout; replacing it without retiring the original might leave two redeemable authorizations. What we observed Environment: Xcode 27.2 (27B52), macOS 26.6.2, physical iPhone 16 with iOS 27.0, Apple Sandbox. Verify the Apple-signed AppTransaction and issue one eligibility JWS. Start the purchase using that JWS, then cancel the payment sheet (userCancelled). Terminate and relaunch the app without resetting purchase history or replacing the authorization. Retry the same product and Apple account using the byte-for-byte identical JWS. Complete the Sandbox purchase and verify the Apple-signed transaction on an isolated server. The successful verification happened about 164 seconds after issuance. This is one successful cancellation/retry observation, not a reproduced expiration failure or proof of a general validity/nonce-consumption rule. It also did not verify our intended introductory price. The purchase call uses the persisted value: let result = try await product.purchase(options: [ .introductoryOfferEligibility(compactJWS: existingJWS) ])Supported lifecycle questions What is the validity period derived from iat, including clock-skew allowances and any Sandbox/production differences? Which API or documented error can establish that a previously issued authorization is definitively unusable, rather than merely absent from transaction history? When is the nonce consumed, and what happens after userCancelled, interruption, pending approval, expiration, or an uncertain result? Is invalidOfferSignature specific enough to distinguish expiration from consumption, account mismatch, or a bad signature? Is it supported to keep the same nonce and all bindings, update only iat, and re-sign? If so, is one-time use enforced across every version and for what retention period? If renewal requires a new nonce, how can the prior authorization be invalidated first? Is there another recommended recovery flow that prevents duplicate redemption without indefinitely blocking checkout? DTS directed us to ask this lifecycle question on the forums. A sanitized minimal Xcode project was already supplied privately to DTS; it demonstrates the observed cancel/relaunch/same-JWS retry, not an expiration failure. We would appreciate a supported procedure or documentation reference.
Replies
1
Boosts
0
Views
76
Activity
20h
Supported publication of HID Report Reference descriptor 0x2908 on iOS
Supported publication of HID Report Reference descriptor 0x2908 on iOS We request clarification of the supported Core Bluetooth API contract for one specific call: publishing the Bluetooth Low Energy HID Report Reference descriptor (0x2908) using CBMutableDescriptor on iOS. Our iPhone application uses CBPeripheralManager to publish a Human Interface Device over GATT Profile (HOGP) service for an iPad. Its Report characteristics use two-byte Report Reference values containing the report ID and report type. The relevant call is: let reportReference = CBMutableDescriptor( type: CBUUID(string: "00002908-0000-1000-8000-00805F9B34FB"), value: Data([0x01, 0x01]) )The initializer documentation lists only CBUUIDCharacteristicUserDescriptionString and CBUUIDCharacteristicFormatString as supported types. The iPhoneOS 27.2 SDK CBDescriptor.h also limits supported mutable descriptors to those types. This appears to exclude Report Reference. Is constructing and publishing 0x2908 in this way officially supported on iOS? If yes, please identify the supported OS versions, value format, restrictions, and an authoritative document or sample. Does spelling the assigned UUID in its full Bluetooth-base 128-bit form change the support contract? If it is unsupported, please confirm that runtime acceptance should not be relied on, and whether a documentation correction or API enhancement request is appropriate. Existing physical-device records from the full application include remote descriptor reads and notification subscriptions through an independent Central application. The owner also reports successful input, automatic reconnection and a 60-minute session after initiating a connection with nRF Connect. These are observations, not evidence of a supported API contract or a verified bonding sequence. They are not results from the attached reduced sample. The attached ReportReferenceProbe.swift isolates construction and an add-service request. It does not advertise or send input and is not a complete HOGP implementation. It has been compiled to an object file against the iPhoneOS SDK but has not been linked, installed or run on a device. No runtime error or success is claimed for it. A didAdd error would not, by itself, identify the descriptor as the rejected attribute because other required HID attributes are omitted. A successful callback would not prove official support or remote visibility. Please answer the API-support question independently of our observed interoperability. We are not asking for private APIs or an App Review approval prediction. Attachments: ReportReferenceProbe.swift, README.en.md, ENVIRONMENT.en.md. References: https://developer.apple.com/documentation/corebluetooth/cbmutabledescriptor https://developer.apple.com/documentation/corebluetooth/cbmutabledescriptor/init(type:value:)
Replies
1
Boosts
0
Views
33
Activity
22h
Healthkit forces 30 days of data or all data
I feel like this should be a bug report but at the very least it is a bad design decision. Confirmed by this documentation Authorizing access to health data Apple is now only allowing the backfill of either 30 days or all data. This is ridiculous as this amounts to all or nothing. We used to sync 90 days worth of data as that was enough to establish trends from the start, 30 days wasn't enough. Now our only option is to backfill years worth of data that is going to bloat our database unacceptably. Could somebody please explain to me the logic behind this? We should absolutely have the ability to choose the backfill window that meets our requirements. My app has been doing the initial backfill for over an hour now and I am storing 100Ks of rows of data I don't want.
Replies
1
Boosts
0
Views
99
Activity
23h
APNs delivery when the device has both cellular and Wi-Fi: which connection gets a pending push?
We're investigating alert pushes (priority 10, mutable-content, short expiration) that occasionally arrive 2-4 minutes late, usually around a cellular -> Wi-Fi handover. The provider gets HTTP 200 immediately. What we've observed with the APNs logging profile: apsd logs mention "dual channel is enabled" and "Connected on N interfaces", and per-interface states for WWAN and NonCellular. When we silently break the device's push connection (no RST), the device notices via a keep-alive timeout, reconnects (falling back to 443 when 5223 is blocked), and pending pushes arrive within ~0.3s of the new connection. In the field, a push sent later can arrive first: push 1 is held ~2-3 minutes while push 2, sent ~60s after it, arrives over Wi-Fi in ~2s. Questions: Can a device hold APNs connections over cellular and Wi-Fi at the same time, and if so, how does APNs choose which one to deliver a push on? If a push was sent on one connection and isn't acknowledged, does APNs keep retrying on that same connection until it times out, or does it move the push to another live connection for the same device? What is the retry/backoff behaviour for an unacknowledged push (intervals, any jitter, the timeout before it's re-queued), and does that schedule reset when the device opens a new connection? On the device side, how does apsd back off between reconnect attempts after consecutive failures, and can anything (for example network changes) cut that wait short? Anything a provider can do to reduce this (for example apns-expiration choices), or is this expected best-effort behaviour? Separate from the questions: I've been genuinely impressed by how resilient the push connection is. Even when I silently killed the connection, the device noticed, fell back to another port, and APNs delivered the pending pushes within a fraction of a second of reconnecting. It's an amazing piece of engineering and I wish I got to see behind the curtain.
Replies
1
Boosts
0
Views
50
Activity
1d
Can an iOS CarPlay app reliably identify the active head unit to enable features on approved units only?
We’re developing, an iOS navigation app with CarPlay support, for a managed vehicle fleet. We want to enable certain in-app features only when the application is running on an approved physical CarPlay head unit. The allowlist needs to distinguish individual units, including several units of the same make and model. The same iPhone may be used in both approved and unapproved vehicles. This application already uses CPTemplateApplicationSceneDelegate to detect when its CarPlay scene connects or disconnects. The scene’s persistentIdentifier is scoped to the UIKit scene session. We’re looking for a stable identity that we can reliably associate with the head unit hosting the current CarPlay session. Is there a supported iOS API or Apple-approved entitlement that lets a third-party navigation app identify the currently active physical CarPlay head unit? Ideally, the identifier would persist across reconnects and be available over both wired and wireless CarPlay. If this requires manufacturer cooperation, is an iAP2 protocol exposed through ExternalAccessory and EASession the recommended approach? Can the app reliably determine that the EAAccessory it is communicating with is the same head unit hosting the current CarPlay session, or would the manufacturer need to provide that binding in firmware? Can an existing third-party navigation app use this integration with an OEM partner, or is it reserved for apps published by the vehicle manufacturer? Which entitlement or Apple program would apply? The use case is to gate specific features inside this application while keeping the standard CarPlay experience available. We can request the required entitlements and work with the hardware vendor if needed. Any guidance on the supported integration path or the appropriate Apple team to contact would be appreciated.
Replies
1
Boosts
0
Views
54
Activity
1d
How does the driverkit serve as a bridge between hardware and the system?
I want to use DriverKit to develop a USBDriver, which serves as a bridge between USB devices and the system. All messages between USB devices and the system will be forwarded through the USBDriver. Can anyone give me some tips or suggestions? What API should I use? I couldn't find anything like this in the documentation or sample code. class MyUSBDriver: public IOUserClient { public: virtual bool init() override; virtual kern_return_t Start(IOService * provider) override; virtual kern_return_t Stop(IOService * provider) override; virtual void free() override; virtual kern_return_t GetRegistryEntryID(uint64_t * registryEntryID) override; virtual kern_return_t NewUserClient(uint32_t type, IOUserClient** userClient) override; virtual kern_return_t ExternalMethod(uint64_t selector, IOUserClientMethodArguments* arguments, const IOUserClientMethodDispatch* dispatch, OSObject* target, void* reference) override; }; I am now able to retrieve the device descriptor in the Start method IOUSBHostDevice *device = OSDynamicCast(IOUSBHostDevice, provider); if (device) { const IOUSBDeviceDescriptor *deviceDescriptor = device->CopyDeviceDescriptor(); if (deviceDescriptor) { uint16_t idVendor = deviceDescriptor->idVendor; uint16_t idProduct = deviceDescriptor->idProduct; uint8_t iSerialNumber = deviceDescriptor->iSerialNumber; IOUSBHostFreeDescriptor(deviceDescriptor); } }
Replies
1
Boosts
0
Views
915
Activity
1d
StoreKit Sandbox returns 0 products after TN3186 checks — catalog association issue?
We are testing two auto-renewable subscriptions for a new iOS app. In Apple Sandbox, StoreKit 2 Product.products(for:) always returns an empty array. Legacy SKProductsRequest independently reports both identifiers as invalid. Environment: Physical iPhone, iOS 26.5 Xcode development build with automatic signing Storefront.current = USA (143441) canMakePayments = true No StoreKit Configuration file in the Run scheme Direct native StoreKit 2 diagnostics, without a third-party purchase SDK TN3186 checks completed: Explicit Bundle ID matches the App Store Connect app, installed app, and provisioning profile In-App Purchase is enabled for the App ID Developer membership, Paid Apps Agreement, banking, and tax are active One subscription group with monthly and yearly subscriptions is visible Pricing, localizations, and availability are configured Neither subscription is Rejected or Developer Action Needed The scheme is using real Sandbox data, not local StoreKit testing Results: StoreKit 2 batch query: rawCount = 0 SKProductsRequest: valid = []; both identifiers are invalid StoreKit 2 individual monthly query: 0 StoreKit 2 individual yearly query: 0 No exception; failure occurs during product discovery Related reports: https://developer.apple.com/forums/thread/839313 — TN3186 verified and USA storefront valid, but still 0 products; no replies https://developer.apple.com/forums/thread/841722 — Apple Commerce mentioned Rejected or Developer Action Needed states; ours have neither https://developer.apple.com/forums/thread/820656 — Apple DTS confirmed first subscriptions do not require review before Sandbox testing Apple Developer Support referred us to TN3186, and we verified every applicable item. The individual subscription detail route in App Store Connect also intermittently remains on a loading indicator although the group and both records are visible. Could an App Store Commerce or StoreKit engineer check whether the app-to-IAP catalog association is populated correctly on Apple's side, or advise an official escalation path to have the Sandbox catalog reprocessed? Exact account, app, product, device, email, and support-case identifiers are omitted from this public post and can be supplied privately through Apple Support or Feedback Assistant.
Replies
1
Boosts
0
Views
511
Activity
1d