Posts under App & System Services topic

Post

Replies

Boosts

Views

Activity

Supported App Sandbox metadata access through an inherited file descriptor, without reopening the object
We are evaluating a macOS design before implementation. This is a platform-support question, not a report of an observed runtime failure. No actual controller executable, helper executable or deployment configuration has been selected or qualified. Our objective is to let a small, fixed-purpose helper examine an already-open object’s identity and access control list (ACL) metadata while preserving the original object reference and ownership. Proposed configuration A trusted controller would launch one native C helper using an explicit executable target and posix_spawn. The controller would need to qualify for the documented nonsandboxed-parent case, without a conflicting inherited sandbox. We do not assume that an absent entitlement alone establishes this. The helper would be the main executable of its own signed application bundle, using its own ordinary App Sandbox rather than inheriting the controller’s sandbox. Its proposed additional filesystem permission would be com.apple.security.temporary-exception.files.absolute-path.read-only, containing exactly two fixed directory entries beneath a user’s development workspace, outside the helper’s container. The literal private paths are omitted. The controller would already own a read-only file descriptor for one regular file or directory within those trees. It would retain its original reference and pass a known duplicate referring to the same open object to the helper. The helper would perform one metadata observation. The proposed operations are fstat for object identity and fgetattrlist requesting ATTR_CMN_RETURNED_ATTRS and ATTR_CMN_EXTENDED_SECURITY, with FSOPT_REPORT_FULLSIZE and without FSOPT_PACK_INVAL_ATTRS. Required boundaries The helper must use only the inherited descriptor for the subject observation. The design does not permit: Reopening or resolving the subject through a pathname, bookmark or URL. Substituting a copied file or another object reference. Reading file contents, enumerating directories, changing permissions, or modifying shared descriptor flags or file offsets. Expanding the exception to an ancestor, home directory or broader workspace. Disabling an existing sandbox or using private interfaces to bypass restrictions. Parent-side traversal and identity checks remain separate responsibilities. Strict ancestor directories are not additional helper ACL subjects. We understand that a read-only directory exception and ordinary sandbox/container capabilities may permit more than the helper’s intended metadata operations. Metadata-only behaviour would be a reviewed-code obligation; we are not asking App Sandbox to enforce an exact metadata-only operation filter. Questions Supported launch and activation: Under what documented parent-context, signing, bundle, container and launch conditions can this helper establish its own ordinary App Sandbox? What supported mechanism or guarantee establishes that the sandbox is effective before the subject metadata operation? Required metadata permissions: Does the proposed read-only exception support fstat and the specified extended-security fgetattrlist request through an inherited read-only descriptor, for both regular files and directories? Please distinguish descriptor survival from permission to perform these operations, including any additional access-control conditions. Scope, identity and lifetime: What documented behaviour applies to descendant coverage, aliases, hard links, renames or namespace changes after the descriptor was opened? Please distinguish the lifetime of the inherited object reference from the lifetime of the static directory permission. Our design would refuse uncertainty rather than reopen, rebind or widen access. Evidence and support boundary: Which public documentation, interfaces or authoritative guarantees support this exact composition, and which parts require separate empirical qualification? Please identify applicable macOS versions and restrictions. Distinguish platform support from any distribution or App Store review requirements. If these requirements cannot be combined through supported interfaces, please identify the incompatible requirement. An approach requiring broader access, pathname reopening, a replacement object or a different handoff architecture would be a design change requiring separate review, not a solution within the stated constraints. Permission failures, unsupported requests and incomplete metadata would remain unavailable results. They would not be interpreted as proof that an ACL is absent. Complete ACL-absence interpretation and native resource bounds remain separate qualification questions. Apple material already reviewed Resolving App Sandbox Inheritance Problems App Sandbox Temporary Exception Entitlements Accessing files from the macOS App Sandbox
1
0
246
6d
macOS 26.x App Sandbox: supported disk-image attachment and descriptor-based unmount
On macOS 26.x, is there a public, supported way for an App Sandbox application with com.apple.security.virtualization to attach a new disk-image file held in its own container and mount its APFS filesystem within that container, without root, privileged or unsandboxed helpers, private/temporary exception entitlements, Full Disk Access, or weakening App Sandbox? We distinguish attaching the image from mounting an already attached virtual device. DTS threads 655767, 780965 and 775835 discuss these separately. Which supported API or tool and entitlements apply on 26.x, if such a path exists? If attachment is supported, what ownership and sandbox checks apply to a later normal funmount(fd, 0) using an O_EVTONLY descriptor for that exact mount's root? Is this combination supported, including when another actor has already unmounted it? If not, what supported alternative preserves the same restricted profile and avoids force or whole-disk unmount?
0
0
95
6d
NSE Filtering for encrypted incoming calls: what technical details should we provide?
We’re preparing the first iOS release of ANTINTWRK, an encrypted messenger with voice and video calls. Our push gateway sends a regular APNs alert for the encrypted event. On the receiving device, the Notification Service Extension fetches and decrypts the referenced event. In this delivery path, the gateway does not distinguish an encrypted call invitation from a message. When the extension identifies a ringing call, it uses CXProvider.reportNewIncomingVoIPPushPayload to hand it off to CallKit and suppresses the original alert to avoid a duplicate notification. Our Notification (NSE) Filtering request, submitted on October 25, 2026, still shows Submitted. Xcode cannot export the App Store build because the extension’s distribution profile lacks com.apple.developer.usernotifications.filtering. We have contacted Developer Support under Entitlements. We’re following Apple’s Sending End-to-End Encrypted VoIP Calls guidance. For this design, what technical details or test evidence should we provide so the capability review team can evaluate the request? Is there a recommended way to attach that material to the existing request? We can provide the request and support case IDs privately if needed.
0
0
146
1w
HKStatisticsCollectionQueryDescriptor intermittently returns no data for certain date ranges on iOS 27
We are seeing inconsistent results from HKStatisticsCollectionQueryDescriptor on iOS 27. Using the same quantity type, statistics options, anchor date, interval components, and predicate configuration, some date ranges return the expected statistics, while other ranges unexpectedly return empty results or buckets with no quantity. The affected ranges do contain HealthKit samples: HKSampleQueryDescriptor finds samples in the same date range. HKStatisticsQueryDescriptor returns the expected value when run separately for an affected bucket. HKStatisticsCollectionQueryDescriptor returns no quantity for that same bucket. Slightly expanding or shifting the date range may cause the collection query to return data again. A simplified version of the query looks like this: let datePredicate = HKQuery.predicateForSamples( withStart: startDate, end: endDate, options: .strictStartDate ) let descriptor = HKStatisticsCollectionQueryDescriptor( predicate: .quantitySample( type: quantityType, predicate: datePredicate ), options: .cumulativeSum, anchorDate: anchorDate, intervalComponents: DateComponents(day: 1) ) let collection = try await descriptor.result(for: healthStore) collection.enumerateStatistics(from: startDate, to: endDate) { statistics, _ in let quantity = statistics.sumQuantity() print(statistics.startDate, quantity as Any) } Expected behavior Every interval containing matching samples should return the corresponding statistics, regardless of the overall requested date range. Actual behavior Some date ranges produce missing or empty buckets even though matching samples exist and an individual HKStatisticsQueryDescriptor can calculate the expected value. Changing only the date range can make the data appear or disappear. The samples are visible to the app in the affected range, so this does not appear to be explained solely by iOS 27’s Limited History authorization. This behavior was not observed with the same query flow on earlier iOS versions. Is this a known regression in HKStatisticsCollectionQueryDescriptor on iOS 27, or has the expected date-range or predicate behavior changed?
4
3
1.2k
1w
CKKS zone deletions return success but stale 2019 key hierarchy persists; "Delete iCloud Data" not offered (FB25584356)
I filed FB25584356 for this and would appreciate any pointers. Setup: Mac mini on macOS 27.2.1 and an iPhone on iOS 27.2, same Apple Account, HSA2, Advanced Data Protection off. Octagon is healthy (otctl status = Ready, both devices trusted). Symptoms: AirDrop to my own iPhone fails when I pick the device (rapportd: "PairVerify server M2 failed: copy identity"), iPhone Mirroring ends with noCompatiblePhone, and various clients log CKErrorDomain 111 / 5009 "Keychain Sync Incomplete". What I see in CKKS: seven zones (AutoUnlock, LimitedPeersAllowed, Home, Manatee, ApplePay, Health, Engram) still hold a key hierarchy from August 2019. The only TLKShare was issued by a device that left the account years ago, so neither current device can unwrap the TLK ("No trusted TLKShares", CKKSErrorDomain 35). secd then does what I would expect: tlkmissing → "No other devices claim to have the TLK. Resetting zone..." → CKKSDeleteCKZoneOperation → local reset → zones recreated. The delete appears to succeed. With debug logging, cloudd shows status:200 and "RecordZone ... was successfully deleted from the server" for each zone. The next fetch returns the same 2019 records with an identical server change token. Over about 11 hours: 124 reset cycles, 869 zone deletions reported as successful, 182 consecutive AutoUnlock fetches with the same change token. When not throttled it runs roughly every 5 seconds; CloudKit eventually answers "Service Unavailable" (CKError 6/2009) with a ~59 minute retry-after, and all 23 zones are unusable in the meantime. ckksctl -v AutoUnlock reset-cloudkit behaves the same way. Reset path: Support article 120862 describes "Don't know passcode?" → "Forgot all passcodes?" → "More Options" → "Delete iCloud Data". With the Mac signed in but not yet trusted, "Don't know passcode?" only offered Cancel / Try Again Later, in the sign-in flow and in both System Settings follow-ups. After three wrong passcodes the alert had a single "Try Again Later" button, although cdpd logged "Creating recovery option: Reset". Questions: Is a client-side zone delete expected to be a no-op for these zones while the server reports success? Is there a supported way to reach performCKServerUnreadableDataRemoval / "Delete iCloud Data" when another trusted device exists? I am fine with losing the end-to-end encrypted data in those zones. Logs and a sysdiagnose are attached to the Feedback.
0
0
135
1w
Screen Time reaches the iPhone Instagram limit after Mac sync while Instagram is in the background (on both devices)
Hello, I have noticed, once again, that DeviceActivityMonitors are quite frustrating to work with, since they are triggering random events when using mulitple Apple devices. For example, if Instagram is open on some random (non-foreground!) tab in Safari on the Mac, the iPhone will reach its limit quite consistently & quickly atfer setting up a new device activity threshold. It would be great if A) background usage was not counted towards screen time: FB25576930 B) there was an API for us developers to set-up device-specific deviceactivitymonitors: FB24363336 Hope that helps!
0
0
102
1w
URGENT: Muse is AUTOMATICALLY blocked by Screen Time API whenever Facebook is blocked, 100% of the time with no workaround for users or developers
PLEASE ADDRESS ASAP: Muse is #1 on the App Store and Facebook is one of the most blocked apps for Screen Time. This is affecting a vast number of people. Description: When a user blocks the Facebook app in any screen time app, iOS automatically blocks Meta's Muse app too. Every single time. There is no way for the user or the developer to prevent it, allowlist it, or even detect that it happened. Muse shows Apple's generic blocked-app "Restricted" screen instead of the screen time app's own design, so the user cannot tell who blocked it or why. This is the worst possible combination of apps: Facebook is one of the most-blocked app across screen time apps. Muse ("Muse from Meta") is currently the #1 app on the App Store. A huge and fast-growing number of users block Facebook and silently lose the world's most popular new app. Muse launched Sept 8, 2026 and adoption is still climbing -- the affected population grows every day. There is NO workaround whatsoever: Users cannot allowlist Muse: it does not appear in Settings > Screen Time > Always Allowed. Developers cannot prevent it: the Screen Time API offers no opt-out; the extra block is applied by iOS itself. Developers cannot detect it: Muse's token (listed separately in FamilyActivityPicker) corresponds to no real identity anywhere else. Developers cannot customize it: their ShieldConfiguration is never consulted -- Muse gets Apple's generic shield UI. Users cannot even block Muse on purpose: shielding Muse's own token is a silent no-op. Screen Time usage statistics are also wrong -- Muse usage is recorded as "Facebook." This Screen Time bug renders screen time apps effectively unusable if users want to block Facebook and use Muse, which is extremely common. How can we get this addressed as soon as possible? Feedback Assistant ticket: https://feedbackassistant.apple.com/feedback/25019832
9
9
690
1w
Looks like the Home application breaks non-trivial predicates?
I've just bumped into a weird problem. I've created a pretty simple and perfectly working event trigger with a non-trivial predicate which is just an AND of two characteristic-value ones. Then I've renamed it in the Home application, and it stopped working. A dump suggests the cause is a broken predicate, for it looks now rather differently — one of those two characteristic-value predicates seems broken (details below). Is that possible? Could please someone check whether the problem is really a Home ugly bug, or whether I do something wrong at my side? Versions are macOS 27.0.1 (26A434) Home 11.0 (1241.1.7.4.3) I would gladly post a bugreport, but at this moment I am far from being sure I do not overlook something of importance myself. Thanks! My code — for clarity stripped of all the error checking and reporting — looks like this: HMHome *home=...; HMCharacteristic *knihovna=..., *stul=..., *sklenik=...; // light chars, found by names HMActionSet *switchOff=...; // trivial switch-off for another light, found by name HMEventTrigger *trigger=...; // found by name if exists if (trigger) { // if found, the predicate is logged out and that's all NSLog(@"=== predicate found: %@", trigger.predicate); return; } // otherwise, the trigger and predicate are created this way: HMCharacteristicEvent *evt=[[HMCharacteristicEvent alloc] initWithCharacteristic:sklenik triggerValue:@1]; NSPredicate *p1=[HMEventTrigger predicateForEvaluatingTriggerWithCharacteristic:knihovna relatedBy:NSEqualToPredicateOperatorType toValue:@1]; NSPredicate *p2=[HMEventTrigger predicateForEvaluatingTriggerWithCharacteristic:stul relatedBy:NSEqualToPredicateOperatorType toValue:@1]; NSPredicate *p=[[NSCompoundPredicate alloc] initWithType:NSAndPredicateType subpredicates:@[p1, p2]]; NSLog(@"=== predicate created: %@", p); HMEventTrigger *et=[[HMEventTrigger alloc] initWithName:@"TEST-ONLY" events:@[evt] predicate:p]; [home addTrigger:et completionHandler:^(NSError *error) { if (!error) [et addActionSet:switchOff completionHandler:^(NSError *error) { if (!error) [et enable:YES completionHandler:^(NSError *error) { if (!error) NSLog(@"=== success!"); }]; }]; }];As already mentioned, the trigger works perfectly (is trivial enough, would be weird if not). When created and also later, when found by name, the predicate looks like this: === predicate found: (characteristic == <HMCharacteristic, Unique ID: 7444CCCE-1C1B-5234-8CAE-01DA4D811B8B, Type: 00000025, Instance ID: 9> AND characteristicValue == 1) AND (characteristic == <HMCharacteristic, Unique ID: 2E046807-9127-565F-B9CB-10BC0F2EE6D2, Type: 00000025, Instance ID: 9> AND characteristicValue == 1)Then, I just go to the Home application and rename the trigger, nothing else. It immediately stops working properly, and from that moment on, the predicate looks like this — note the difference: === predicate found: characteristic == <HMCharacteristic, Unique ID: 2E046807-9127-565F-B9CB-10BC0F2EE6D2, Type: 00000025, Instance ID: 9> AND (characteristic == <HMCharacteristic, Unique ID: 7444CCCE-1C1B-5234-8CAE-01DA4D811B8B, Type: 00000025, Instance ID: 9> AND characteristicValue == 1)
7
0
375
1w
MacOS HomeKit private (and still useable) application
How do I write a macOS (no need for i/Pad/OS at all) application which supports HomeKit for my own private use only? My original idea was simply to write it non-provisioned, which would solve all problems perfectly. Nevertheless, that, triple alas, seems not possible (https://developer.apple.com/forums/thread/849288). Provisioned one built locally has a lot of problems, among which there are e.g., big hurdles to run it at my other Macs (https://developer.apple.com/forums/thread/848903?answerId=907702022#907702022) an extremely limited expiration date A reasonable solution would be a creation of a provisioning profile which would include an expiration date in a far future and also IDs of my other computers; alas, far as I understand, this does not seem to be possible with a Personal Team, as detailed in https://developer.apple.com/forums/thread/848903?answerId=907568022#907568022. Whilst a sort of solution might be the nuclear one, it does not feel right the slightest — I need just to run my own app, nothing more, and whilst naturally I gladly accept all the possible dangers of code I wrote myself, I definitely would not want to get completely vulnerable where all the other apps (and other threats) in the world are concerned. I've checked also the Unlisted distribution; far as understand it properly, won't do either, e.g., since it will be declined ... if your app is in a beta or prerelease state — the application is fully intended to be used locally for months or even years in this state. Besides, even Unlisted apps must go through Review, and when I bump to a need to make any kind of change, I want simply do that locally and launch my new version immediately without any Review hassle. Also, in this case the otherwise important privacy issues like e.g., determining the proper computer name (https://developer.apple.com/forums/thread/813853?answerId=874075022#874075022) are of absolutely no importance. Are there other, more reasonable options? What's the proper solution, after all? Thanks!
4
0
629
1w
MacOS HomeKit non-provisioned application
How to write a macOS (no need for i/Pad/OS at all) application which supports HomeKit without provisioning? Might be a non-GUI plain tool or essentially anything runnable, if it makes the solution easier. All it ever needs is access to HomeKit (in future probably also Matter) APIs and the filesystem (to write logs). The application will be used locally and privately only. Thus, the provisioning is definitely not needed, and it brings ugly problems (compare please the thread https://developer.apple.com/forums/thread/848903). Thanks!
5
0
626
1w
Develop DriverKit extensions locally with a Free Developer Account
I have not seen a clear answer about this: https://developer.apple.com/forums/thread/776657 https://developer.apple.com/forums/thread/743639?answerId=775701022#775701022 Thus, I would like to pose: Is there any way to install a DriverKit extension locally for development and testing using a free developer account, or is development of DriverKit extensions gated behind a paid developer account? Thank you!
8
0
945
1w
Supported vnode metadata access after same-UID libproc queries return EPERM
We need a documented interface/access contract for a test supervisor that requires a bounded, process-birth-bound census of every process with its effective UID and every current vnode descriptor of those processes. Coverage includes ambient same-UID processes and descriptors opened before observation starts. The census needs vnode identity, type, ownership, mode, link count and open flags. Unavailable metadata is a refusal; we cannot treat EPERM as absence or exempt a process or descriptor. This is a cooperative same-user lifetime-containment check. We do not require file contents or an atomic global snapshot. Environment of the retained observation: Apple Silicon, macOS 27.3 (26C5035r), Darwin 27.3.0; Xcode 27.0 (27A5194q), macOS 27.0 SDK. The subject was the system LaunchServices process /usr/libexec/lsd. Caller real/effective UID matched the subject's UID/RUID/SVUID. Before/after process identity, version, start time, credentials and executable-path samples matched. PROC_PIDLISTFDS returned 32 bytes and listed FD 3 as PROX_FDTYPE_VNODE. The following call shape returned 0 bytes with errno == EPERM (1), with 176 bytes expected in that environment: struct vnode_fdinfo info = {0}; errno = 0; int count = proc_pidfdinfo(pid, fd, PROC_PIDFDVNODEINFO, &info, sizeof(info)); int saved_errno = errno;Errno was cleared before and saved immediately after the call. Credentials were not modified. The matching process bracket does not establish an FD-to-path association or a specific protection category. In particular, we have not established a Data Vault subtype. Questions: Is there a supported public interface or provisionable entitlement that supplies the required metadata for every current same-UID vnode descriptor, including descriptors whose metadata is protected internally by macOS? If Endpoint Security is the supported route, which documented API supplies a complete current FD inventory and vnode mapping, including already-open descriptors and ambient same-UID processes? A solution limited to descendants of the client does not cover this requirement. If supported interfaces cannot guarantee that coverage, could you clarify the limitation? Is there a supported diagnostic that identifies the protection category responsible for this metadata denial without inspecting private policy state? I have considered the "On File System Permissions" guidance, including its mandatory-access-control and Endpoint Security discussion: https://developer.apple.com/forums/thread/678819 This post describes a retained observation, not a new reproduction or qualification result. All retained results remain UNPROVED. A focused SDK/libproc Xcode example was previously built for arm64 and x86_64 but was not run; it illustrates the call sequence and a BSD start/credential bracket, not complete kernel birth binding or the required census. It does not guarantee reproduction on another system. The sanitized supporting record and example are already attached to the private DTS request that directed me to the forums. Please advise the supported interface/access contract, or its limitation, and any further diagnostic scope needed.
1
0
109
1w
[iPadOS] USB 4 port unusable after driver uninstallation
I have developed a Driverkit extension which matches for iPadOS which is bundled within my app. My driver contains two different personalities - one with provider as USBDriverKit (USB3) and the other as PCIDriverKit (USB4). When enabled, the driver matches/loads, initiates the class with the right personality based on whether the device it matches is operating in USB3/USB4 and works as intended. Recently, we have come across an issue when operating in USB4 and below are the steps to reproduce. Platform: iPadOS 17 (Issue not observed following the same steps in iPadOS 18 and higher, though cannot be full sure as I have tested only in some versions of 18, 26 and 27). Tested in iPad having a USB-4 compatible port. Steps to reproduce: Install the app. This provides an option to enable the driver which comes with it from the app's system settings. Enable the driver from settings. Connect a device which is compatible with the driver in USB4 so that the specific personality can be loaded. Verify the same. With the driver still loaded, uninstall the application. On uninstallation, ideally, Apple's inbox IONVMeController should match on the re-enumeration and get loaded without a re-plug being required. However, this is not the case as the port becomes unresponsive. Even after a re-plug of the device, it still does not get detected by the iPad itself. This is specifically when I try to connect the device in USB 4 as this specific driver personality matches only in this case. If I try the same steps as above or try to connect the same device in USB3 after this issue has occured, the device gets detected and works as intended. To make the port to be usable again in USB4, a restart of the iPad has to be performed. Has this been observed before with PCIDriverKit or DriverKit in general? Is this some kind of a known limitation in iPadOS 17 in particular or does this have something to do with the iPad model too? Is there any workaround that can be followed to prevent this from happening?
5
0
433
1w
DriverKit eligibility and interface matching for Roland DEMORA on macOS
We’re evaluating an AudioDriverKit driver on Apple Silicon for Roland DEMORA (VID 0x0582, PID 0x01AF). The proposed audio service would use IF1 for isochronous output and IF2 for isochronous input, leaving IF3 available for separate use. Before requesting entitlements, we’d appreciate guidance on three points. Can an independent developer team use the development-only USB transport entitlement for this device? Is a distribution entitlement available for this exact VID/PID, and does either path require written authorization from Roland? If so, what documentation should we provide? Can a service matched to IOUSBHostDevice open IF1 and IF2 concurrently as a client while leaving IF3 available for separate matching and ownership? If Apple supports a different provider arrangement for this use case, which arrangement should we use? What is the narrowest supported match scope and entitlement combination for this arrangement? Can the USB grant or matching personality be limited to 0582:01AF and IF1/IF2, and can the same driver App ID and provisioning profile carry the USB transport and AudioDriverKit family entitlements for development and distribution?
1
0
84
1w
DriverKit entitlement eligibility and USB matching for Roland DEMORA audio on macOS
We’re evaluating a macOS driver for Roland DEMORA audio on Apple Silicon. The device identifies as USB VID 0x0582, PID 0x01AF. The audio scope is IF1 for isochronous output and IF2 for isochronous input. We want to leave IF3 available to another client. Apple’s documentation describes descriptor-based USB transport entitlements and team-specific entitlement approval. Our candidate is a service matched to IOUSBHostDevice that opens IF1 and IF2 concurrently as a client. We’d like to confirm which service and provider arrangement Apple supports for this device. Can an independent developer team qualify for USBDriverKit transport access to this third-party VID/PID during development? What additional requirements apply for distribution? Is Roland’s authorization required for either stage? If so, what documentation should the developer provide? What is the narrowest supported matching and provider arrangement for opening IF1 and IF2 while leaving IF3 available? Is a device-matched service supported, or should separate IOUSBHostInterface services coordinate with the audio service? For the supported arrangement, which services need the USBDriverKit transport and AudioDriverKit family entitlements, and can those entitlements be provisioned together where needed? Links to the applicable documentation or a supported example would be appreciated. USB transport entitlement · DriverKit entitlement request process
2
0
108
1w
Autoupdate of macOS app via Sparkle library fails
We have a macOS app distributed outside of the Mac App Store (Developer ID distribution). It incorporates the popular Sparkle library which offers autoupdate functionality (downloading and installing new versions of the app). The autoupdate functionality worked flawlessly for many years but with recent versions of macOS it started failing. To demonstrate the problem we offer the dts-demo.tar.gz file (available for download, see the links below). That file contains the sources of DemoApp and a build_packages.sh script that builds installer packages for two versions of DemoApp. The old version is 3.0 and the new version is 4.0. Manually installing the old version (by double-clicking the .pkg file and following the instructions) succeeds. After launching the app, trying to autoupdate to the new version by clicking the "Check for update" button fails. The process starts, the old version is deinstalled but then the installation of the new version fails. Manually installing the new version of the app over the old version succeeds. The problem has been observed on macOS versions after 26.6.2 (e.g. 26.7.3 or 27.0). The files we offer for download are: dts-demo.tar.gz at https://dx4vjdxzj60n5.cloudfront.net/dts-demo.tar.gz manual_install_v3.log at https://dx4vjdxzj60n5.cloudfront.net/manual_install_v3.log autoupdate_to_v4.log at https://dx4vjdxzj60n5.cloudfront.net/autoupdate_to_v4.log manual_install_v4_over_v3.log at https://dx4vjdxzj60n5.cloudfront.net/manual_install_v4_over_v3.log The provided log files are excerpts from the system's install.log file. The Sparkle framework included in dts-demo.tar.gz comes from the official Sparkle 2.10.1 release. That version is the latest available at the time of writing. Please note that it only accepts URLs with 'http' or 'https' schemes. This autoupdate problem looks like a bug in macOS because manual updates succeed. We'd like to have the bug fixed or at least find a workaround.
3
0
80
1w
Control fan speed on MacPro 3.1 (early 2008)
I have a MacPro 3.1 (early 2008) that has Linux installed on it (Debian 13). One of the fans runs at top speed of 2500 rpm because some sensor in the board can't report the temperature to the OS. The Motherboard model is Mac-F42C88C8. Searched the internet to see if I could find any information on how to identify the sensor or the SMC information about it. I would like to know if there is any information that anyone knows about this and where could I start trying to make a driver to control the fan by getting the value in the sensor. I can see in the program coolercontrol in Linux the Fan that is at high speed is named 'IO' Thanks Regards
0
0
97
1w
Storefront.current.countryCode returns USA despite Apple ID, device region, and sandbox tester all set to Spain
We have an app in TestFlight with a subscription group and an introductory offer (7-day free trial) configured only for Spain. When testing a sandbox purchase, no free trial is offered and prices display in USD. We confirmed via Purchases.storefront (RevenueCat SDK, which wraps Storefront.current.countryCode from StoreKit natively) that the resolved storefront is consistently USA, not ESP. Steps taken to troubleshoot, all showing Spain, with no change to the result: Real Apple ID's country/region (Settings > Name > Media & Purchases > View Account): Spain, locked/non-editable. 3 different sandbox tester accounts (App Store Connect > Sandbox): all set to Spain. Device Language & Region (Settings > General): Spain. No VPN active. Fully deleted and reinstalled the app (not just relaunched). Fully signed out and back in to the real Apple ID on the device, and changed/reverted device Region. Checked billing address on file for the payment method: Spain. Checked carrier/SIM: Orange España. App-level Pricing and Availability in App Store Connect was never configured before (base country defaulted to United States) — we corrected it to Spain and waited 24+ hours. No change. Unchecked "Pause subscription renewals for this tester" (an Apple sandbox testing flag meant to simulate a Terms & Conditions interruption) — no change. In every case, storefront.countryCode continues to return "USA". Question: What else determines the resolved StoreKit storefront for a given Apple ID/device combination, given that every visible setting indicates Spain? Is there a known cache that needs to be cleared on Apple's side? (We also have an open Apple Developer Support case, #30000007050484, still awaiting a first response after 6 days.)
2
0
116
1w
CXCallUpdate not working for outgoing calls
I try to update remoteHandle using CXCallUpdate for outgoing call, but this works only on iOS 15 but not on 17 or 18 (16 didn't test). This problem actual only for outgoing calls, but for incoming calls update works fine. func startOutgoingCall(with callID: UUID, userID: String) { let handle = CXHandle(type: .generic, value: userID) let action = CXStartCallAction(call: callID, handle: handle) callController.requestTransaction(with: action) { [weak self] error in // ... } } func updateOutgoingCall(with callID: UUID, groupID: String) { let update = CXCallUpdate() update.remoteHandle = CXHandle(type: .generic, value: groupID) provider.reportCall(with: callID, updated: update) } I also tried phoneNumber type but it seems initial handle that I set to CXStartCallAction not possible to change (value or even type). I use this handle value to implement recall by tap on call in Recents tab of system address book. But since my calls can transform from p2p to group call, I need to update handle value or find some another way to pass call identification info.
5
1
1.4k
1w
Sandboxed app can't delete AppleDouble (._) files of quarantined files on exFAT, even with user-selected access.
My Mac App Store utility removes Mac metadata files (._* AppleDouble files and .DS_Store) from external drives formatted exFAT or FAT32. The app is sandboxed with com.apple.security.files.user-selected.read-write and com.apple.security.files.bookmarks.app-scope. The user grants access to the volume through NSOpenPanel. Deleting a ._ file works, except when its companion file carries com.apple.quarantine (for example, a file downloaded with Safari and copied to the drive with Finder). Then: let partner = volume + "/Nirvana.zip" // has com.apple.quarantine let appleDouble = volume + "/._Nirvana.zip" removexattr(partner, "com.apple.quarantine", XATTR_NOFOLLOW) // -1, errno 1 (EPERM) unlink(appleDouble) // -1, errno 1 (EPERM)The same unlink succeeds from Terminal (not sandboxed, without Full Disk Access). On exFAT the extended attributes live in the ._ file, so I assume the sandbox refuses because deleting it would remove the quarantine attribute. Tried, with no change: Full Disk Access granted to the app. Adding com.apple.security.files.user-selected.executable (verified in the built app with codesign -d --entitlements -). On a real drive (a music library), about 71% of the ._ files belong to quarantined files, so this is most of what users want removed. Is there a supported way for a sandboxed app to remove these files: an entitlement, an API, or an explicit user action? If not, can you confirm it's intended behavior, so I can explain it accurately to users? macOS 27.2, Xcode 27.2, Apple silicon. Focused sample project available (one source file). DTS case 23240704.
5
0
174
1w
Supported App Sandbox metadata access through an inherited file descriptor, without reopening the object
We are evaluating a macOS design before implementation. This is a platform-support question, not a report of an observed runtime failure. No actual controller executable, helper executable or deployment configuration has been selected or qualified. Our objective is to let a small, fixed-purpose helper examine an already-open object’s identity and access control list (ACL) metadata while preserving the original object reference and ownership. Proposed configuration A trusted controller would launch one native C helper using an explicit executable target and posix_spawn. The controller would need to qualify for the documented nonsandboxed-parent case, without a conflicting inherited sandbox. We do not assume that an absent entitlement alone establishes this. The helper would be the main executable of its own signed application bundle, using its own ordinary App Sandbox rather than inheriting the controller’s sandbox. Its proposed additional filesystem permission would be com.apple.security.temporary-exception.files.absolute-path.read-only, containing exactly two fixed directory entries beneath a user’s development workspace, outside the helper’s container. The literal private paths are omitted. The controller would already own a read-only file descriptor for one regular file or directory within those trees. It would retain its original reference and pass a known duplicate referring to the same open object to the helper. The helper would perform one metadata observation. The proposed operations are fstat for object identity and fgetattrlist requesting ATTR_CMN_RETURNED_ATTRS and ATTR_CMN_EXTENDED_SECURITY, with FSOPT_REPORT_FULLSIZE and without FSOPT_PACK_INVAL_ATTRS. Required boundaries The helper must use only the inherited descriptor for the subject observation. The design does not permit: Reopening or resolving the subject through a pathname, bookmark or URL. Substituting a copied file or another object reference. Reading file contents, enumerating directories, changing permissions, or modifying shared descriptor flags or file offsets. Expanding the exception to an ancestor, home directory or broader workspace. Disabling an existing sandbox or using private interfaces to bypass restrictions. Parent-side traversal and identity checks remain separate responsibilities. Strict ancestor directories are not additional helper ACL subjects. We understand that a read-only directory exception and ordinary sandbox/container capabilities may permit more than the helper’s intended metadata operations. Metadata-only behaviour would be a reviewed-code obligation; we are not asking App Sandbox to enforce an exact metadata-only operation filter. Questions Supported launch and activation: Under what documented parent-context, signing, bundle, container and launch conditions can this helper establish its own ordinary App Sandbox? What supported mechanism or guarantee establishes that the sandbox is effective before the subject metadata operation? Required metadata permissions: Does the proposed read-only exception support fstat and the specified extended-security fgetattrlist request through an inherited read-only descriptor, for both regular files and directories? Please distinguish descriptor survival from permission to perform these operations, including any additional access-control conditions. Scope, identity and lifetime: What documented behaviour applies to descendant coverage, aliases, hard links, renames or namespace changes after the descriptor was opened? Please distinguish the lifetime of the inherited object reference from the lifetime of the static directory permission. Our design would refuse uncertainty rather than reopen, rebind or widen access. Evidence and support boundary: Which public documentation, interfaces or authoritative guarantees support this exact composition, and which parts require separate empirical qualification? Please identify applicable macOS versions and restrictions. Distinguish platform support from any distribution or App Store review requirements. If these requirements cannot be combined through supported interfaces, please identify the incompatible requirement. An approach requiring broader access, pathname reopening, a replacement object or a different handoff architecture would be a design change requiring separate review, not a solution within the stated constraints. Permission failures, unsupported requests and incomplete metadata would remain unavailable results. They would not be interpreted as proof that an ACL is absent. Complete ACL-absence interpretation and native resource bounds remain separate qualification questions. Apple material already reviewed Resolving App Sandbox Inheritance Problems App Sandbox Temporary Exception Entitlements Accessing files from the macOS App Sandbox
Replies
1
Boosts
0
Views
246
Activity
6d
macOS 26.x App Sandbox: supported disk-image attachment and descriptor-based unmount
On macOS 26.x, is there a public, supported way for an App Sandbox application with com.apple.security.virtualization to attach a new disk-image file held in its own container and mount its APFS filesystem within that container, without root, privileged or unsandboxed helpers, private/temporary exception entitlements, Full Disk Access, or weakening App Sandbox? We distinguish attaching the image from mounting an already attached virtual device. DTS threads 655767, 780965 and 775835 discuss these separately. Which supported API or tool and entitlements apply on 26.x, if such a path exists? If attachment is supported, what ownership and sandbox checks apply to a later normal funmount(fd, 0) using an O_EVTONLY descriptor for that exact mount's root? Is this combination supported, including when another actor has already unmounted it? If not, what supported alternative preserves the same restricted profile and avoids force or whole-disk unmount?
Replies
0
Boosts
0
Views
95
Activity
6d
NSE Filtering for encrypted incoming calls: what technical details should we provide?
We’re preparing the first iOS release of ANTINTWRK, an encrypted messenger with voice and video calls. Our push gateway sends a regular APNs alert for the encrypted event. On the receiving device, the Notification Service Extension fetches and decrypts the referenced event. In this delivery path, the gateway does not distinguish an encrypted call invitation from a message. When the extension identifies a ringing call, it uses CXProvider.reportNewIncomingVoIPPushPayload to hand it off to CallKit and suppresses the original alert to avoid a duplicate notification. Our Notification (NSE) Filtering request, submitted on October 25, 2026, still shows Submitted. Xcode cannot export the App Store build because the extension’s distribution profile lacks com.apple.developer.usernotifications.filtering. We have contacted Developer Support under Entitlements. We’re following Apple’s Sending End-to-End Encrypted VoIP Calls guidance. For this design, what technical details or test evidence should we provide so the capability review team can evaluate the request? Is there a recommended way to attach that material to the existing request? We can provide the request and support case IDs privately if needed.
Replies
0
Boosts
0
Views
146
Activity
1w
HKStatisticsCollectionQueryDescriptor intermittently returns no data for certain date ranges on iOS 27
We are seeing inconsistent results from HKStatisticsCollectionQueryDescriptor on iOS 27. Using the same quantity type, statistics options, anchor date, interval components, and predicate configuration, some date ranges return the expected statistics, while other ranges unexpectedly return empty results or buckets with no quantity. The affected ranges do contain HealthKit samples: HKSampleQueryDescriptor finds samples in the same date range. HKStatisticsQueryDescriptor returns the expected value when run separately for an affected bucket. HKStatisticsCollectionQueryDescriptor returns no quantity for that same bucket. Slightly expanding or shifting the date range may cause the collection query to return data again. A simplified version of the query looks like this: let datePredicate = HKQuery.predicateForSamples( withStart: startDate, end: endDate, options: .strictStartDate ) let descriptor = HKStatisticsCollectionQueryDescriptor( predicate: .quantitySample( type: quantityType, predicate: datePredicate ), options: .cumulativeSum, anchorDate: anchorDate, intervalComponents: DateComponents(day: 1) ) let collection = try await descriptor.result(for: healthStore) collection.enumerateStatistics(from: startDate, to: endDate) { statistics, _ in let quantity = statistics.sumQuantity() print(statistics.startDate, quantity as Any) } Expected behavior Every interval containing matching samples should return the corresponding statistics, regardless of the overall requested date range. Actual behavior Some date ranges produce missing or empty buckets even though matching samples exist and an individual HKStatisticsQueryDescriptor can calculate the expected value. Changing only the date range can make the data appear or disappear. The samples are visible to the app in the affected range, so this does not appear to be explained solely by iOS 27’s Limited History authorization. This behavior was not observed with the same query flow on earlier iOS versions. Is this a known regression in HKStatisticsCollectionQueryDescriptor on iOS 27, or has the expected date-range or predicate behavior changed?
Replies
4
Boosts
3
Views
1.2k
Activity
1w
CKKS zone deletions return success but stale 2019 key hierarchy persists; "Delete iCloud Data" not offered (FB25584356)
I filed FB25584356 for this and would appreciate any pointers. Setup: Mac mini on macOS 27.2.1 and an iPhone on iOS 27.2, same Apple Account, HSA2, Advanced Data Protection off. Octagon is healthy (otctl status = Ready, both devices trusted). Symptoms: AirDrop to my own iPhone fails when I pick the device (rapportd: "PairVerify server M2 failed: copy identity"), iPhone Mirroring ends with noCompatiblePhone, and various clients log CKErrorDomain 111 / 5009 "Keychain Sync Incomplete". What I see in CKKS: seven zones (AutoUnlock, LimitedPeersAllowed, Home, Manatee, ApplePay, Health, Engram) still hold a key hierarchy from August 2019. The only TLKShare was issued by a device that left the account years ago, so neither current device can unwrap the TLK ("No trusted TLKShares", CKKSErrorDomain 35). secd then does what I would expect: tlkmissing → "No other devices claim to have the TLK. Resetting zone..." → CKKSDeleteCKZoneOperation → local reset → zones recreated. The delete appears to succeed. With debug logging, cloudd shows status:200 and "RecordZone ... was successfully deleted from the server" for each zone. The next fetch returns the same 2019 records with an identical server change token. Over about 11 hours: 124 reset cycles, 869 zone deletions reported as successful, 182 consecutive AutoUnlock fetches with the same change token. When not throttled it runs roughly every 5 seconds; CloudKit eventually answers "Service Unavailable" (CKError 6/2009) with a ~59 minute retry-after, and all 23 zones are unusable in the meantime. ckksctl -v AutoUnlock reset-cloudkit behaves the same way. Reset path: Support article 120862 describes "Don't know passcode?" → "Forgot all passcodes?" → "More Options" → "Delete iCloud Data". With the Mac signed in but not yet trusted, "Don't know passcode?" only offered Cancel / Try Again Later, in the sign-in flow and in both System Settings follow-ups. After three wrong passcodes the alert had a single "Try Again Later" button, although cdpd logged "Creating recovery option: Reset". Questions: Is a client-side zone delete expected to be a no-op for these zones while the server reports success? Is there a supported way to reach performCKServerUnreadableDataRemoval / "Delete iCloud Data" when another trusted device exists? I am fine with losing the end-to-end encrypted data in those zones. Logs and a sysdiagnose are attached to the Feedback.
Replies
0
Boosts
0
Views
135
Activity
1w
Screen Time reaches the iPhone Instagram limit after Mac sync while Instagram is in the background (on both devices)
Hello, I have noticed, once again, that DeviceActivityMonitors are quite frustrating to work with, since they are triggering random events when using mulitple Apple devices. For example, if Instagram is open on some random (non-foreground!) tab in Safari on the Mac, the iPhone will reach its limit quite consistently & quickly atfer setting up a new device activity threshold. It would be great if A) background usage was not counted towards screen time: FB25576930 B) there was an API for us developers to set-up device-specific deviceactivitymonitors: FB24363336 Hope that helps!
Replies
0
Boosts
0
Views
102
Activity
1w
URGENT: Muse is AUTOMATICALLY blocked by Screen Time API whenever Facebook is blocked, 100% of the time with no workaround for users or developers
PLEASE ADDRESS ASAP: Muse is #1 on the App Store and Facebook is one of the most blocked apps for Screen Time. This is affecting a vast number of people. Description: When a user blocks the Facebook app in any screen time app, iOS automatically blocks Meta's Muse app too. Every single time. There is no way for the user or the developer to prevent it, allowlist it, or even detect that it happened. Muse shows Apple's generic blocked-app "Restricted" screen instead of the screen time app's own design, so the user cannot tell who blocked it or why. This is the worst possible combination of apps: Facebook is one of the most-blocked app across screen time apps. Muse ("Muse from Meta") is currently the #1 app on the App Store. A huge and fast-growing number of users block Facebook and silently lose the world's most popular new app. Muse launched Sept 8, 2026 and adoption is still climbing -- the affected population grows every day. There is NO workaround whatsoever: Users cannot allowlist Muse: it does not appear in Settings > Screen Time > Always Allowed. Developers cannot prevent it: the Screen Time API offers no opt-out; the extra block is applied by iOS itself. Developers cannot detect it: Muse's token (listed separately in FamilyActivityPicker) corresponds to no real identity anywhere else. Developers cannot customize it: their ShieldConfiguration is never consulted -- Muse gets Apple's generic shield UI. Users cannot even block Muse on purpose: shielding Muse's own token is a silent no-op. Screen Time usage statistics are also wrong -- Muse usage is recorded as "Facebook." This Screen Time bug renders screen time apps effectively unusable if users want to block Facebook and use Muse, which is extremely common. How can we get this addressed as soon as possible? Feedback Assistant ticket: https://feedbackassistant.apple.com/feedback/25019832
Replies
9
Boosts
9
Views
690
Activity
1w
Looks like the Home application breaks non-trivial predicates?
I've just bumped into a weird problem. I've created a pretty simple and perfectly working event trigger with a non-trivial predicate which is just an AND of two characteristic-value ones. Then I've renamed it in the Home application, and it stopped working. A dump suggests the cause is a broken predicate, for it looks now rather differently — one of those two characteristic-value predicates seems broken (details below). Is that possible? Could please someone check whether the problem is really a Home ugly bug, or whether I do something wrong at my side? Versions are macOS 27.0.1 (26A434) Home 11.0 (1241.1.7.4.3) I would gladly post a bugreport, but at this moment I am far from being sure I do not overlook something of importance myself. Thanks! My code — for clarity stripped of all the error checking and reporting — looks like this: HMHome *home=...; HMCharacteristic *knihovna=..., *stul=..., *sklenik=...; // light chars, found by names HMActionSet *switchOff=...; // trivial switch-off for another light, found by name HMEventTrigger *trigger=...; // found by name if exists if (trigger) { // if found, the predicate is logged out and that's all NSLog(@"=== predicate found: %@", trigger.predicate); return; } // otherwise, the trigger and predicate are created this way: HMCharacteristicEvent *evt=[[HMCharacteristicEvent alloc] initWithCharacteristic:sklenik triggerValue:@1]; NSPredicate *p1=[HMEventTrigger predicateForEvaluatingTriggerWithCharacteristic:knihovna relatedBy:NSEqualToPredicateOperatorType toValue:@1]; NSPredicate *p2=[HMEventTrigger predicateForEvaluatingTriggerWithCharacteristic:stul relatedBy:NSEqualToPredicateOperatorType toValue:@1]; NSPredicate *p=[[NSCompoundPredicate alloc] initWithType:NSAndPredicateType subpredicates:@[p1, p2]]; NSLog(@"=== predicate created: %@", p); HMEventTrigger *et=[[HMEventTrigger alloc] initWithName:@"TEST-ONLY" events:@[evt] predicate:p]; [home addTrigger:et completionHandler:^(NSError *error) { if (!error) [et addActionSet:switchOff completionHandler:^(NSError *error) { if (!error) [et enable:YES completionHandler:^(NSError *error) { if (!error) NSLog(@"=== success!"); }]; }]; }];As already mentioned, the trigger works perfectly (is trivial enough, would be weird if not). When created and also later, when found by name, the predicate looks like this: === predicate found: (characteristic == <HMCharacteristic, Unique ID: 7444CCCE-1C1B-5234-8CAE-01DA4D811B8B, Type: 00000025, Instance ID: 9> AND characteristicValue == 1) AND (characteristic == <HMCharacteristic, Unique ID: 2E046807-9127-565F-B9CB-10BC0F2EE6D2, Type: 00000025, Instance ID: 9> AND characteristicValue == 1)Then, I just go to the Home application and rename the trigger, nothing else. It immediately stops working properly, and from that moment on, the predicate looks like this — note the difference: === predicate found: characteristic == <HMCharacteristic, Unique ID: 2E046807-9127-565F-B9CB-10BC0F2EE6D2, Type: 00000025, Instance ID: 9> AND (characteristic == <HMCharacteristic, Unique ID: 7444CCCE-1C1B-5234-8CAE-01DA4D811B8B, Type: 00000025, Instance ID: 9> AND characteristicValue == 1)
Replies
7
Boosts
0
Views
375
Activity
1w
MacOS HomeKit private (and still useable) application
How do I write a macOS (no need for i/Pad/OS at all) application which supports HomeKit for my own private use only? My original idea was simply to write it non-provisioned, which would solve all problems perfectly. Nevertheless, that, triple alas, seems not possible (https://developer.apple.com/forums/thread/849288). Provisioned one built locally has a lot of problems, among which there are e.g., big hurdles to run it at my other Macs (https://developer.apple.com/forums/thread/848903?answerId=907702022#907702022) an extremely limited expiration date A reasonable solution would be a creation of a provisioning profile which would include an expiration date in a far future and also IDs of my other computers; alas, far as I understand, this does not seem to be possible with a Personal Team, as detailed in https://developer.apple.com/forums/thread/848903?answerId=907568022#907568022. Whilst a sort of solution might be the nuclear one, it does not feel right the slightest — I need just to run my own app, nothing more, and whilst naturally I gladly accept all the possible dangers of code I wrote myself, I definitely would not want to get completely vulnerable where all the other apps (and other threats) in the world are concerned. I've checked also the Unlisted distribution; far as understand it properly, won't do either, e.g., since it will be declined ... if your app is in a beta or prerelease state — the application is fully intended to be used locally for months or even years in this state. Besides, even Unlisted apps must go through Review, and when I bump to a need to make any kind of change, I want simply do that locally and launch my new version immediately without any Review hassle. Also, in this case the otherwise important privacy issues like e.g., determining the proper computer name (https://developer.apple.com/forums/thread/813853?answerId=874075022#874075022) are of absolutely no importance. Are there other, more reasonable options? What's the proper solution, after all? Thanks!
Replies
4
Boosts
0
Views
629
Activity
1w
MacOS HomeKit non-provisioned application
How to write a macOS (no need for i/Pad/OS at all) application which supports HomeKit without provisioning? Might be a non-GUI plain tool or essentially anything runnable, if it makes the solution easier. All it ever needs is access to HomeKit (in future probably also Matter) APIs and the filesystem (to write logs). The application will be used locally and privately only. Thus, the provisioning is definitely not needed, and it brings ugly problems (compare please the thread https://developer.apple.com/forums/thread/848903). Thanks!
Replies
5
Boosts
0
Views
626
Activity
1w
Develop DriverKit extensions locally with a Free Developer Account
I have not seen a clear answer about this: https://developer.apple.com/forums/thread/776657 https://developer.apple.com/forums/thread/743639?answerId=775701022#775701022 Thus, I would like to pose: Is there any way to install a DriverKit extension locally for development and testing using a free developer account, or is development of DriverKit extensions gated behind a paid developer account? Thank you!
Replies
8
Boosts
0
Views
945
Activity
1w
Supported vnode metadata access after same-UID libproc queries return EPERM
We need a documented interface/access contract for a test supervisor that requires a bounded, process-birth-bound census of every process with its effective UID and every current vnode descriptor of those processes. Coverage includes ambient same-UID processes and descriptors opened before observation starts. The census needs vnode identity, type, ownership, mode, link count and open flags. Unavailable metadata is a refusal; we cannot treat EPERM as absence or exempt a process or descriptor. This is a cooperative same-user lifetime-containment check. We do not require file contents or an atomic global snapshot. Environment of the retained observation: Apple Silicon, macOS 27.3 (26C5035r), Darwin 27.3.0; Xcode 27.0 (27A5194q), macOS 27.0 SDK. The subject was the system LaunchServices process /usr/libexec/lsd. Caller real/effective UID matched the subject's UID/RUID/SVUID. Before/after process identity, version, start time, credentials and executable-path samples matched. PROC_PIDLISTFDS returned 32 bytes and listed FD 3 as PROX_FDTYPE_VNODE. The following call shape returned 0 bytes with errno == EPERM (1), with 176 bytes expected in that environment: struct vnode_fdinfo info = {0}; errno = 0; int count = proc_pidfdinfo(pid, fd, PROC_PIDFDVNODEINFO, &info, sizeof(info)); int saved_errno = errno;Errno was cleared before and saved immediately after the call. Credentials were not modified. The matching process bracket does not establish an FD-to-path association or a specific protection category. In particular, we have not established a Data Vault subtype. Questions: Is there a supported public interface or provisionable entitlement that supplies the required metadata for every current same-UID vnode descriptor, including descriptors whose metadata is protected internally by macOS? If Endpoint Security is the supported route, which documented API supplies a complete current FD inventory and vnode mapping, including already-open descriptors and ambient same-UID processes? A solution limited to descendants of the client does not cover this requirement. If supported interfaces cannot guarantee that coverage, could you clarify the limitation? Is there a supported diagnostic that identifies the protection category responsible for this metadata denial without inspecting private policy state? I have considered the "On File System Permissions" guidance, including its mandatory-access-control and Endpoint Security discussion: https://developer.apple.com/forums/thread/678819 This post describes a retained observation, not a new reproduction or qualification result. All retained results remain UNPROVED. A focused SDK/libproc Xcode example was previously built for arm64 and x86_64 but was not run; it illustrates the call sequence and a BSD start/credential bracket, not complete kernel birth binding or the required census. It does not guarantee reproduction on another system. The sanitized supporting record and example are already attached to the private DTS request that directed me to the forums. Please advise the supported interface/access contract, or its limitation, and any further diagnostic scope needed.
Replies
1
Boosts
0
Views
109
Activity
1w
[iPadOS] USB 4 port unusable after driver uninstallation
I have developed a Driverkit extension which matches for iPadOS which is bundled within my app. My driver contains two different personalities - one with provider as USBDriverKit (USB3) and the other as PCIDriverKit (USB4). When enabled, the driver matches/loads, initiates the class with the right personality based on whether the device it matches is operating in USB3/USB4 and works as intended. Recently, we have come across an issue when operating in USB4 and below are the steps to reproduce. Platform: iPadOS 17 (Issue not observed following the same steps in iPadOS 18 and higher, though cannot be full sure as I have tested only in some versions of 18, 26 and 27). Tested in iPad having a USB-4 compatible port. Steps to reproduce: Install the app. This provides an option to enable the driver which comes with it from the app's system settings. Enable the driver from settings. Connect a device which is compatible with the driver in USB4 so that the specific personality can be loaded. Verify the same. With the driver still loaded, uninstall the application. On uninstallation, ideally, Apple's inbox IONVMeController should match on the re-enumeration and get loaded without a re-plug being required. However, this is not the case as the port becomes unresponsive. Even after a re-plug of the device, it still does not get detected by the iPad itself. This is specifically when I try to connect the device in USB 4 as this specific driver personality matches only in this case. If I try the same steps as above or try to connect the same device in USB3 after this issue has occured, the device gets detected and works as intended. To make the port to be usable again in USB4, a restart of the iPad has to be performed. Has this been observed before with PCIDriverKit or DriverKit in general? Is this some kind of a known limitation in iPadOS 17 in particular or does this have something to do with the iPad model too? Is there any workaround that can be followed to prevent this from happening?
Replies
5
Boosts
0
Views
433
Activity
1w
DriverKit eligibility and interface matching for Roland DEMORA on macOS
We’re evaluating an AudioDriverKit driver on Apple Silicon for Roland DEMORA (VID 0x0582, PID 0x01AF). The proposed audio service would use IF1 for isochronous output and IF2 for isochronous input, leaving IF3 available for separate use. Before requesting entitlements, we’d appreciate guidance on three points. Can an independent developer team use the development-only USB transport entitlement for this device? Is a distribution entitlement available for this exact VID/PID, and does either path require written authorization from Roland? If so, what documentation should we provide? Can a service matched to IOUSBHostDevice open IF1 and IF2 concurrently as a client while leaving IF3 available for separate matching and ownership? If Apple supports a different provider arrangement for this use case, which arrangement should we use? What is the narrowest supported match scope and entitlement combination for this arrangement? Can the USB grant or matching personality be limited to 0582:01AF and IF1/IF2, and can the same driver App ID and provisioning profile carry the USB transport and AudioDriverKit family entitlements for development and distribution?
Replies
1
Boosts
0
Views
84
Activity
1w
DriverKit entitlement eligibility and USB matching for Roland DEMORA audio on macOS
We’re evaluating a macOS driver for Roland DEMORA audio on Apple Silicon. The device identifies as USB VID 0x0582, PID 0x01AF. The audio scope is IF1 for isochronous output and IF2 for isochronous input. We want to leave IF3 available to another client. Apple’s documentation describes descriptor-based USB transport entitlements and team-specific entitlement approval. Our candidate is a service matched to IOUSBHostDevice that opens IF1 and IF2 concurrently as a client. We’d like to confirm which service and provider arrangement Apple supports for this device. Can an independent developer team qualify for USBDriverKit transport access to this third-party VID/PID during development? What additional requirements apply for distribution? Is Roland’s authorization required for either stage? If so, what documentation should the developer provide? What is the narrowest supported matching and provider arrangement for opening IF1 and IF2 while leaving IF3 available? Is a device-matched service supported, or should separate IOUSBHostInterface services coordinate with the audio service? For the supported arrangement, which services need the USBDriverKit transport and AudioDriverKit family entitlements, and can those entitlements be provisioned together where needed? Links to the applicable documentation or a supported example would be appreciated. USB transport entitlement · DriverKit entitlement request process
Replies
2
Boosts
0
Views
108
Activity
1w
Autoupdate of macOS app via Sparkle library fails
We have a macOS app distributed outside of the Mac App Store (Developer ID distribution). It incorporates the popular Sparkle library which offers autoupdate functionality (downloading and installing new versions of the app). The autoupdate functionality worked flawlessly for many years but with recent versions of macOS it started failing. To demonstrate the problem we offer the dts-demo.tar.gz file (available for download, see the links below). That file contains the sources of DemoApp and a build_packages.sh script that builds installer packages for two versions of DemoApp. The old version is 3.0 and the new version is 4.0. Manually installing the old version (by double-clicking the .pkg file and following the instructions) succeeds. After launching the app, trying to autoupdate to the new version by clicking the "Check for update" button fails. The process starts, the old version is deinstalled but then the installation of the new version fails. Manually installing the new version of the app over the old version succeeds. The problem has been observed on macOS versions after 26.6.2 (e.g. 26.7.3 or 27.0). The files we offer for download are: dts-demo.tar.gz at https://dx4vjdxzj60n5.cloudfront.net/dts-demo.tar.gz manual_install_v3.log at https://dx4vjdxzj60n5.cloudfront.net/manual_install_v3.log autoupdate_to_v4.log at https://dx4vjdxzj60n5.cloudfront.net/autoupdate_to_v4.log manual_install_v4_over_v3.log at https://dx4vjdxzj60n5.cloudfront.net/manual_install_v4_over_v3.log The provided log files are excerpts from the system's install.log file. The Sparkle framework included in dts-demo.tar.gz comes from the official Sparkle 2.10.1 release. That version is the latest available at the time of writing. Please note that it only accepts URLs with 'http' or 'https' schemes. This autoupdate problem looks like a bug in macOS because manual updates succeed. We'd like to have the bug fixed or at least find a workaround.
Replies
3
Boosts
0
Views
80
Activity
1w
Control fan speed on MacPro 3.1 (early 2008)
I have a MacPro 3.1 (early 2008) that has Linux installed on it (Debian 13). One of the fans runs at top speed of 2500 rpm because some sensor in the board can't report the temperature to the OS. The Motherboard model is Mac-F42C88C8. Searched the internet to see if I could find any information on how to identify the sensor or the SMC information about it. I would like to know if there is any information that anyone knows about this and where could I start trying to make a driver to control the fan by getting the value in the sensor. I can see in the program coolercontrol in Linux the Fan that is at high speed is named 'IO' Thanks Regards
Replies
0
Boosts
0
Views
97
Activity
1w
Storefront.current.countryCode returns USA despite Apple ID, device region, and sandbox tester all set to Spain
We have an app in TestFlight with a subscription group and an introductory offer (7-day free trial) configured only for Spain. When testing a sandbox purchase, no free trial is offered and prices display in USD. We confirmed via Purchases.storefront (RevenueCat SDK, which wraps Storefront.current.countryCode from StoreKit natively) that the resolved storefront is consistently USA, not ESP. Steps taken to troubleshoot, all showing Spain, with no change to the result: Real Apple ID's country/region (Settings > Name > Media & Purchases > View Account): Spain, locked/non-editable. 3 different sandbox tester accounts (App Store Connect > Sandbox): all set to Spain. Device Language & Region (Settings > General): Spain. No VPN active. Fully deleted and reinstalled the app (not just relaunched). Fully signed out and back in to the real Apple ID on the device, and changed/reverted device Region. Checked billing address on file for the payment method: Spain. Checked carrier/SIM: Orange España. App-level Pricing and Availability in App Store Connect was never configured before (base country defaulted to United States) — we corrected it to Spain and waited 24+ hours. No change. Unchecked "Pause subscription renewals for this tester" (an Apple sandbox testing flag meant to simulate a Terms & Conditions interruption) — no change. In every case, storefront.countryCode continues to return "USA". Question: What else determines the resolved StoreKit storefront for a given Apple ID/device combination, given that every visible setting indicates Spain? Is there a known cache that needs to be cleared on Apple's side? (We also have an open Apple Developer Support case, #30000007050484, still awaiting a first response after 6 days.)
Replies
2
Boosts
0
Views
116
Activity
1w
CXCallUpdate not working for outgoing calls
I try to update remoteHandle using CXCallUpdate for outgoing call, but this works only on iOS 15 but not on 17 or 18 (16 didn't test). This problem actual only for outgoing calls, but for incoming calls update works fine. func startOutgoingCall(with callID: UUID, userID: String) { let handle = CXHandle(type: .generic, value: userID) let action = CXStartCallAction(call: callID, handle: handle) callController.requestTransaction(with: action) { [weak self] error in // ... } } func updateOutgoingCall(with callID: UUID, groupID: String) { let update = CXCallUpdate() update.remoteHandle = CXHandle(type: .generic, value: groupID) provider.reportCall(with: callID, updated: update) } I also tried phoneNumber type but it seems initial handle that I set to CXStartCallAction not possible to change (value or even type). I use this handle value to implement recall by tap on call in Recents tab of system address book. But since my calls can transform from p2p to group call, I need to update handle value or find some another way to pass call identification info.
Replies
5
Boosts
1
Views
1.4k
Activity
1w
Sandboxed app can't delete AppleDouble (._) files of quarantined files on exFAT, even with user-selected access.
My Mac App Store utility removes Mac metadata files (._* AppleDouble files and .DS_Store) from external drives formatted exFAT or FAT32. The app is sandboxed with com.apple.security.files.user-selected.read-write and com.apple.security.files.bookmarks.app-scope. The user grants access to the volume through NSOpenPanel. Deleting a ._ file works, except when its companion file carries com.apple.quarantine (for example, a file downloaded with Safari and copied to the drive with Finder). Then: let partner = volume + "/Nirvana.zip" // has com.apple.quarantine let appleDouble = volume + "/._Nirvana.zip" removexattr(partner, "com.apple.quarantine", XATTR_NOFOLLOW) // -1, errno 1 (EPERM) unlink(appleDouble) // -1, errno 1 (EPERM)The same unlink succeeds from Terminal (not sandboxed, without Full Disk Access). On exFAT the extended attributes live in the ._ file, so I assume the sandbox refuses because deleting it would remove the quarantine attribute. Tried, with no change: Full Disk Access granted to the app. Adding com.apple.security.files.user-selected.executable (verified in the built app with codesign -d --entitlements -). On a real drive (a music library), about 71% of the ._ files belong to quarantined files, so this is most of what users want removed. Is there a supported way for a sandboxed app to remove these files: an entitlement, an API, or an explicit user action? If not, can you confirm it's intended behavior, so I can explain it accurately to users? macOS 27.2, Xcode 27.2, Apple silicon. Focused sample project available (one source file). DTS case 23240704.
Replies
5
Boosts
0
Views
174
Activity
1w