Hello, I have encountered an issue with an iPhone 15PM with iOS 18.5. The NSHTTPCookieStorage failed to clear cookies, after clearing them, I was still able to retrieve them. However, on the same system NSHTTPCookie *cookie; NSHTTPCookieStorage *storage = [NSHTTPCookieStorage sharedHTTPCookieStorage]; for (cookie in [storage cookies]) { [storage deleteCookie:cookie]; } NSArray *cookies = [[NSHTTPCookieStorage sharedHTTPCookieStorage] cookiesForURL:[[self url] absoluteURL]]; // still able to get cookies,why???

I'm trying to use NEHotspotNetwork to configure an IoT. I've read all the issues that have plagued other developers when using this framework, and I was under the impression that bugs were filed and fixed. Here are my issues in hopes that someone can catch my bug, or has finally figured this out and it's not a bug in the framework with no immediate fix on the horizon. If I use the following code: let config = NEHotspotConfiguration(ssid: ssid) config.joinOnce = true KiniStatusBanner.shared.show(text: "Connecting to Kini", in: presentingVC.view) NEHotspotConfigurationManager.shared.apply(config) { error in DispatchQueue.main.async { if let nsError = error as NSError?, nsError.domain == NEHotspotConfigurationErrorDomain, nsError.code == NEHotspotConfigurationError.alreadyAssociated.rawValue { print("Already connected to \(self.ssid)") KiniStatusBanner.shared.dismiss() self.presentCaptivePortal(from: presentingVC, activationCode: activationCode) } else if let error = error { // This doesn't happen print("❌ Failed to connect: \(error.localizedDescription)") KiniStatusBanner.shared.update(text: "Failed to Connect to Kini. Try again later.") KiniStatusBanner.shared.dismiss(after: 2.5) } else { // !!!! Most often, this is the path the code takes NEHotspotNetwork.fetchCurrent { current in if let ssid = current?.ssid, ssid == self.ssid { log("✅✅ 1st attempt: connected to \(self.ssid)") KiniStatusBanner.shared.dismiss() self.presentCaptivePortal(from: presentingVC, activationCode: activationCode) } else { // Dev forums talked about giving things a bit of time to settle and then try again DispatchQueue.main.asyncAfter(deadline: .now() + 2) { NEHotspotNetwork.fetchCurrent { current in if let ssid = current?.ssid, ssid == self.ssid { log("✅✅✅ 2nd attempt: connected to \(self.ssid)") KiniStatusBanner.shared.dismiss() self.presentCaptivePortal(from: presentingVC, activationCode: activationCode) } else { log("❌❌❌ 2nd attempt: Failed to connect: \(self.ssid)") KiniStatusBanner.shared.update(text: "Could not join Kini network. Try again.") KiniStatusBanner.shared.dismiss(after: 2.5) self.cleanupHotspot() DispatchQueue.main.asyncAfter(deadline: .now() + 2) { print("cleanup again") self.cleanupHotspot() } } } } log("❌❌ 1st attempt: Failed to connect: \(self.ssid)") KiniStatusBanner.shared.update(text: "Could not join Kini network. Try again.") KiniStatusBanner.shared.dismiss(after: 2.5) self.cleanupHotspot() } As you can see, one can't just use NEHotspotConfigurationManager.shared.apply and has to double-check to make sure that it actually succeeds, by checking to see if the SSID desired, matches the one that the device is using. Ok, but about 50% of the time, the call to NEHotspotNetwork.fetchCurrent gives me this error: NEHotspotNetwork nehelper sent invalid result code [1] for Wi-Fi information request Well, there is a workaround for that randomness too. At some point before calling this code, one can: let locationManager = CLLocationManager() locationManager.requestWhenInUseAuthorization() That eliminates the NEHotspotNetwork nehelper sent invalid result code [1] for Wi-Fi information request BUT... three issues. The user is presented with an authorization alert: Allow "Kini" to use your location? This app needs access to you Wi-Fi name to connect to your Kini device. Along with a map with a location pin on it. This gives my users a completely wrong impression, especially for a device/app where we promise users not to track their location. They actually see a map with their location pinned on it, implying something that would freak out anyone who was expecting no tracking. I understand why an authorization is normally required, but since all we are getting is our own IoT's SSID, there should be no need for an authorization for this, and no map associated with the request. Again, they are accessing my IoT's network, NOT their home/location Wi-Fi SSID. My app already knows and specifies that network, and all I am trying to do is to work around a bug that makes it look like I have a successful return from NEHotspotConfigurationManager.shared.apply() when in fact the network I was looking for wasn't even on. Not only do I get instances where the network doesn't connect, and result codes show no errors, but I also get instances where I get an alert that says that the network is unreachable, yet my IoT shows that the app is connected to its Wi-Fi. On the iOS device, I go to the Wi-Fi settings, and see that I am on the IoT's network. So basically, sometimes I connect, but the frameworks says that there is no connection, and sometimes it reports a connection when there is none. As you can see in the code, I call cleanupHotspot() to make the iOS device get off of my temp Wi-Fi SSID. This is the code: func cleanupHotspot() { NEHotspotConfigurationManager.shared.removeConfiguration(forSSID: ssid) } That code gets called by the above code when things aren't as I expect and need to cleanup. And I also call it when the user dismisses the viewcontroller that is attempting to make the connection. It doesn't always work. I get stuck on the tempo SSID, unless I go through this whole thing again: try to make the connection again, this time it succeeds quickly, and then I can disconnect. Any ideas? I'm on iOS18.5, and have tried this on multiple iPhones including 11, 13 and 16.

Hi, I developed a network extension program on macOS. I tried to update the program by changing the version number. My update process was to first turn off network filtering via "NEFilterManager.sharedManager.enabled = NO", and then use "[OSSystemExtensionRequest activationRequestForExtension:bundleid queue:dispatch_get_global_queue(DISPATCH_QUEUE_PRIORITY_HIGH, 0)];" to let the system replace the old network extension program. However, sometimes the old network extension process will become a zombie process like pid=86621 in the figure. As long as the zombie process exists, the network cannot be used. After about 10 minutes, it will be cleared and the network will be available. Restarting Wi-Fi can also clear the zombie process immediately. Why is this? How to avoid this problem?

We are developing a client server application using TCP bsd sockets. When our client is connected to the server, copying another client .app bundle from a file server on the same machine (using Finder or terminal using cp), occasionally causes the first client to disconnect. The client receives an EBROKENPIPE error when attempting to write to its socket. In the Console, the following message appears just before the disconnection: necp_socket_find_policy_match: Marking socket in state 258 as defunct This issue seems to occur only when copying an .app bundle signed with the same TeamIdentifier as the running client. Copying arbitrary files or bundles with a different TeamIdentifier does not trigger the problem. We are running on macOS 15.5. The issue appears specific to macOS 15 and was not observed on earlier versions. Any help or pointers would be greatly appreciated!

sometimes app from background to foreground , then send a Http request will got network lost response, and if you delay 0.1 seconds to send request, it's work fine. Does any one can explian this?

On "Accessory Interface Specification CarPlay Addendum R10", it says that it is recommended that the accessory uses a MIMO (2x2) hardware configuration, does this imply that WiFi 5 and SISO (1X1) will be phased out in the near future? When will WiFi 6 MIMO (2x2) become mandatory? On "Accessory Interface Specification CarPlay Addendum R10", it says that Spatial Audio is mandatory. However, for aftermarket in-vehicle infotainment (IVI) system due to the number of speakers are less than 6, is it allowed not to support spatial audio for this type of aftermarket IVI system?

Hello team, Would this mean that content filters intended for all browsing can only be implemented for managed devices using MDM? My goal would be to create a content filtering app for all users, regardless of if their device is managed/supervised. thanks.

I am wondering wether iOS allow apps to detect users' proxy.

What is the best way to detect if the Wifi is being used for Wireless Carplay or is just a normal network interface?

Hello, I’m reviewing the open-source mDNSResponder repository and have a question regarding licensing/provenance in mDNSCore/DNSDigest.c file. That file contains an embedded notice stating that parts of the MD5/digest implementation were derived from older OpenSSL sources and therefore include the legacy OpenSSL/SSLeay license text, even though OpenSSL itself is now Apache-2.0 starting from version 3.0. The legacy OpenSSL/SSLeay license is widely understood to impose additional attribution and notice requirements compared to Apache-2.0, and some downstream projects prefer to avoid it when a permissively licensed alternative is available. Repository: https://github.com/apple-oss-distributions/mDNSResponder File: https://github.com/apple-oss-distributions/mDNSResponder/blob/main/mDNSCore/DNSDigest.c#L66 I’d like to clarify a few points: Is the MD5/digest code in DNSDigest.c still based on pre–OpenSSL-3.0 sources, such that retaining the legacy OpenSSL/SSLeay license block is intentional and required? If the goal were to simplify licensing (Apache-2.0 only), would Apple consider replacing this MD5 implementation with an Apache-2.0–licensed alternative (for example, code derived from OpenSSL 3.x or another permissive implementation)? Are there any technical or policy reasons (compatibility, crypto policy, platform APIs) that make such a replacement undesirable? Since GitHub issues and PRs are restricted for this repository, I’m asking here for guidance. If maintainers agree that such an update would be useful, I’d be happy to help by preparing a PR for review. I've also created a feedback report for this topic, the reference ID is FB21269078. Thanks for any clarification.

Transport Layer Security (TLS) is the most important security protocol on the Internet today. Most notably, TLS puts the S into HTTPS, adding security to the otherwise insecure HTTP protocol. IMPORTANT TLS is the successor to the Secure Sockets Layer (SSL) protocol. SSL is no longer considered secure and it’s now rarely used in practice, although many folks still say SSL when they mean TLS. TLS is a complex protocol. Much of that complexity is hidden from app developers but there are places where it’s important to understand specific details of the protocol in order to meet your requirements. This post explains the fundamentals of TLS, concentrating on the issues that most often confuse app developers. Note The focus of this is TLS-PKI, where PKI stands for public key infrastructure. This is the standard TLS as deployed on the wider Internet. There’s another flavour of TLS, TLS-PSK, where PSK stands for pre-shared key. This has a variety of uses, but an Apple platforms we most commonly see it with local traffic, for example, to talk to a Wi-Fi based accessory. For more on how to use TLS, both TLS-PKI and TLS-PSK, in a local context, see TLS For Accessory Developers. Server Certificates For standard TLS to work the server must have a digital identity, that is, the combination of a certificate and the private key matching the public key embedded in that certificate. TLS Crypto Magic™ ensures that: The client gets a copy of the server’s certificate. The client knows that the server holds the private key matching the public key in that certificate. In a typical TLS handshake the server passes the client a list of certificates, where item 0 is the server’s certificate (the leaf certificate), item N is (optionally) the certificate of the certificate authority that ultimately issued that certificate (the root certificate), and items 1 through N-1 are any intermediate certificates required to build a cryptographic chain of trust from 0 to N. Note The cryptographic chain of trust is established by means of digital signatures. Certificate X in the chain is issued by certificate X+1. The owner of certificate X+1 uses their private key to digitally sign certificate X. The client verifies this signature using the public key embedded in certificate X+1. Eventually this chain terminates in a trusted anchor, that is, a certificate that the client trusts by default. Typically this anchor is a self-signed root certificate from a certificate authority. Note Item N is optional for reasons I’ll explain below. Also, the list of intermediate certificates may be empty (in the case where the root certificate directly issued the leaf certificate) but that’s uncommon for servers in the real world. Once the client gets the server’s certificate, it evaluates trust on that certificate to confirm that it’s talking to the right server. There are three levels of trust evaluation here: Basic X.509 trust evaluation checks that there’s a cryptographic chain of trust from the leaf through the intermediates to a trusted root certificate. The client has a set of trusted root certificates built in (these are from well-known certificate authorities, or CAs), and a site admin can add more via a configuration profile. This step also checks that none of the certificates have expired, and various other more technical criteria (like the Basic Constraints extension). Note This explains why the server does not have to include the root certificate in the list of certificates it passes to the client; the client has to have the root certificate installed if trust evaluation is to succeed. In addition, TLS trust evaluation (per RFC 2818) checks that the DNS name that you connected to matches the DNS name in the certificate. Specifically, the DNS name must be listed in the Subject Alternative Name extension. Note The Subject Alternative Name extension can also contain IP addresses, although that’s a much less well-trodden path. Also, historically it was common to accept DNS names in the Common Name element of the Subject but that is no longer the case on Apple platforms. App Transport Security (ATS) adds its own security checks. Basic X.509 and TLS trust evaluation are done for all TLS connections. ATS is only done on TLS connections made by URLSession and things layered on top URLSession (like WKWebView). In many situations you can override trust evaluation; for details, see Technote 2232 HTTPS Server Trust Evaluation). Such overrides can either tighten or loosen security. For example: You might tighten security by checking that the server certificate was issued by a specific CA. That way, if someone manages to convince a poorly-managed CA to issue them a certificate for your server, you can detect that and fail. You might loosen security by adding your own CA’s root certificate as a trusted anchor. IMPORTANT If you rely on loosened security you have to disable ATS. If you leave ATS enabled, it requires that the default server trust evaluation succeeds regardless of any customisations you do. Mutual TLS The previous section discusses server trust evaluation, which is required for all standard TLS connections. That process describes how the client decides whether to trust the server. Mutual TLS (mTLS) is the opposite of that, that is, it’s the process by which the server decides whether to trust the client. Note mTLS is commonly called client certificate authentication. I avoid that term because of the ongoing industry-wide confusion between certificates and digital identities. While it’s true that, in mTLS, the server authenticates the client certificate, to set this up on the client you need a digital identity, not a certificate. mTLS authentication is optional. The server must request a certificate from the client and the client may choose to supply one or not (although if the server requests a certificate and the client doesn’t supply one it’s likely that the server will then fail the connection). At the TLS protocol level this works much like it does with the server certificate. For the client to provide this certificate it must apply a digital identity, known as the client identity, to the connection. TLS Crypto Magic™ assures the server that, if it gets a certificate from the client, the client holds the private key associated with that certificate. Where things diverge is in trust evaluation. Trust evaluation of the client certificate is done on the server, and the server uses its own rules to decided whether to trust a specific client certificate. For example: Some servers do basic X.509 trust evaluation and then check that the chain of trust leads to one specific root certificate; that is, a client is trusted if it holds a digital identity whose certificate was issued by a specific CA. Some servers just check the certificate against a list of known trusted client certificates. When the client sends its certificate to the server it actually sends a list of certificates, much as I’ve described above for the server’s certificates. In many cases the client only needs to send item 0, that is, its leaf certificate. That’s because: The server already has the intermediate certificates required to build a chain of trust from that leaf to its root. There’s no point sending the root, as I discussed above in the context of server trust evaluation. However, there are no hard and fast rules here; the server does its client trust evaluation using its own internal logic, and it’s possible that this logic might require the client to present intermediates, or indeed present the root certificate even though it’s typically redundant. If you have problems with this, you’ll have to ask the folks running the server to explain its requirements. Note If you need to send additional certificates to the server, pass them to the certificates parameter of the method you use to create your URLCredential (typically init(identity:certificates:persistence:)). One thing that bears repeating is that trust evaluation of the client certificate is done on the server, not the client. The client doesn’t care whether the client certificate is trusted or not. Rather, it simply passes that certificate the server and it’s up to the server to make that decision. When a server requests a certificate from the client, it may supply a list of acceptable certificate authorities [1]. Safari uses this to filter the list of client identities it presents to the user. If you are building an HTTPS server and find that Safari doesn’t show the expected client identity, make sure you have this configured correctly. If you’re building an iOS app and want to implement a filter like Safari’s, get this list using: The distinguishedNames property, if you’re using URLSession The sec_protocol_metadata_access_distinguished_names routine, if you’re using Network framework [1] See the certificate_authorities field in Section 7.4.4 of RFC 5246, and equivalent features in other TLS versions. Self-Signed Certificates Self-signed certificates are an ongoing source of problems with TLS. There’s only one unequivocally correct place to use a self-signed certificate: the trusted anchor provided by a certificate authority. One place where a self-signed certificate might make sense is in a local environment, that is, securing a connection between peers without any centralised infrastructure. However, depending on the specific circumstances there may be a better option. TLS For Accessory Developers discusses this topic in detail. Finally, it’s common for folks to use self-signed certificates for testing. I’m not a fan of that approach. Rather, I recommend the approach described in QA1948 HTTPS and Test Servers. For advice on how to set that up using just your Mac, see TN2326 Creating Certificates for TLS Testing. TLS Standards RFC 6101 The Secure Sockets Layer (SSL) Protocol Version 3.0 (historic) RFC 2246 The TLS Protocol Version 1.0 RFC 4346 The Transport Layer Security (TLS) Protocol Version 1.1 RFC 5246 The Transport Layer Security (TLS) Protocol Version 1.2 RFC 8446 The Transport Layer Security (TLS) Protocol Version 1.3 RFC 4347 Datagram Transport Layer Security RFC 6347 Datagram Transport Layer Security Version 1.2 RFC 9147 The Datagram Transport Layer Security (DTLS) Protocol Version 1.3 Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = "eskimo" + "1" + "@" + "apple.com" Revision History: 2025-11-21 Clearly defined the terms TLS-PKI and TLS-PSK. 2024-03-19 Adopted the term mutual TLS in preference to client certificate authentication throughout, because the latter feeds into the ongoing certificate versus digital identity confusion. Defined the term client identity. Added the Self-Signed Certificates section. Made other minor editorial changes. 2023-02-28 Added an explanation mTLS acceptable certificate authorities. 2022-12-02 Added links to the DTLS RFCs. 2022-08-24 Added links to the TLS RFCs. Made other minor editorial changes. 2022-06-03 Added a link to TLS For Accessory Developers. 2021-02-26 Fixed the formatting. Clarified that ATS only applies to URLSession. Minor editorial changes. 2020-04-17 Updated the discussion of Subject Alternative Name to account for changes in the 2019 OS releases. Minor editorial updates. 2018-10-29 Minor editorial updates. 2016-11-11 First posted.

Esim activation. Assuming I already have card data, I use the universal link https://esimsetup.apple.com/esim_qrcode_provisioning?carddata= to install it. However, it always ends up in the system Settings app. The flow: 1. Click the link -> 2. Redirect to Settings -> 3. Show activation dialog. Is there anyway to make the activation flow stay within the app? I couldn't find any documentation for that. This is an example from Revolut app, where the whole flow above happens without leaving the app.

I am currently developing a custom-protocol VPN application for iOS using PacketTunnelProvider. I have also integrated an HTTP proxy service, which is launched via a dylib. The overall flow is as follows: App -> VPN TUN -> Local HTTP Proxy -> External Network I have a question: I am capturing all traffic, and normally, requests sent out by the HTTP proxy are also captured again by the VPN. However, when I send requests using createUdpSession in my code, they are not being captured by the virtual interface (TUN). What could be the reason for this? override func startTunnel(options: [String : NSObject]?, completionHandler: @escaping (Error?) -> Void) { let tunnelNetworkSettings = NEPacketTunnelNetworkSettings(tunnelRemoteAddress: "192.168.18.0") tunnelNetworkSettings.mtu=1400 let ipv4Settings = NEIPv4Settings(addresses: ["192.169.10.10"], subnetMasks: ["255.255.255.0"]) ipv4Settings.includedRoutes=[NEIPv4Route.default()] ipv4Settings.excludedRoutes = [NEIPv4Route(destinationAddress: "10.0.0.0", subnetMask: "255.0.0.0"), NEIPv4Route(destinationAddress: "172.16.0.0", subnetMask: "255.240.0.0"), NEIPv4Route(destinationAddress: "192.168.0.0", subnetMask: "255.255.0.0"), NEIPv4Route(destinationAddress:"127.0.0.0", subnetMask: "255.0.0.0"), ] tunnelNetworkSettings.ipv4Settings = ipv4Settings // Configure proxy settings let proxySettings = NEProxySettings() proxySettings.httpEnabled = true proxySettings.httpServer = NEProxyServer(address: "127.0.0.1", port: 7890) proxySettings.httpsEnabled = true proxySettings.httpsServer = NEProxyServer(address: "127.0.0.1", port: 7890) proxySettings.excludeSimpleHostnames = true proxySettings.exceptionList=["localhost","127.0.0.1"] tunnelNetworkSettings.proxySettings = proxySettings setTunnelNetworkSettings(tunnelNetworkSettings) { [weak self] error in if error != nil { completionHandler(error) return } completionHandler(nil) let stack = TUNInterface(packetFlow: self!.packetFlow) RawScoketFactory.TunnelProvider=self stack.register(stack: UDPDirectStack()) stack.register(stack: TCPDirectStack()) stack.start() } } NWUdpSession.swift // // NWUDPSocket.swift // supervpn // // Created by TobbyQuinn on 2025/2/3. // import Foundation import NetworkExtension import CocoaLumberjack public protocol NWUDPSocketDelegate: AnyObject{ func didReceive(data:Data,from:NWUDPSocket) func didCancel(socket:NWUDPSocket) } public class NWUDPSocket:NSObject{ private let session:NWUDPSession private let timeout:Int private var pendingWriteData: [Data] = [] private var writing = false private let queue:DispatchQueue=QueueFactory.getQueue() public weak var delegate:NWUDPSocketDelegate? public init?(host:String,port:UInt16,timeout:Int=Opt.UDPSocketActiveTimeout){ guard let udpSession = RawScoketFactory.TunnelProvider?.createUDPSession(to: NWHostEndpoint(hostname: host, port: "\(port)"), from: nil) else{ return nil } session = udpSession self.timeout=timeout super.init() session.addObserver(self, forKeyPath: #keyPath(NWUDPSession.state),options: [.new], context: nil) session.setReadHandler({ dataArray, error in self.queueCall{ guard error == nil, let dataArray = dataArray else { print("Error when reading from remote server or connection reset") return } for data in dataArray{ self.delegate?.didReceive(data: data, from: self) } } }, maxDatagrams: 32) } /** Send data to remote. - parameter data: The data to send. */ public func write(data: Data) { pendingWriteData.append(data) checkWrite() } public func disconnect() { session.cancel() } public override func observeValue(forKeyPath keyPath: String?, of object: Any?, change: [NSKeyValueChangeKey : Any]?, context: UnsafeMutableRawPointer?) { guard keyPath == "state" else { return } switch session.state { case .cancelled: queueCall { self.delegate?.didCancel(socket: self) } case .ready: checkWrite() default: break } } private func checkWrite() { guard session.state == .ready else { return } guard !writing else { return } guard pendingWriteData.count > 0 else { return } writing = true session.writeMultipleDatagrams(self.pendingWriteData) {_ in self.queueCall { self.writing = false self.checkWrite() } } self.pendingWriteData.removeAll(keepingCapacity: true) } private func queueCall(block:@escaping ()->Void){ queue.async { block() } } deinit{ session.removeObserver(self, forKeyPath: #keyPath(NWUDPSession.state)) } }

Hi, I'm experiencing intermittent delays with URLSession where requests take 3-4 seconds to be sent, even though the actual server processing is fast. This happens randomly, maybe 10-20% of requests. The pattern I've noticed is I create my request I send off my request using try await urlSession.data(for: request) My middleware ends up receiving this request 4-7s after its been fired from the client-side The round trip ends up taking 4-7s! This hasn't been reproducible consistently at all on my end. I've also tried ephemeral URLSessions (so recreating the session instead of using .shared so no dead connections, but this doesn't seem to help at all) Completely lost on what to do. Please help!

Hello, How long does it usually take for a URL Filter request to be reviewed? It's been 2.5 weeks since we submitted the request form but we haven't received any feedback yet. Just in case, the request ID is D3633USVZZ

When running my app with Xcode16.4, it crashed with the error: dyld[1045]: Symbol not found: ___res_9_state Referenced from: <8B329554-5BEF-38D0-BFCD-1731FA6120CB> /private/var/containers/Bundle/Application/00C941BA-E397-4D0B-B280-E75583FF2890/xxx.app/xxx.debug.dylib Expected in: <7D74C679-8F55-3A01-9AA2-C205A4A19D3E> /usr/lib/libresolv.9.dylib The ___res_9_state related code in my app is: let state = __res_9_state() res_9_ninit(state) var servers = [res_9_sockaddr_union](repeating: res_9_sockaddr_union(), count: Int(MAXNS)) let found = Int(res_9_getservers(state, &servers, Int32(MAXNS))) res_9_nclose(state) if found > 0 { return Array(servers[0..<found]).filter() { $0.sin.sin_len > 0 } } else { return [] } Previously, __res_9_state() could run normally in Xcode 16.1 How to fix this problem?

I'm looking for help with a network extension filtering issue. Specifically, we have a subclass of NEFilterDataProvider that is used to filter flows based upon a set of rules, including source IP and destination IP. We've run into an issue where the source IP is frequently 0.0.0.0 (or the IPv6 equivalent) on outgoing flows. This has made it so rules based upon source IP don't work. This is also an issue as we report these connections, but we're lacking critical data. We were able to work around the issue somewhat by keeping a list of flows that we allow that we periodically check to see if the source IP is available, and then report after it becomes available. We also considered doing a "peekBytes" to allow a bit of data to flow and then recheck the flow, but we don't want to allow data leakage on connections that should be blocked because of the source IP. Is there a way to force the operating system or network extension frameworks to determine the source IP for an outbound flow without allowing any bytes to flow to the network? STEPS TO REPRODUCE Create a network filtering extension for filtering flows using NEFilterDataProvider See that when handleNewFlow: is called, the outgoing flow lacks the source IP (is 0.0.0.0) in most cases There is this post that is discussing a similar question, though for a slightly different reason. I imagine the answer to this and the other post will be related, at least as far as NEFilterDataProvider:handleNewFlow not having source IP is considered. Thanks!

I’m developing an app designed for hospital environments, where public internet access may not be available. The app includes two components: the main app and a Local Connectivity Extension. Both rely on persistent TCP socket connections to communicate with a local server. We’re observing a recurring issue where the extension’s socket becomes unresponsive every 1–3 hours, but only when the device is on the lock screen, even if the main app remains in the foreground. When the screen is not locked, the connection is stable and no disconnections occur. ❗ Issue Details: • What’s going on: The extension sends a keep-alive ping packet every second, and the server replies with a pong and a system time packet. • The bug: The server stops receiving keep alive packets from the extension.  • On the server, we detect about 30 second gap on the server, a gap that shows no packets were received by the extension. This was confirmed via server logs and Wireshark).  • On the extension, from our logs there was no gap in sending packets. From it’s perspective, all packets were sent with no error.  • Because no packet are being received by the server, no packets will be sent to the extension. Eventually the server closes the connection due to keep-alive timeout.  • FYI we log when the NEAppPushProvider subclass sleeps and it did NOT go to sleep while we were debugging. 🧾 Example Logs: Extension log: 2025-03-24 18:34:48.808 sendKeepAliveRequest() 2025-03-24 18:34:49.717 sendKeepAliveRequest() 2025-03-24 18:34:50.692 sendKeepAliveRequest() ... // continuous sending of the ping packet to the server, no problems here 2025-03-24 18:35:55.063 sendKeepAliveRequest() 2025-03-24 18:35:55.063 keepAliveTimer IS TIME OUT... in CoreService. // this is triggered because we did not receive any packets from the server 2025-03-24 18:34:16.298 No keep-alive received for 16 seconds... connection ID=95b3... // this shows that there has been no packets being received by the extension ... 2025-03-24 18:34:30.298 Connection timed out on keep-alive. connection ID=95b3... // eventually closes due to no packets being received 2025-03-24 18:34:30.298 Remote Subsystem Disconnected {name=iPhone|Replica-Ext|...} ✅ Observations: • The extension process continues running and logging keep-alive attempts. • However, network traffic stops reaching the server, and no inbound packets are received by the extension. • It looks like the socket becomes silently suspended or frozen, without being properly closed or throwing an error. ❓Questions: • Do you know why this might happen within a Local Connectivity Extension, especially under foreground conditions and locked ? • Is there any known system behavior that might cause the socket to be suspended or blocked in this way after running for a few hours? Any insights or recommendations would be greatly appreciated. Thank you!

Please consider this trivial C code which deals with BSD sockets. This will illustrate an issue with sendto() which seems to be impacted by the recent "Local Network" restrictions on 15.3.1 macos. #include <stdio.h> #include <stdlib.h> #include <netinet/in.h> #include <arpa/inet.h> #include "sys/socket.h" #include <string.h> #include <unistd.h> #include <ifaddrs.h> #include <net/if.h> // prints out the sockaddr_in6 void print_addr(const char *msg_prefix, struct sockaddr_in6 sa6) { char addr_text[INET6_ADDRSTRLEN] = {0}; printf("%s%s:%d, addr family=%u\n", msg_prefix, inet_ntop(AF_INET6, &sa6.sin6_addr, (char *) &addr_text, INET6_ADDRSTRLEN), sa6.sin6_port, sa6.sin6_family); } // creates a datagram socket int create_dgram_socket() { const int fd = socket(AF_INET6, SOCK_DGRAM, 0); if (fd < 0) { perror("Socket creation failed"); return -1; } return fd; } // returns a string representing the current local time char *current_time() { time_t seconds_since_epoch; time(&seconds_since_epoch); char *res = ctime(&seconds_since_epoch); const size_t len = strlen(res); // strip off the newline character that's at the end of the ctime() output res[len - 1] = '\0'; return res; } // Creates a datagram socket and then sends a messages (through sendto()) to a valid // multicast address. This it does two times, to the exact same destination address from // the exact same socket. // // Between the first and the second attempt to sendto(), there is // a sleep of 1 second. // // The first time, the sendto() succeeds and claims to have sent the expected number of bytes. // However system logs (generated through "log collect") seem to indicate that the message isn't // actually sent (there's a "cfil_service_inject_queue:4466 CFIL: sosend() failed 65" in the logs). // // The second time the sendto() returns a EHOSTUNREACH ("No route to host") error. // // If the sleep between these two sendto() attempts is removed then both the attempts "succeed". // However, the system logs still suggest that the message isn't actually sent. int main() { printf("current process id:%ld parent process id: %ld\n", (long) getpid(), (long) getppid()); // valid multicast address as specified in // https://www.iana.org/assignments/ipv6-multicast-addresses/ipv6-multicast-addresses.xhtml const char *ip6_addr_str = "ff01::1"; struct in6_addr ip6_addr; int rv = inet_pton(AF_INET6, ip6_addr_str, &ip6_addr); if (rv != 1) { fprintf(stderr, "failed to parse ipv6 addr %s\n", ip6_addr_str); exit(EXIT_FAILURE); } // create a AF_INET6 SOCK_DGRAM socket const int sock_fd = create_dgram_socket(); if (sock_fd < 0) { exit(EXIT_FAILURE); } printf("created a socket, descriptor=%d\n", sock_fd); const int dest_port = 12345; // arbitrary port struct sockaddr_in6 dest_sock_addr; memset((char *) &dest_sock_addr, 0, sizeof(struct sockaddr_in6)); dest_sock_addr.sin6_addr = ip6_addr; // the target multicast address dest_sock_addr.sin6_port = htons(dest_port); dest_sock_addr.sin6_family = AF_INET6; print_addr("test will attempt to sendto() to destination host:port -> ", dest_sock_addr); const char *msg = "hello"; const size_t msg_len = strlen(msg) + 1; for (int i = 1; i <= 2; i++) { if (i != 1) { // if not the first attempt, then sleep a while before attempting to sendto() again int num_sleep_seconds = 1; printf("sleeping for %d second(s) before calling sendto()\n", num_sleep_seconds); sleep(num_sleep_seconds); } printf("%s attempt %d to sendto() %lu bytes\n", current_time(), i, msg_len); const size_t num_sent = sendto(sock_fd, msg, msg_len, 0, (struct sockaddr *) &dest_sock_addr, sizeof(dest_sock_addr)); if (num_sent == -1) { fprintf(stderr, "%s ", current_time()); perror("sendto() failed"); close(sock_fd); exit(EXIT_FAILURE); } printf("%s attempt %d of sendto() succeeded, sent %lu bytes\n", current_time(), i, num_sent); } return 0; } What this program does is, it uses the sendto() system call to send a message over a datagram socket to a (valid) multicast address. It does this twice, from the same socket to the same target address. There is a sleep() of 1 second between these two sendto() attempts. Copy that code into noroutetohost.c and compile: clang noroutetohost.c Then run: ./a.out This generates the following output: current process id:58597 parent process id: 21614 created a socket, descriptor=3 test will attempt to sendto() to destination host:port ->ff01::1:14640, addr family=30 Fri Mar 14 20:34:09 2025 attempt 1 to sendto() 6 bytes Fri Mar 14 20:34:09 2025 attempt 1 of sendto() succeeded, sent 6 bytes sleeping for 1 second(s) before calling sendto() Fri Mar 14 20:34:10 2025 attempt 2 to sendto() 6 bytes Fri Mar 14 20:34:10 2025 sendto() failed: No route to host Notice how the first call to sendto() "succeeds", even the return value (that represents the number of bytes sent) matches the number of bytes that were supposed to be sent. Then notice how the second attempt fails with a EHOSTUNREACH ("No route to host") error. Looking through the system logs, it appears that the first attempt itself has failed: 2025-03-14 20:34:09.474797 default kernel cfil_hash_entry_log:6082 <CFIL: Error: sosend_reinject() failed>: [58597 a.out] <UDP(17) out so 891be95f3a70c605 22558774573152560 22558774573152560 age 0> lport 0 fport 12345 laddr :: faddr ff01::1 hash 1003930 2025-03-14 20:34:09.474806 default kernel cfil_service_inject_queue:4466 CFIL: sosend() failed 65 (notice the time on that log messages, they match the first attempt from the program's output log) So even though the first attempt failed, it never got reported back to the application. Then after sleeping for (an arbitrary amount of) 1 second, the second call fails with the EHOSTUNREACH. The system logs don't show any error (at least not the one similar to that previous one) for the second call. If I remove that sleep() between those two attempts, then both the sendto() calls "succeed" (and return the expected value for the number of bytes sent). However, the system logs show that the first call (and very likely even the second) has failed with the exact same log message from the kernel like before. If I'm not wrong then this appears to be some kind of a bug in the "local network" restrictions. Should this be reported? I can share the captured logs but I would prefer to do it privately for this one. Another interesting thing in all this is that there's absolutely no notification to the end user (I ran this program from the Terminal) about any of the "Local Network" restrictions.

I'm trying to use ThreadNetwork API to manage TheradNetworks on device (following this documentation: https://developer.apple.com/documentation/threadnetwork/), but while some functions on THClient work (such as getPreferedNetwork), most don't (storeCredentials, retrieveAllCredentials). When calling these functions I get the following warning/error: Client: -[THClient getConnectionEntitlementValidity]_block_invoke - Error: -[THClient storeCredentialsForBorderAgent:activeOperationalDataSet:completion:]_block_invoke:701: - Error: Error Domain=NSCocoaErrorDomain Code=4099 "The connection to service with pid 414 named com.apple.ThreadNetwork.xpc was invalidated from this process." UserInfo={NSDebugDescription=The connection to service with pid 414 named com.apple.ThreadNetwork.xpc was invalidated from this process.} Error Domain=NSCocoaErrorDomain Code=4099 "The connection to service with pid 414 named com.apple.ThreadNetwork.xpc was invalidated from this process." UserInfo={NSDebugDescription=The connection to service with pid 414 named com.apple.ThreadNetwork.xpc was invalidated from this process.} Failed to store Thread credentials: Couldn’t communicate with a helper application. STEPS TO REPRODUCE Create new project Add Thread Network capability via Xcode UI (com.apple.developer.networking.manage-thread-network-credentials) Trigger storeCredentials let extendedMacData = "9483C451DC3E".hexadecimal let tlvHex = "0e080000000000010000000300001035060004001fffe002083c66f0dc9ef53f1c0708fdb360c72874da9905104094dce45388fd3d3426e992cbf0697b030d474c2d5332302d6e65773030310102250b04106c9f919a4da9b213764fc83f849381080c0402a0f7f8".hexadecimal // Initialize the THClient let thClient = THClient() // Store the credentials await thClient.storeCredentials(forBorderAgent: extendedMacData!, activeOperationalDataSet: tlvHex!) { error in if let error = error { print(error) print("Failed to store Thread credentials: \(error.localizedDescription)") } else { print("Successfully stored Thread credentials") } } NOTES: I tried with first calling getPreferedNetwork to initiate network permission dialog Tried adding meshcop to bojur services Tried with different release and debug build configurations