I tried to install the flang-new compiler from Homebrew on Sequoia OSX. Complex division is broken because file divdc3 is missing. This file comes from libclang_rt.osx.a, a standard LLVM library. This library is missing on OSX. program test integer, parameter :: n=2 complex(kind=8), dimension(n,n) :: V complex(kind=8) :: PER V(1,1)=cmplx(4.0,2.0) V(2,2)=cmplx(5.0,3.0) V(1,2)=0.0 V(2,1)=0.5 PER=cmplx(1.2,1.2) V(:,:)=V(:,:)/PER end program test alainhebert@Alains-MacBook-Air-2 test_complex % flang-new test.f90 Undefined symbols for architecture arm64: “___divdc3”, referenced from: __QQmain in test-fc2bb3.o ld: symbol(s) not found for architecture arm64 flang-new: error: linker command failed with exit code 1 (use -v to see invocation)

In my project, i am initialising bytes with some character in cpp function, func CreateByteWithVal (), and passing to a function, func CreateNSStringFromCString(_ pPtr : UnsafeMutableRawPointer, _ pLength : Int), in swift using Swift-Cpp interop. CreateByteWithVal () allocates bytes on heap with value "AAAAAAAAAA", also calls swift function CreateNSStringFromCString. And func CreateNSStringFromCString (_ pPtr : UnsafeMutableRawPointer, _ pLength : Int) creates a NSString instance using NSString's BytesNoCopy initialiser using the bytes (pPtr) passed to it in parameter. Cpp code: void CppClass::CreateByteWithVal () { char * bytesForString = (char *) malloc (10); memset (bytesForString, 65, 10); Interop_Swift::CreateNSStringFromCString (bytesForString, 10); } Swift code: public func CreateNSStringFromCString (_ pPtr : UnsafeMutableRawPointer, _ pLength : Int) { let ns_string:NSString = NSString (bytesNoCopy: pPtr, length: pLength, encoding: String.Encoding.utf8.rawValue, freeWhenDone: false) } If we modify the byte values in memory from C++ directly, the NSString instance, which is supposed to be immutable by nature, reflects these changes. Is this approach appropriate, or is there something we're overlooking or need to address? In our project, we are mutating the byte values directly like this, and the changes are being reflected in the NSString instance : memset (bytesForString, 66, 5); Essentially, I want to confirm whether we can use this method to modify values through C++ without directly mutating the NSString instance. For the UI, we'll be reading the NSString instance by creating a native Swift String instance from it, like this: let str:String = ns_string as String Will the value of str remain consistent as long as we ensure that the correct bytes are provided by C++?

I have a simple shell script as follows: #!/bin/bash OUTPUT="network.$(date +'%d-%m-%y').info.txt" SUPPORT_ID="email" echo "---------------------------------------------------" > $OUTPUT echo "Run date and time: $(date)" >> $OUTPUT echo "---------------------------------------------------" >> $OUTPUT ifconfig >> $OUTPUT echo "---------------------------------------------------" >> $OUTPUT echo "Network info written to file: $OUTPUT." echo "Please email this file to: $SUPPORT_ID." It just dumps the network config into a file. At some point I will have the file emailed out, but right now I'm just trying to figure out why the output looks like the following? bash ./test.sh .etwork info written to file: network.26-01-25.info.txt .lease email this file to: email Why in the world does the initial character of the last couple of "echo" commands get clipped and turned into periods? The echos for the output of the commands piped into the output file are fine. Strange... Any ideas?

I've created a Julia interface for Apple Accelerate's libSparse, via calling the library functions as if they were C (@ccall). I'm interested in using this in the context of power systems, where the sparse matrix is the Jacobian or the ABA matrix from a sparse grid network. However, I'm puzzled by the performance. I ran a sampling profiler on repeated in-place solves of Ax = b for a large sparse matrix A and random dense vectors b. (A is size 30k, positive definite so Cholesky factorization.) The 2 functions with the largest impact are _SparseConvertFromCoordinate_Double from libSparse.dylib, and BLASStateRelease from libBLAS.dylib. That strikes me as bizarre. This is an in-place solve: there should be minimal overheard from allocating/deallocating memory. Also, it seems strange that the library would repeatedly convert from coordinate form. Is this expected behavior? Thinking it might be an artifact of the Julia-C interface, I wrote up a similar program in C/Objective-C. I didn't profile it, but timing the same operation (repeated in-place solves of Ax = b for random vectors b, with the same matrix A as in the Julia) gave the same duration. I've attached the C/Objective-C below.profiling-comparison.m.txt If you're familiar with Julia, the following will give you the matrix I was working with: using PowerSystems, PowerNetworkMatrices sys = System("pglib_opf_case30000_goc.m") A = PowerNetworkMatrices.ABA_Matrix(sys).data where you can find the .m file here. (As a crude way to transfer A from Julia to C, I wrote the 3 arrays A.nzval, A.colptr, and A.rowval to .txt files as space-separated lists of numbers: the above C/objective-C reads in those files.) To duplicate my Julia profiling, do pkg> add AppleAccelerate#libSparse Profile--note the #libSparse part, these features aren't on the main branch--then run using AppleAccelerate, Profile # run previous code snippet to define A M, N = 10000, size(A)[1] bs = [rand(N) for _ in 1:M] aa_fact = AAFactorization(A) factor!(aa_fact) solve!(aa_fact, bs[1]) # pre-compile before we profile. Profile.init(n = 10^6, delay = 0.0003) @profile (for i in 1:M; solve!(aa_fact, bs[i]); end;) Profile.print(C = true, format = :flat, sortedby = :count)

I’m creating an app using SwiftUI, and I would like to incorporate a small Java codebase that I created for the Android version of the app. Is there a way to package the Java code to work on iOS and macOS

hi, Is it possible to compare two vectors and get a boolean answer? example : uint642_t a; uint642_t b; .../... if(a == b) .../... how to do it ? thank

After swapping the -objectAtIndex: method using method_exchangeImplementations, it will cause continuous memory growth. Connect the iPhone and run the provided project. Continuously tap the iPhone screen. Observe Memory; it will keep growing. Sample code

Crash Log We have a issue with our watch app. When we do a release build with xcode 16 the watch app will not launch and crashes on watchOS 10 and below devices. It does not do this on debug builds...and it does not do this on xcode 15 release/debug builds. Anybody running into watch crashes on xcode 16? Thanks

I have a class object created dynamically using Runtime, and I want to release some manually allocated memory resources when this object is deallocated. To achieve this, I added a custom implementation of the dealloc method using the following code: SEL aSel = NSSelectorFromString(@"dealloc"); class_addMethod(kvoClass, aSel, (IMP)custom_dealloc, method_getTypeEncoding(class_getInstanceMethod(kvoClass, aSel))); However, I encountered some issues. If I don't call the superclass's dealloc method in the cus_dealloc function, the superclass's dealloc implementation will not be executed. On the other hand, if I explicitly call the superclass's dealloc method, the program crashes. Here is the implementation of the cus_dealloc function: void custom_dealloc(id self, SEL _cmd) { // Release other memory ![]("https://developer.apple.com/forums/content/attachment/c7b0c16b-be23-4776-b8db-f22b661c5e7d" "title=iShot_2025-01-03_19.31.34.png;width=1080;height=1895") Class superClass = class_getSuperclass(object_getClass(self)); void (*originIMP)(struct objc_super *, SEL, ...) = (void *)objc_msgSendSuper; struct objc_super *objcSuper = &(struct objc_super){self, superClass}; originIMP(objcSuper, _cmd); } demo

I have a simple shell script as follows: #!/bin/bash OUTPUT="network.$(date +'%d-%m-%y').info.txt" SUPPORT_ID="emailaddress" echo "---------------------------------------------------" > $OUTPUT echo "Run date and time: $(date)" >> $OUTPUT echo "---------------------------------------------------" >> $OUTPUT ifconfig >> $OUTPUT echo "---------------------------------------------------" >> $OUTPUT echo "Network info written to file: $OUTPUT." echo "Please email this file to: $SUPPORT_ID." It just dumps the network config into a file. At some point I will have the file emailed out, but right now I'm just trying to figure out why the output looks like the following? bash ./test.sh .etwork info written to file: network.26-01-25.info.txt .lease email this file to: emailaddress Why in the world does the initial character of the last couple of "echo" commands get clipped and turned into periods? The echos for the output of the commands piped into the output file are fine. Strange... Any ideas?

Hello everyone, There is one thing about Objective-C's memory management that confuses me, which is a returned object's lifetime from methods with names doesn't start with "alloc", "new", "copy", or "mutableCopy". Take this as an example, when using NSBitmapImageRep's representationUsingType:properties: method, it returns an NSData object (reference: https://developer.apple.com/documentation/appkit/nsbitmapimagerep/representation(using:properties:)?language=objc). While testing this out, the NSData seemed to be an owned object (it doesn't get released until the end of the program). From what I understand, this may be an auto-released object which is released at the end of an autorelease pool block. Could someone explain this in more detail? What if I want to release that NSData object before the end of the autorelease pool block? How can I know which object is autoreleased, borrowed, or owned?

I have a Swift Package that contains an Objective-C target. The target contains Objective-C literals but unfortunately the compiler says "Initializer element is not a compile-time constant", what am I doing wrong? Based on the error triggering in the upper half, I take it that objc_array_literals is on. My target definition looks like: .target( name: "MyTarget", path: "Sources/MySourcesObjC", publicHeadersPath: "include", cxxSettings: [ .unsafeFlags("-fobjc-constant-literals") ] ), I believe Objective-C literals are enabled since a long time but I still tried passing in the -fobjc-constant-literals flag and no luck. To be clear I'm not interested in a run-time initialization, I really want it to be compile time. Does anyone know what I can do?

I've been teaching myself Objective-C and I wanted to start creating projects that don't use ARC to become better at memory management and learn how it all works. I've been attempting to build and run applications, but I'm not really sure where to start as modern iOS development is used with Swift and memory management is handled. Is there any way to create modern applications that use Objective-C, UIKit, and not use ARC?

At least with macOS Sequoia 15.5 and Xcode 16.3: $ cat test.cc #include <locale.h> #include <string.h> #include <xlocale.h> int main(void) { locale_t l = newlocale(LC_ALL_MASK, "el_GR.UTF-8", 0); strxfrm_l(NULL, "ό", 0, l); return 0; } $ c99 test.c && ./a.out Assertion failed: (p->val == key), function lookup_substsearch, file collate.c, line 596. Abort trap: 6

Hi! We are seeing a bit surprising behavior of dispatch_main on macOS where it seems to spawn a different thread instead of preserving the one it gets called from. Managed to reproduce it in a completely empty command-line tool project in Xcode int main(int argc, const char * argv[]) { @autoreleasepool { dispatch_main(); return 0; } } I put a breakpoint on the line with dispatch_main and see that I am on Thread 1 and inside main function. That makes sense. I resume execution and pause again. Looking at Thread output in Xcode, I can only see Thread 2. Thread 1 is gone and the executable keeps on running. So dispatch_main did what was expected (prevented the process from termination) but throws out the thread it was called from and creates a new one? Is that behavior expected or am I missing something? Just a brain teaser at this point. But we could not make sense out of it. :)

Hello Apple Team, I'm trying to import the Audodesk FBX SDK to my Objective-C iOS Project. The SDK is written in C++, but has support for iOS and the iOS simulator architectures. I've added the path to the include folder in the Header Search Path I've also added the paths to libfbxsdk.a in the Library Search Paths Finally, I've added the libfbxsdk.a file to the Link Binary with Libraries. However, when I build the project, I get the following error: building for 'iOS', but linking in object file (/Users/Lond/Documents/v2/Autodesk/iOS/2020.3.7/lib/ios/debug/libfbxsdk.a[28](fbxalloc.cxx.o)) built for 'macOS' In the terminal, if I type the command: 
lipo -info libfbxsdk.a I get the message Non-fat file: libfbxsdk.a is architecture: arm64 confirming that I'm using the library for the correct architecture.   Do I need to add any other confifuration option? (Like the other linker flag or something else) I'm quite new to C++, and integrating a C++ SDK into iOS is not easy.   I'm using Mac Os Sonoma 14.6.1 Tested on Xcode 15.4 and 16.2 Target Device: iPhone 13 Pro (iOS 17.6.1) iOS FBX SDK version: 2020.3.7 Link to the SDK if needed: https://aps.autodesk.com/developer/overview/fbx-sdk   Any help would be greatly appreciated Thank you

My company wants to be insure that if my Objective-C to Swift conversions fail in anyway, that the app can revert to using the older Objective-C code. By using a remotely controllable flag, the app can switch which code runs as, both are compiled into the app. Essentially, I create a protocol that describes the original class, then both classes (with a "s" or "o" appended to them) conform to the protocol. Protocol: Object Objective-C class: oObject Swift class: sObject That said, I hit one issue that I just can't seem reason out. I create a Objective-C function that returns the appropriate class: Class<Object> classObject(void) { if (myFlag) { return [sObject class]; } else { return [oObject class]; } } Swift deals with this really well - I can create an initialized object using: let object = classObject().init() but I cannot find a way to do this in Objective-C: Object *object = [[classSalesForceData() alloc] init]; fails with "No known class method for selector 'alloc'" Is there a way to do this? David PS: my workaround is to return an allocated object: Object *createObject(void) { if (myFlag) { return [sObject alloc]; } else { return [oObject alloc]; } }

We developing Native App with C++17 for iOS. We override global new and delete operators. This App deallocate all allocated memories correctly by Run on Xcode (Command + R), but exception occurs launch from xcrun or App icon on iPhone. I debugged the exception. Overriding new operation was called correctly, but overriding delete operation was not called. The default delete was called. I'm not sure why is that. STEPS TO REPRODUCE Build xcode project. Run "xcrun devicectl device install app --device " Run "xcrun devicectl device process launch --console --device " PLATFORM AND VERSION iOS Development environment: Xcode 16.4, macOS macOS Sequoia 15.5 Run-time configuration: iOS 18.5 main.cpp I attached is sample code to reproduce this problem. main.cpp

Hi there, I’m having issue with the python3 installation provided by Xcode’s toolchain. I’m currently writing a LLDB plugin, using the LLDB python API, to allow the user to visualize audio data from the current debugged program in a GUI, using tkinter and matplotlib. I'm using those because I'm developing a cross-platform plugin, as I'm initially a Linux developer who wants to make this available to my fellow Apple audio devs. My issue arise at least on two setups MacOS 12.7.6 Monterey with Xcode 14.2 MacOS 14 with Xcode 15.4 (not my machine) Because I wanna support Xcode’s toolchain, I want to use Xcode’s lldb. Xcode’s lldb uses Xcode’s provided python, which I’m having issues with when loading tkinter.The issue can be reproduced like this : > xcrun python3 -c "import tkinter;tkinter._test()" macOS 12 (1207) or later required, have instead 12 (1206) ! zsh: abort xcrun python3 -c "import tkinter;tkinter._test()" On MacOS 14 the version numbers are : macOS 14 (1407) or later required, have instead 14 (1406) ! You can see it fails to load tkinter. From what I understood so far, it looks like the tkinter/tcl/tk version distributed with Xcode is not supported by MacOS ?I checked and the imported tkinter module is definitely the one provided by Xcode’s toolchain : # Checking where tkinter is installed > fd "^tkinter$" /Applications/Xcode.app /Applications/Xcode.app/Contents/Developer/Library/Frameworks/Python3.framework/Versions/3.8/lib/python3.8/tkinter/ /Applications/Xcode.app/Contents/Developer/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/tkinter/ /Applications/Xcode.app/Contents/Developer/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/site-packages/future/moves/tkinter/ # Checking that Xcode python uses the right module - it matches > xcrun python3 -c "import tkinter;print(tkinter.sys.modules['tkinter'])" <module 'tkinter' from '/Applications/Xcode.app/Contents/Developer/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/tkinter/__init__.py'> I can get a working tkinter working by installing it using homebrew or macports, but I’m not able to use it with Xcode’s python installation. I tried overwriting sys.path to force Xcode’s python to import homebrew’s tkinter module, but it still loads Xcode’s tkinter .so. In the crash report I can see it indeed loads tcl/tk 8.5 and loads _tkinter.cpython-39-darwin.so from Xcode. I could disable the SIP (System Integrity Protection) to force to load another version of the library, but that wouldn’t be something I can ask the users. On the LLDB forum, they advise against using another python interpreter that the one provided by the toolchain. So is there a way to get the provided tkinter/tcl/tk installation to work ? If not I’m confused about why it’s provided in the first place. Thanks a lot for your time and please tell me if you have any questions. PS: if possible i'll post the head of the crash report in the comment of this post

I have a string of the form “Mon 22nd April”. I’m trying to extract the day (i.e. Mon), the date (i.e. 22nd) and the month (i.e. April) using this Applescript: set originalDateString to “Mon 22nd April” -- Extract the components by splitting the string set AppleScript's text item delimiters to " " set dayOfWeekAbbrev to text item 1 of originalDateString set dayOfMonth to text item 2 of originalDateString set monthName to text item 3 of originalDateString When I run this on its own it works as expected: dayOfWeekAbbrev is set to “Mon” dayOfMonth is set to “22nd” monthName is set to “April” When I run this inside a bigger script involving Numbers, the text item delimiters fails to work, no compile or run time errors occur and I end up with: dayOfWeekAbbrev is set to “M” dayOfMonth is set to “o” monthName is set to “n” I.e the first three characters of the string. If I replace originalDateString with the literal string “Mon 22nd April” I get the same result. In other words, text items and being recognized as individual characters, no delimiter (or delimiter is null). Totally confused.