Thanks, my feedback report # is FB11845080.

Update, back to square 1... After switching code signing setting to Run Locally and back to Developer, and cleaning, the xcodebuild test command succeeds from the Terminal with this target. Still fails on Xcode Cloud however, including after a clean environment.

Additional info: I discovered that this is not just an issue with Xcode cloud, I can reproduce by just trying to run the testing target from the command-line locally, with xcodebuild -scheme (XCTest-scheme-name) test I get the same error shown above, which does not occur when I run the Test action from the IDE. In this case it's looking for the test file in the Debug .app build in derived data - and the file is actually there. Any suggestions? Is this a code-signing issue?

Many of us would like this! I'm mainly interested in use on Mac in NSImage, but it's the same issue - clearly NSImage has the capability since it can be created from an asset, why isn't it possible from loose .svg resource or svg NSData?

I think I answered my question by searching for "macos sandbox entitlement font auto-activation" online... I came across a troubleshooting article on limitations of Suitcase auto-activation, which won't work for Sandboxed apps (including Pages etc.), which unfortunately this forum won't let me post a link to for some reason. (Can't post the search link either, gah). So it would seem that auto-activation for sandboxed apps is unsupported. I will file a feedback request to ask for an entitlement for this feature, it's an important use case for designers that have large collections of fonts and Sandboxed apps shouldn't be denied the ability to leverage it. A global font auto-activation entitlement would make sense for apps that use typography in significant ways.

Also a follow-up: could there be an entitlement needed for auto-activation to work from a Sandboxed app?

Oh I also tried CTFontManagerSetAutoActivationSetting(kCTFontManagerBundleIdentifier, kCTFontManagerAutoActivationEnabled); which is supposed to set "global auto-activation", but no difference in behavior.

Thanks @edford for the answer - I was able to get it working as expected with my existing project by just adding the guards you mention to my source, so the symbols were not referenced for the not-yet-supported arm64 architecture - I used #if defined(__x86_64__)

After experimenting a bit I'm concluding that this just isn't really a use case supported by the weak-linking mechanism. It looks to me like any time you're embedding a framework in your bundle, the Xcode build sequence is going to try and link its symbols, and if it's a universal app this will mean for both architectures. There's no workaround here even if you're doing things like setting different search paths and bundle locations for x86 vs. arm64. I think to make this work we would need a 'stub library' implementation of the framework for arm64 to get us past the link errors, but I'm still not confident the right thing would happen at runtime (which is, x86 would link with the x86 framework, arm64 would behave as if the optional framework is missing). If anyone from Apple or with expertise can confirm this conclusion, that would be appreciated! And this question then becomes, more generally, what's the approach to have a universal app make use of a library/framework only available for one architecture - is XPC the only way?

Posting a reply to my own question here in case others encounter the same... While I'm not certain about the underlying design or when a non-precomposed RGBA setting would be possible or valid for a bitmap CGContext, that configuration turned out to not be needed. I'm now using this code to generate my CGContext - it's sRGB color space, and the alpha value is set to premultiplied. I was concerned this would mean I would somehow need to do some premultiplied calculation for the use of semi-opaque colors in my drawing code, but this turns out not to be the case - I can obtain an image with a transparent background, and also set partial alpha values in colors set for fill or stroke while drawing, and everything works properly. if let colorspace = CGColorSpace(name: CGColorSpace.sRGB) {       if let cgc = CGContext(data: nil,                     width: Int(pixelWidth),                     height: Int(pixelHeight),                     bitsPerComponent: 16,                     bytesPerRow: 0,                     space: colorspace,                     bitmapInfo: CGImageAlphaInfo.premultipliedLast.rawValue |                     CGBitmapInfo.byteOrder16Little.rawValue |                     CGBitmapInfo.floatComponents.rawValue) { ... So my answer in a nutshell is, if you want to use a bitmap CGContext for use in a Mac app with RGB color space that includes transparency, you can just use sRGB color space and premultiplied alpha values for the bitmapInfo parameter.

With more investigation I think I've solved this issue! Hope this helps others who are investigating performance with XCTest and SIMD or other Accelerate technologies... XCTest performance tests can work great for benchmarking and investing alternate implementations even with micro performance, but the trick is to make sure you're not testing code built for debug or running under debugging. I now have XCTest running the performance tests from my original post and showing meaningful (and actionable) results. On my current machine, the 100000 regular Double calculation block has an average measurement of 0.000328 s, while the simd_double2 test block has an average measurement of 0.000257 s, which is about 78% of the non-SIMD time, very close to the difference I measured in my release build. So now I can reliably measure what performance gains I'll get from SIMD and other Accelerate APIs as I decide whether to adopt. Here's the approach I recommend: Put all of your performance XCTests in separate files from functional tests, so you can have a separate target compile them. Create a separate Performance Test target in the Xcode project. If you already have a UnitTest target, it's easy just to duplicate it and rename. Separate your tests between these targets, with the functional tests only in the original Unit Test target, and the performance tests in the Performance Test target. Create a new Performance Test Scheme associated with the Performance Test Target. THE IMPORTANT PART: Edit the Performance Test Scheme, Test action, and set its Build Configuration to Release, uncheck Debug Executable, and uncheck everything under Diagnostics. This will make sure that when you run Project->Test, it's Release-optimized code that's getting run for your performance measurements. There are a couple of additional steps if you want to be able to run performance tests ad hoc from the editor, with your main app set as the current scheme. First you'll need to add the Performance Test target to your main app scheme's Test section. The problem now is that your main app's scheme only has one setting for test configuration (Debug vs. Release), so assuming it's set to Debug when you run your performance test ad hoc it will display the behavior in my OP, with SIMD code especially orders of magnitude slower. I do want my main app's test configuration to remain Debug for working with functional unit test code. So to make performance tests work tolerably in this scenario, I edited the build settings of the Performance Test target (only) so that it's Debug settings were more like Release - the key setting being Swift Compiler Code Generation, changing Debug to Optimize for Speed [-O]. While I don't think this is going to be quite as accurate as running under the Performance Test scheme with Release configuration and all other debug options disabled, I'm now able to run the performance test under my main app's scheme and see reasonable results - it again shows SIMD time measurement in the 75-80% range compared to non-SIMD for the test in question.