Propagate vfs overlays and -fbuiltin-headers-in-system-modules #79926

hjyamauchi · 2025-03-11T22:02:06Z

This fixes explicit module builds for a hello world program on Windows as well as the ucrt import build failure as in the included test.

hjyamauchi · 2025-03-11T22:03:31Z

This is a new fix to what #78184 intended to fix and the win-crt.swift test comes from there.

hjyamauchi · 2025-03-11T22:04:22Z

@swift-ci please test Windows platform

hjyamauchi · 2025-03-11T22:04:38Z

@swift-ci please test

hjyamauchi · 2025-03-11T22:11:51Z

@swift-ci please test

hjyamauchi · 2025-03-12T17:58:03Z

@swift-ci please test

hjyamauchi · 2025-03-12T21:49:41Z

@swift-ci please test Linux platform

lib/ClangImporter/ClangImporter.cpp

lib/ClangImporter/ClangModuleDependencyScanner.cpp

cachemeifyoucan · 2025-03-13T19:21:16Z

lib/ClangImporter/ClangModuleDependencyScanner.cpp

@@ -185,6 +193,13 @@ ModuleDependencyVector ClangImporter::bridgeClangModuleDependencies(
        swiftArgs.push_back("-vfsoverlay");
        swiftArgs.push_back(remapPath(overlay));
      }
+
+      // Pass along the SDK path


Can you explain why this is needed? The swift compile PCM job is pretty much only using cc1 arguments and I don't see why a swift argument her is useful.

This is how building hello.swift with -explicit-module-build fails on Windows without this due to diag::warn_module_config_mismatch. I think this was caused by a mismatch due to a missing -sdk flag in the ASTReader code. With this, this succeeds and hello.swift builds correctly.

If this doesn't make sense, I can dig more.

>swiftc hello.swift -explicit-module-build -v "C:\Users\hiroshi\source\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\bin\swift-frontend.exe" -frontend -scan-dependencies -target x86_64-unknown-windows-msvc -disable-objc-interop -sdk "S:\Program Files\Swift\Platforms\Windows.platform\Developer\SDKs\Windows.sdk" -color-diagnostics -new-driver-path "C:\Users\hiroshi\source\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\bin\swift-driver.exe" -empty-abi-descriptor -resource-dir "C:\Users\hiroshi\source\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\lib\swift" -no-auto-bridging-header-chaining -module-name hello -in-process-plugin-server-path "C:\Users\hiroshi\source\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\bin\SwiftInProcPluginServer.dll" -plugin-path "C:\Users\hiroshi\source\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\bin" -plugin-path "C:\Users\hiroshi\source\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\local\bin" -scanner-module-validation -no-auto-bridging-header-chaining -scanner-output-dir C:\Users\hiroshi\AppData\Local\Temp\TemporaryDirectory.RYev3J\scanner hello.swift Swift version 6.2-dev (LLVM aa33e9831968e53, Swift ff6a1e1138782ce) Target: x86_64-unknown-windows-msvc "C:\Users\hiroshi\source\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\bin\swift-frontend.exe" -frontend -compile-module-from-interface -target x86_64-unknown-windows-msvc -swift-version 5 -sdk "S:\Program Files\Swift\Platforms\Windows.platform\Developer\SDKs\Windows.sdk" -suppress-warnings -disable-objc-attr-requires-foundation-module -enable-ossa-modules -no-clang-include-tree -track-system-dependencies "S:\Program Files\Swift\Platforms\Windows.platform\Developer\SDKs\Windows.sdk\usr\lib\swift\windows\SwiftOnoneSupport.swiftmodule\x86_64-unknown-windows-msvc.private.swiftinterface" -module-name SwiftOnoneSupport -disable-objc-attr-requires-foundation-module -target x86_64-unknown-windows-msvc -disable-objc-interop -enable-experimental-feature NoncopyableGenerics2 -enable-experimental-feature SuppressedAssociatedTypes -enable-experimental-feature SE427NoInferenceOnExtension -enable-experimental-feature NonescapableTypes -enable-experimental-feature LifetimeDependence -enable-upcoming-feature MemberImportVisibility -enable-library-evolution -module-link-name swiftSwiftOnoneSupport -parse-stdlib -swift-version 5 -O -library-level api -enforce-exclusivity=unchecked -disable-objc-interop -target-min-inlining-version min -module-name SwiftOnoneSupport -strict-memory-safety -enable-lexical-lifetimes=false -enable-ossa-modules -interface-compiler-version 6.2 -explicit-interface-module-build -disable-implicit-swift-modules -Xcc -fno-implicit-modules -Xcc -fno-implicit-module-maps "-swift-module-file=Swift=S:\Program Files\Swift\Platforms\Windows.platform\Developer\SDKs\Windows.sdk\usr\lib\swift\windows\Swift.swiftmodule\x86_64-unknown-windows-msvc.swiftmodule" -Xcc -fmodule-file=SwiftShims=C:\Users\hiroshi\AppData\Local\clang\ModuleCache\SwiftShims-A4U7ZTF76P8Z79XYETHS7YEGN.pcm -Xcc -fmodule-file=_Builtin_stdint=C:\Users\hiroshi\AppData\Local\clang\ModuleCache\_Builtin_stdint-BP8JEOXOE7GZI1QX2C0AR3XAE.pcm -o SwiftOnoneSupport-2A4D0UPZ41A5O.swiftmodule -load-resolved-plugin "C:\Users\hiroshi\source\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\bin\SwiftMacros.dll##SwiftMacros" error: compile-module-from-interface command failed with exit code 1 (use -v to see invocation) <unknown>:0: error: builtin headers belong to system modules, and _Builtin_ modules are ignored for cstdlib headers was disabled in PCH file but is currently enabled <unknown>:0: error: module file C:\Users\hiroshi\AppData\Local\clang\ModuleCache\SwiftShims-A4U7ZTF76P8Z79XYETHS7YEGN.pcm cannot be loaded due to a configuration mismatch with the current compilation S:\Program Files\Swift\Platforms\Windows.platform\Developer\SDKs\Windows.sdk\usr\lib\swift\windows\SwiftOnoneSupport.swiftmodule\x86_64-unknown-windows-msvc.private.swiftinterface:5:8: error: missing required module 'SwiftShims' 3 | // swift-module-flags: -disable-objc-attr-requires-foundation-module -target x86_64-unknown-windows-msvc -disable-objc-interop -enable-experimental-feature NoncopyableGenerics2 -enable-experimental-feature SuppressedAssociatedTypes -enable-experimental-feature SE427NoInferenceOnExtension -enable-experimental-feature NonescapableTypes -enable-experimental-feature LifetimeDependence -enable-upcoming-feature MemberImportVisibility -enable-library-evolution -module-link-name swiftSwiftOnoneSupport -parse-stdlib -swift-version 5 -O -library-level api -enforce-exclusivity=unchecked -disable-objc-interop -target-min-inlining-version min -module-name SwiftOnoneSupport 4 | // swift-module-flags-ignorable: -strict-memory-safety -enable-lexical-lifetimes=false -enable-ossa-modules -interface-compiler-version 6.2 5 | import Swift | `- error: missing required module 'SwiftShims' 6 |

It doesn't really make sense. That doesn't look like the output from the new swift-driver. I don't think building explicit module with legacy driver is really supported and we don't go out of way to make that working.

Can you check which swiftc you are running? Maybe windows bot is missing the new driver, thus it is not going down the right code path. Check lib/DriverTool/driver.cpp and make sure it goes down the path to run swift-driver.

Interesting :)

The swiftc that I'm using is S:\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\bin\swiftc.exe which is produced by my local toolchain build with build.ps1.

The new swift-driver is what's from the swift-driver repo and is written in Swift, and the legacy driver is the one from the swift repo and is written in mostly C++ (which I think is the same binary as swift-frontend.exe below), right? (Note we don't use symlinks on windows)

I think the windows build process in build.ps1 installs the new driver over the legacy driver.

I understand that the tests may have not the new driver. The command line is a manual one, not part of a test.

According to ls, diff and dumpbin, I think the swiftc.exe is the same as the new driver swift-driver.exe and swift-frontend.exe is the legacy driver.

So I think I am using the new driver. But I'm not sure what's different here.

S:\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\bin>ls -l swift*.exe -rwxr-xr-x 1 hiroshi 197121 154601984 Mar 14 15:33 swift-api-digester.exe -rwxr-xr-x 1 hiroshi 197121 148999168 Mar 14 15:33 swift-ast-script.exe -rwxr-xr-x 1 hiroshi 197121 154601984 Mar 14 15:33 swift-autolink-extract.exe -rwxr-xr-x 1 hiroshi 197121 46592 Mar 14 15:25 swift-build-sdk-interfaces.exe -rwxr-xr-x 1 hiroshi 197121 1731072 Mar 14 15:24 swift-build-tool.exe -rwxr-xr-x 1 hiroshi 197121 23552 Mar 14 15:28 swift-build.exe -rwxr-xr-x 1 hiroshi 197121 154601984 Mar 14 15:33 swift-cache-tool.exe -rwxr-xr-x 1 hiroshi 197121 341504 Mar 14 15:33 swift-compatibility-symbols.exe -rwxr-xr-x 1 hiroshi 197121 885760 Mar 14 15:33 swift-def-to-strings-converter.exe -rwxr-xr-x 1 hiroshi 197121 759296 Mar 14 15:33 swift-demangle.exe -rwxr-xr-x 1 hiroshi 197121 41984 Mar 14 15:35 swift-driver.exe -rwxr-xr-x 1 hiroshi 197121 23552 Mar 14 15:28 swift-experimental-sdk.exe -rwxr-xr-x 1 hiroshi 197121 188416 Mar 14 15:29 swift-format.exe -rwxr-xr-x 1 hiroshi 197121 154608640 Mar 14 15:35 swift-frontend.exe -rwxr-xr-x 1 hiroshi 197121 49664 Mar 14 15:24 swift-help.exe -rwxr-xr-x 1 hiroshi 197121 151496192 Mar 14 15:33 swift-ide-test.exe -rwxr-xr-x 1 hiroshi 197121 194048 Mar 14 15:30 swift-inspect.exe -rwxr-xr-x 1 hiroshi 197121 23552 Mar 14 15:28 swift-package.exe -rwxr-xr-x 1 hiroshi 197121 22528 Mar 14 14:45 swift-plugin-server.exe -rwxr-xr-x 1 hiroshi 197121 5738496 Mar 14 15:33 swift-reflection-dump.exe -rwxr-xr-x 1 hiroshi 197121 23552 Mar 14 15:28 swift-run.exe -rwxr-xr-x 1 hiroshi 197121 394752 Mar 14 15:33 swift-scan-test.exe -rwxr-xr-x 1 hiroshi 197121 23552 Mar 14 15:28 swift-sdk.exe -rwxr-xr-x 1 hiroshi 197121 747520 Mar 14 15:33 swift-serialize-diagnostics.exe -rwxr-xr-x 1 hiroshi 197121 154601984 Mar 14 15:33 swift-symbolgraph-extract.exe -rwxr-xr-x 1 hiroshi 197121 154601984 Mar 14 15:33 swift-synthesize-interface.exe -rwxr-xr-x 1 hiroshi 197121 23552 Mar 14 15:28 swift-test.exe -rwxr-xr-x 1 hiroshi 197121 41984 Mar 14 15:35 swift.exe -rwxr-xr-x 1 hiroshi 197121 41984 Mar 14 15:35 swiftc.exe S:\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\bin>diff swiftc.exe swift-driver.exe S:\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\bin>echo %errorlevel% 0 S:\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\bin>dumpbin /dependents swiftc.exe Microsoft (R) COFF/PE Dumper Version 14.42.34436.0 Copyright (C) Microsoft Corporation. All rights reserved. Dump of file swiftc.exe File Type: EXECUTABLE IMAGE Image has the following dependencies: mimalloc-redirect.dll mimalloc.dll SwiftDriverExecution.dll SwiftDriver.dll SwiftOptions.dll TSCBasic.dll swiftCore.dll swiftWinSDK.dll KERNEL32.dll swiftCRT.dll FoundationEssentials.dll swiftDispatch.dll BlocksRuntime.dll VCRUNTIME140.dll api-ms-win-crt-runtime-l1-1-0.dll api-ms-win-crt-heap-l1-1-0.dll api-ms-win-crt-math-l1-1-0.dll api-ms-win-crt-stdio-l1-1-0.dll api-ms-win-crt-locale-l1-1-0.dll Summary 1000 .data 1000 .pdata 3000 .rdata 2000 .reloc 1000 .sw5acfn 1000 .sw5asty 1000 .sw5bltn 1000 .sw5cptr 1000 .sw5entr 1000 .sw5flmd 1000 .sw5hash 1000 .sw5mpen 1000 .sw5prt 1000 .sw5prtc 1000 .sw5ratt 1000 .sw5repl 1000 .sw5reps 1000 .sw5rfst 1000 .sw5test 1000 .sw5tymd 1000 .sw5tyrf 3000 .text S:\Program Files\Swift\Toolchains\0.0.0+Asserts\usr\bin>dumpbin /dependents swift-frontend.exe Microsoft (R) COFF/PE Dumper Version 14.42.34436.0 Copyright (C) Microsoft Corporation. All rights reserved. Dump of file swift-frontend.exe File Type: EXECUTABLE IMAGE Image has the following delay load dependencies: SHELL32.dll ole32.dll Image has the following dependencies: mimalloc-redirect.dll mimalloc.dll swiftCore.dll _CompilerSwiftCompilerPluginMessageHandling.dll _CompilerSwiftSyntaxMacroExpansion.dll _CompilerSwiftSyntaxMacros.dll _CompilerSwiftIDEUtils.dll _CompilerSwiftIfConfig.dll _CompilerSwiftOperators.dll _CompilerSwiftSyntaxBuilder.dll _CompilerSwiftParserDiagnostics.dll _CompilerSwiftParser.dll _CompilerSwiftBasicFormat.dll _CompilerSwiftDiagnostics.dll _CompilerSwiftSyntax.dll RPCRT4.dll cmark-gfm.dll VERSION.dll ADVAPI32.dll ntdll.dll KERNEL32.dll OLEAUT32.dll MSVCP140.dll VCRUNTIME140.dll VCRUNTIME140_1.dll api-ms-win-crt-runtime-l1-1-0.dll api-ms-win-crt-heap-l1-1-0.dll api-ms-win-crt-environment-l1-1-0.dll api-ms-win-crt-stdio-l1-1-0.dll api-ms-win-crt-math-l1-1-0.dll api-ms-win-crt-time-l1-1-0.dll api-ms-win-crt-convert-l1-1-0.dll api-ms-win-crt-string-l1-1-0.dll api-ms-win-crt-utility-l1-1-0.dll api-ms-win-crt-filesystem-l1-1-0.dll api-ms-win-crt-locale-l1-1-0.dll _CompilerSwiftLexicalLookup.dll Summary 188000 .data 391000 .pdata 1F4A000 .rdata D7000 .reloc 1000 .rsrc 1000 .sw5acfn 2000 .sw5asty 1000 .sw5bltn 1000 .sw5cptr 9000 .sw5flmd 1000 .sw5hash 1000 .sw5mpen 1000 .sw5prt 2000 .sw5prtc 1000 .sw5ratt 1000 .sw5repl 1000 .sw5reps 6000 .sw5rfst 1000 .sw5test 1000 .sw5tymd 5000 .sw5tyrf 6EE6000 .text

I can also build http://github.com/compnerd/cassowary this way with the above change.

>git clone http://github.com/compnerd/cassowary >cd cassowary >swift build -Xswiftc -explicit-module-build Building for debugging... [23/23] Linking C:\Users\hiroshi\cassowary\.build\x86_64-unknown-windows-msvc\debug\cassowary.dll Build complete! (2.89s)

This also fails like hello.swift without this change.

I don't remember, it was too long ago

In that case, it may be a different bug

@cachemeifyoucan WDYT?

We need to fix -fbuiltin-headers-in-system-modules flag. It is missing from dependency scanner args.

I think it really should't be initialized outside getClangCC1Arguments and it should ideally be inside getClangDriverArguments.

I am still against hacking -sdk here if possible.

@cachemeifyoucan

I assume you mean we want to move the code that adds the -fbuiltin-headers-in-system-modules to getClangDriverArguments? I moved it to addCommonInvocationArguments for which getClangDriverArguments is the only caller and next to the code that adds the ivfsoverlay flag (though we could alternatively have it in getClangDriverArguments). Does that work?

I don't think we need to use -sdk.

test/ScanDependencies/hello-explicit-module-build.swift

hjyamauchi · 2025-03-14T22:52:18Z

@swift-ci please test

hjyamauchi · 2025-03-14T22:52:50Z

@ian-twilightcoder @cachemeifyoucan Please take another look

lib/ClangImporter/ClangImporter.cpp

hjyamauchi · 2025-03-15T00:14:10Z

@swift-ci please test

ian-twilightcoder · 2025-03-18T21:14:00Z

Does this fix #70330 ? Can we re-enable no-implicit-extra-clang-opts.swift?

hjyamauchi · 2025-03-19T00:25:19Z

Does this fix #70330 ? Can we re-enable no-implicit-extra-clang-opts.swift?

Doesn't seem to. Here's the output no-implicit-extra-clang-opts.swift.txt

ian-twilightcoder · 2025-03-19T00:30:48Z

Does this fix #70330 ? Can we re-enable no-implicit-extra-clang-opts.swift?

Doesn't seem to. Here's the output no-implicit-extra-clang-opts.swift.txt

Well that's unfortunate

cachemeifyoucan

LGTM. Thanks!

hjyamauchi · 2025-03-21T16:57:37Z

@ian-twilightcoder Any more thoughts?

hjyamauchi · 2025-03-21T16:58:05Z

@swift-ci please test

lib/ClangImporter/ClangImporter.cpp

hjyamauchi · 2025-03-21T23:08:57Z

@swift-ci please test

ian-twilightcoder · 2025-03-22T00:29:17Z

lib/ClangImporter/ClangImporter.cpp

-    auto swiftTargetClangArgs =
-        getClangCC1Arguments(importer.get(), ctx, VFS, true);
+    auto swiftTargetClangArgs = getClangCC1Arguments(importer.get(), ctx, VFS,
+        /*requiresBuiltinHeadersInSystemModulestrue*/false,


Why do we know this is always false? Do we never use a vfs in this case?

I think this needs to match the getClangCC1Arguments above in the same function.

hjyamauchi · 2025-03-25T03:10:08Z

Please take another look @cachemeifyoucan @ian-twilightcoder

ian-twilightcoder · 2025-03-25T04:40:47Z

lib/ClangImporter/ClangModuleDependencyScanner.cpp

  std::vector<std::string> commandLineArgs =
-      ClangImporter::getClangDriverArguments(ctx);
+      ClangImporter::getClangDriverArguments(ctx,
+          fileMapping.requiresBuiltinHeadersInSystemModules);


I don't understand why we need to set -fbuiltin-headers-in-system-modules here when we aren't actually using the file mapping. It's the file mapping that determines whether you need -fbuiltin-headers-in-system-modules. @cachemeifyoucan is the scanner picking up the Swift provided module maps in another spot?

Can you explain a bit more about your concern? I am not quite sure what you mean by not using the file mapping. All the file mapping is important during scanning to provide the correct module while the main clang importer will only load modules and PCH (and currently very little header file) so the file mapping has little impact there, just need to match so there isn't module loading error.

If I unconditionally pass requiresBuiltinHeadersInSystemModules=false here, hello.swift fails to build due to a diag::err_pch_langopt_mismatch of BuiltinHeadersInSystemModulesmismatch detected in the ASTReader. I think it's the mismatch between the AST writing time and the AST reading time that we discussed above.

At this spot in the code, we're getting the file mapping and then throwing it away without adding any vfs arguments from it. Is there some other spot in the dependency scanner where it gets and uses the file mapping vfs?

Most VFS is configured into the VirtualFileSystem and passed to scanner directly. But if the VFS setting is affecting reader writer configuration, it will cause problem.

Would simply adding an internal/private field in the ClangImporter class or the ClangImporter::Implementation class work, as opposed to part of ClangImporterOptions or ExtraArgs? We just need to save a bool value from ClangImporter::create and later read it in ClangImporter::getHeaderDependencies and ClangImporter::getModuleDependencies (the two callers of getClangDepScanningInvocationArguments) and then pass it here to getClangDepScanningInvocationArguments, no?

Ah I missed the subtlety that the callers were ClangImporter since they're in ClangModuleDependencyScanner.cpp. Yeah I think an internal/private field would work, and then we wouldn't have to pass it as parameters all over the place too 😄

Well, I guess we'd have to make getClangDriverArguments/getClangCC1Arguments non-static to not have to pass it as a parameter. Maybe it's worth it?

Yeah. Updated accordingly. I added a private field. I made getClangDriverArguments/getClangCC1Arguments non-static, which led to additional simplifications because getClangCC1Arguments already took an importer instance as the first argument. I also ended up making getClangDepScanningInvocationArguments an instance member function as it also took an importer instance as the first argument and it looked kind of similar to getClangDriverArguments/getClangCC1Arguments. Thoughts?

This looks great to me, thanks for sticking with it!

hjyamauchi · 2025-03-27T20:40:09Z

@swift-ci please test

This fixes explicit module builds for a hello world program on Windows as well as the ucrt import build failure as in the included test.

hjyamauchi · 2025-03-28T16:15:32Z

Note: the CI passed at the current revision of this PR

hjyamauchi · 2025-03-28T17:54:29Z

@swift-ci please test

hjyamauchi · 2025-03-28T20:32:55Z

@swift-ci please test macos platform

hjyamauchi force-pushed the explicit-module-build branch from b3c60b5 to ebbdc71 Compare March 11, 2025 22:10

hjyamauchi force-pushed the explicit-module-build branch from ebbdc71 to 084c0e5 Compare March 12, 2025 17:57

hjyamauchi marked this pull request as ready for review March 13, 2025 15:37

hjyamauchi requested review from artemcm, zoecarver, hyp, egorzhdan, beccadax, ian-twilightcoder, Xazax-hun, j-hui and fahadnayyar as code owners March 13, 2025 15:37

hjyamauchi requested review from cachemeifyoucan and compnerd March 13, 2025 15:37

compnerd mentioned this pull request Mar 13, 2025

utils: remove swiftinterfaces from the build image #79840

Closed

hjyamauchi mentioned this pull request Mar 13, 2025

[scanner] setup the clang importer injected redirecting overlay vfs f… #78184

Open

ian-twilightcoder reviewed Mar 13, 2025

View reviewed changes

lib/ClangImporter/ClangImporter.cpp Outdated Show resolved Hide resolved

hjyamauchi force-pushed the explicit-module-build branch from 084c0e5 to 7c1f248 Compare March 13, 2025 19:09

hjyamauchi requested a review from ian-twilightcoder March 13, 2025 19:11

cachemeifyoucan reviewed Mar 13, 2025

View reviewed changes

hjyamauchi force-pushed the explicit-module-build branch from 7c1f248 to d5a35ff Compare March 14, 2025 22:50

hjyamauchi requested a review from cachemeifyoucan March 14, 2025 22:52

ian-twilightcoder requested changes Mar 14, 2025

View reviewed changes

lib/ClangImporter/ClangImporter.cpp Outdated Show resolved Hide resolved

hjyamauchi force-pushed the explicit-module-build branch from d5a35ff to 87b19aa Compare March 15, 2025 00:13

hjyamauchi force-pushed the explicit-module-build branch from 87b19aa to 0404200 Compare March 18, 2025 17:14

hjyamauchi force-pushed the explicit-module-build branch 2 times, most recently from 3f9160b to 24dd1cf Compare March 21, 2025 16:06

hjyamauchi requested a review from susmonteiro as a code owner March 21, 2025 16:06

cachemeifyoucan approved these changes Mar 21, 2025

View reviewed changes

ian-twilightcoder reviewed Mar 21, 2025

View reviewed changes

lib/ClangImporter/ClangImporter.cpp Show resolved Hide resolved

hjyamauchi force-pushed the explicit-module-build branch from 24dd1cf to 8698e59 Compare March 21, 2025 20:57

ian-twilightcoder reviewed Mar 22, 2025

View reviewed changes

hjyamauchi force-pushed the explicit-module-build branch from 8698e59 to 89083b7 Compare March 25, 2025 03:06

hjyamauchi changed the title ~~Propagate vfs overlays and sdk to subinvocations~~ Propagate vfs overlays and -fbuiltin-headers-in-system-modules Mar 25, 2025

ian-twilightcoder reviewed Mar 25, 2025

View reviewed changes

Propagate vfs overlays and -fbuiltin-headers-in-system-modules

8312f7a

This fixes explicit module builds for a hello world program on Windows as well as the ucrt import build failure as in the included test.

hjyamauchi force-pushed the explicit-module-build branch from 89083b7 to 8312f7a Compare March 28, 2025 16:16

ian-twilightcoder approved these changes Mar 28, 2025

View reviewed changes

hjyamauchi enabled auto-merge March 28, 2025 20:52

hjyamauchi merged commit 42fe7d9 into swiftlang:main Mar 29, 2025
5 checks passed

Propagate vfs overlays and -fbuiltin-headers-in-system-modules #79926

Propagate vfs overlays and -fbuiltin-headers-in-system-modules #79926

Conversation

hjyamauchi commented Mar 11, 2025

hjyamauchi commented Mar 11, 2025 • edited Loading

hjyamauchi commented Mar 11, 2025

hjyamauchi commented Mar 11, 2025

hjyamauchi commented Mar 11, 2025

hjyamauchi commented Mar 12, 2025

hjyamauchi commented Mar 12, 2025

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

hjyamauchi Mar 14, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

hjyamauchi commented Mar 14, 2025

hjyamauchi commented Mar 14, 2025

hjyamauchi commented Mar 15, 2025

ian-twilightcoder commented Mar 18, 2025

hjyamauchi commented Mar 19, 2025

ian-twilightcoder commented Mar 19, 2025

cachemeifyoucan left a comment

Choose a reason for hiding this comment

hjyamauchi commented Mar 21, 2025

hjyamauchi commented Mar 21, 2025

hjyamauchi commented Mar 21, 2025

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

hjyamauchi commented Mar 25, 2025

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

hjyamauchi Mar 27, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

hjyamauchi Mar 28, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

hjyamauchi commented Mar 27, 2025

hjyamauchi commented Mar 28, 2025

hjyamauchi commented Mar 28, 2025

hjyamauchi commented Mar 28, 2025

hjyamauchi commented Mar 11, 2025 •

edited

Loading

hjyamauchi Mar 14, 2025 •

edited

Loading

hjyamauchi Mar 27, 2025 •

edited

Loading

hjyamauchi Mar 28, 2025 •

edited

Loading