[clang][modules] Build inferred modules

[jansvoboda11]

Summary

This patch enables explicitly building inferred modules.

Effectively a cherry-pick of https://github.com/apple/llvm-project/pull/699 authored by @Bigcheese with libclang and dependency scanner changes omitted.

Contains the following changes:

1. [Clang] Fix the header paths in clang::Module for inferred modules.
   • The UmbrellaAsWritten and NameAsWritten fields in clang::Module are a lie for framework modules. For those they actually are the path to the header or umbrella relative to the clang::Module::Directory.
   • The exception to this case is for inferred modules. Here it actually is the name as written, because we print out the module and read it back in when implicitly building modules. This causes a problem when explicitly building an inferred module, as we skip the printing out step.
   • In order to fix this issue this patch adds a new field for the path we want to use in getInputBufferForModule. It also makes NameAsWritten actually be the name written in the module map file (or that would be, in the case of an inferred module).

2. [Clang] Allow explicitly building an inferred module.
   • Building the actual module still fails, but make sure it fails for the right reason.

Split from D100934.

[dexonsmith]

Looks great — thanks for splitting this out!

clang/include/clang/Basic/Module.h

public:
   llvm::PointerUnion<const FileEntry *, const DirectoryEntry *> Umbrella;
 
   /// The module signature.
   ASTFileSignature Signature;
 
   /// The name of the umbrella entry, as written in the module map.
   std::string UmbrellaAsWritten;
 
+  // The path to the umbrella entry relative to the root module's \c Directory.
+  std::string UmbrellaRelativeToRootModuleDirectory;
+
   /// The module through which entities defined in this module will
   /// eventually be exposed, for use in "private" modules.
   std::string ExportAsModule;
 
   /// Does this Module scope describe part of the purview of a named C++ module?
   bool isModulePurview() const {
     return Kind == ModuleInterfaceUnit || Kind == PrivateModuleFragment;
   }
enum HeaderKind {
     HK_Excluded
   };
   static const int NumHeaderKinds = HK_Excluded + 1;
 
   /// Information about a header directive as found in the module map
   /// file.
   struct Header {
     std::string NameAsWritten;
+    std::string PathRelativeToRootModuleDirectory;
     const FileEntry *Entry;
 
     explicit operator bool() { return Entry; }
   };
 
   /// Information about a directory name as found in the module map
   /// file.
   struct DirectoryName {
     std::string NameAsWritten;
+    std::string PathRelativeToRootModuleDirectory;
     const DirectoryEntry *Entry;
 
     explicit operator bool() { return Entry; }
   };
 
   /// The headers that are part of this module.
   SmallVector<Header, 2> Headers[5];
 
public:
   /// Retrieve the directory for which this module serves as the
   /// umbrella.
   DirectoryName getUmbrellaDir() const;
 
   /// Retrieve the header that serves as the umbrella header for this
   /// module.
   Header getUmbrellaHeader() const {
     if (auto *FE = Umbrella.dyn_cast<const FileEntry *>())
-      return Header{UmbrellaAsWritten, FE};
+      return Header{UmbrellaAsWritten, UmbrellaRelativeToRootModuleDirectory,
+                    FE};
     return Header{};
   }
 
   /// Determine whether this module has an umbrella directory that is
   /// not based on an umbrella header.
   bool hasUmbrellaDir() const {
     return Umbrella && Umbrella.is<const DirectoryEntry *>();
   }

clang/include/clang/Lex/ModuleMap.h

public:
   ///
   /// \returns true if any errors were encountered while resolving conflicts,
   /// false otherwise.
   bool resolveConflicts(Module *Mod, bool Complain);
 
   /// Sets the umbrella header of the given module to the given
   /// header.
   void setUmbrellaHeader(Module *Mod, const FileEntry *UmbrellaHeader,
-                         Twine NameAsWritten);
+                         const Twine &NameAsWritten,
+                         const Twine &PathRelativeToRootModuleDirectory);
 
   /// Sets the umbrella directory of the given module to the given
   /// directory.
   void setUmbrellaDir(Module *Mod, const DirectoryEntry *UmbrellaDir,
-                      Twine NameAsWritten);
+                      const Twine &NameAsWritten,
+                      const Twine &PathRelativeToRootModuleDirectory);
 
   /// Adds this header to the given module.
   /// \param Role The role of the header wrt the module.
   void addHeader(Module *Mod, Module::Header Header,
                  ModuleHeaderRole Role, bool Imported = false);
 
   /// Marks this header as being excluded from the given module.
   void excludeHeader(Module *Mod, Module::Header Header);

clang/lib/Basic/Module.cpp

if (nameParts.empty() || M->Name != nameParts.back())
       return false;
     nameParts = nameParts.drop_back();
   }
   return nameParts.empty();
 }
 
 Module::DirectoryName Module::getUmbrellaDir() const {
   if (Header U = getUmbrellaHeader())
-    return {"", U.Entry->getDir()};
+    return {"", "", U.Entry->getDir()};
 
-  return {UmbrellaAsWritten, Umbrella.dyn_cast<const DirectoryEntry *>()};
+  return {UmbrellaAsWritten, UmbrellaRelativeToRootModuleDirectory,
+          Umbrella.dyn_cast<const DirectoryEntry *>()};
 }
 
 void Module::addTopHeader(const FileEntry *File) {
   assert(File);
   TopHeaders.insert(File);
 }
 
 ArrayRef<const FileEntry *> Module::getTopHeaders(FileManager &FileMgr) {

clang/lib/Frontend/FrontendAction.cpp

static std::error_code collectModuleHeaderIncludes(
   // Add includes for each of these headers.
   for (auto HK : {Module::HK_Normal, Module::HK_Private}) {
     for (Module::Header &H : Module->Headers[HK]) {
       Module->addTopHeader(H.Entry);
       // Use the path as specified in the module map file. We'll look for this
       // file relative to the module build directory (the directory containing
       // the module map file) so this will find the same file that we found
       // while parsing the module map.
-      addHeaderInclude(H.NameAsWritten, Includes, LangOpts, Module->IsExternC);
+      addHeaderInclude(H.PathRelativeToRootModuleDirectory, Includes, LangOpts,
+                       Module->IsExternC);
     }
   }
   // Note that Module->PrivateHeaders will not be a TopHeader.
 
   if (Module::Header UmbrellaHeader = Module->getUmbrellaHeader()) {
     Module->addTopHeader(UmbrellaHeader.Entry);
     if (Module->Parent)
       // Include the umbrella header for submodules.
-      addHeaderInclude(UmbrellaHeader.NameAsWritten, Includes, LangOpts,
-                       Module->IsExternC);
+      addHeaderInclude(UmbrellaHeader.PathRelativeToRootModuleDirectory,
+                       Includes, LangOpts, Module->IsExternC);
   } else if (Module::DirectoryName UmbrellaDir = Module->getUmbrellaDir()) {
     // Add all of the headers we find in this subdirectory.
     std::error_code EC;
     SmallString<128> DirNative;
     llvm::sys::path::native(UmbrellaDir.Entry->getName(), DirNative);
 
     llvm::vfs::FileSystem &FS = FileMgr.getVirtualFileSystem();
     SmallVector<std::pair<std::string, const FileEntry *>, 8> Headers;
for (llvm::vfs::recursive_directory_iterator Dir(FS, DirNative, EC), End;
       if (ModMap.isHeaderUnavailableInModule(*Header, Module))
         continue;
 
       // Compute the relative path from the directory to this file.
       SmallVector<StringRef, 16> Components;
       auto PathIt = llvm::sys::path::rbegin(Dir->path());
       for (int I = 0; I != Dir.level() + 1; ++I, ++PathIt)
         Components.push_back(*PathIt);
-      SmallString<128> RelativeHeader(UmbrellaDir.NameAsWritten);
+      SmallString<128> RelativeHeader(
+          UmbrellaDir.PathRelativeToRootModuleDirectory);
       for (auto It = Components.rbegin(), End = Components.rend(); It != End;
            ++It)
         llvm::sys::path::append(RelativeHeader, *It);
 
       std::string RelName = RelativeHeader.c_str();
       Headers.push_back(std::make_pair(RelName, *Header));
     }
 
if (CI.getLangOpts().CurrentModule.empty()) {
     // default. Then it would be fairly trivial to just "compile" a module
     // map with a single module (the common case).
     return nullptr;
   }
 
   // Dig out the module definition.
   HeaderSearch &HS = CI.getPreprocessor().getHeaderSearchInfo();
   Module *M = HS.lookupModule(CI.getLangOpts().CurrentModule,
-                              /*AllowSearch=*/false);
+                              /*AllowSearch=*/true);
   if (!M) {
     CI.getDiagnostics().Report(diag::err_missing_module)
       << CI.getLangOpts().CurrentModule << ModuleMapFilename;
 
     return nullptr;
   }
 
   // Check whether we can build this module at all.

 static std::unique_ptr<llvm::MemoryBuffer>
 getInputBufferForModule(CompilerInstance &CI, Module *M) {
   FileManager &FileMgr = CI.getFileManager();
 
   // Collect the set of #includes we need to build the module.
   SmallString<256> HeaderContents;
   std::error_code Err = std::error_code();
   if (Module::Header UmbrellaHeader = M->getUmbrellaHeader())
-    addHeaderInclude(UmbrellaHeader.NameAsWritten, HeaderContents,
-                     CI.getLangOpts(), M->IsExternC);
+    addHeaderInclude(UmbrellaHeader.PathRelativeToRootModuleDirectory,
+                     HeaderContents, CI.getLangOpts(), M->IsExternC);
   Err = collectModuleHeaderIncludes(
       CI.getLangOpts(), FileMgr, CI.getDiagnostics(),
       CI.getPreprocessor().getHeaderSearchInfo().getModuleMap(), M,
       HeaderContents);
 
   if (Err) {
     CI.getDiagnostics().Report(diag::err_module_cannot_create_includes)
       << M->getFullModuleName() << Err.message();

clang/lib/Frontend/FrontendActions.cpp

for (StringRef Name : ModuleHeaders) {
     Optional<FileEntryRef> FE = HS.LookupFile(
         Name, SourceLocation(), /*Angled*/ false, nullptr, CurDir, None,
         nullptr, nullptr, nullptr, nullptr, nullptr, nullptr);
     if (!FE) {
       CI.getDiagnostics().Report(diag::err_module_header_file_not_found)
         << Name;
       continue;
     }
-    Headers.push_back({std::string(Name), &FE->getFileEntry()});
+    Headers.push_back(
+        {std::string(Name), std::string(Name), &FE->getFileEntry()});
   }
   HS.getModuleMap().createHeaderModule(CI.getLangOpts().CurrentModule, Headers);
 
   return GenerateModuleAction::BeginSourceFileAction(CI);
 }
 
 std::unique_ptr<raw_pwrite_stream>
 GenerateHeaderModuleAction::CreateOutputFile(CompilerInstance &CI,

clang/lib/Lex/ModuleMap.cpp

if (Optional<FileEntryRef> File =
           findHeader(Mod, Header, RelativePathName, NeedsFramework)) {
     if (Header.IsUmbrella) {
       const DirectoryEntry *UmbrellaDir = &File->getDir().getDirEntry();
       if (Module *UmbrellaMod = UmbrellaDirs[UmbrellaDir])
         Diags.Report(Header.FileNameLoc, diag::err_mmap_umbrella_clash)
           << UmbrellaMod->getFullModuleName();
       else
         // Record this umbrella header.
-        setUmbrellaHeader(Mod, *File, RelativePathName.str());
+        setUmbrellaHeader(Mod, *File, Header.FileName, RelativePathName.str());
     } else {
-      Module::Header H = {std::string(RelativePathName.str()), *File};
+      Module::Header H = {Header.FileName, std::string(RelativePathName.str()),
+                          *File};
       if (Header.Kind == Module::HK_Excluded)
         excludeHeader(Mod, H);
       else
         addHeader(Mod, H, headerKindToRole(Header.Kind));
     }
   } else if (Header.HasBuiltinHeader && !Header.Size && !Header.ModTime) {
     // There's a builtin header but no corresponding on-disk header. Assume
     // this was supposed to modularize the builtin header alone.
bool ModuleMap::resolveAsBuiltinHeader(
   // supplied by Clang. Find that builtin header.
   SmallString<128> Path;
   llvm::sys::path::append(Path, BuiltinIncludeDir->getName(), Header.FileName);
   auto File = SourceMgr.getFileManager().getFile(Path);
   if (!File)
     return false;
 
   auto Role = headerKindToRole(Header.Kind);
-  Module::Header H = {std::string(Path.str()), *File};
+  Module::Header H = {Header.FileName, std::string(Path.str()), *File};
   addHeader(Mod, H, Role);
   return true;
 }
 
 ModuleMap::ModuleMap(SourceManager &SourceMgr, DiagnosticsEngine &Diags,
                      const LangOptions &LangOpts, const TargetInfo *Target,
                      HeaderSearch &HeaderInfo)
     : SourceMgr(SourceMgr), Diags(Diags), LangOpts(LangOpts), Target(Target),
Module *ModuleMap::inferFrameworkModule(const DirectoryEntry *FrameworkDir,
   }
 
   Result->IsSystem |= Attrs.IsSystem;
   Result->IsExternC |= Attrs.IsExternC;
   Result->ConfigMacrosExhaustive |= Attrs.IsExhaustive;
   Result->NoUndeclaredIncludes |= Attrs.NoUndeclaredIncludes;
   Result->Directory = FrameworkDir;
 
+  // Chop off the first framework bit, as that is implied.
+  StringRef RelativePath = UmbrellaName.str().substr(
+      Result->getTopLevelModule()->Directory->getName().size());
+  RelativePath = llvm::sys::path::relative_path(RelativePath);
+
   // umbrella header "umbrella-header-name"
-  //
-  // The "Headers/" component of the name is implied because this is
-  // a framework module.
-  setUmbrellaHeader(Result, *UmbrellaHeader, ModuleName + ".h");
+  setUmbrellaHeader(Result, *UmbrellaHeader, ModuleName + ".h", RelativePath);
 
   // export *
   Result->Exports.push_back(Module::ExportDecl(nullptr, true));
 
   // module * { export * }
   Result->InferSubmodules = true;
   Result->InferExportWildcard = true;
 
Module *ModuleMap::createShadowedModule(StringRef Name, bool IsFramework,
   Result->ShadowingModule = ShadowingModule;
   Result->markUnavailable(/*Unimportable*/true);
   ModuleScopeIDs[Result] = CurrentModuleScopeID;
   ShadowModules.push_back(Result);
 
   return Result;
 }
 
-void ModuleMap::setUmbrellaHeader(Module *Mod, const FileEntry *UmbrellaHeader,
-                                  Twine NameAsWritten) {
+void ModuleMap::setUmbrellaHeader(
+    Module *Mod, const FileEntry *UmbrellaHeader, const Twine &NameAsWritten,
+    const Twine &PathRelativeToRootModuleDirectory) {
   Headers[UmbrellaHeader].push_back(KnownHeader(Mod, NormalHeader));
   Mod->Umbrella = UmbrellaHeader;
   Mod->UmbrellaAsWritten = NameAsWritten.str();
+  Mod->UmbrellaRelativeToRootModuleDirectory =
+      PathRelativeToRootModuleDirectory.str();
   UmbrellaDirs[UmbrellaHeader->getDir()] = Mod;
 
   // Notify callbacks that we just added a new header.
   for (const auto &Cb : Callbacks)
     Cb->moduleMapAddUmbrellaHeader(&SourceMgr.getFileManager(), UmbrellaHeader);
 }
 
 void ModuleMap::setUmbrellaDir(Module *Mod, const DirectoryEntry *UmbrellaDir,
-                               Twine NameAsWritten) {
+                               const Twine &NameAsWritten,
+                               const Twine &PathRelativeToRootModuleDirectory) {
   Mod->Umbrella = UmbrellaDir;
   Mod->UmbrellaAsWritten = NameAsWritten.str();
+  Mod->UmbrellaRelativeToRootModuleDirectory =
+      PathRelativeToRootModuleDirectory.str();
   UmbrellaDirs[UmbrellaDir] = Mod;
 }
 
 void ModuleMap::addUnresolvedHeader(Module *Mod,
                                     Module::UnresolvedHeaderDirective Header,
                                     bool &NeedsFramework) {
   // If there is a builtin counterpart to this file, add it now so it can
   // wrap the system header.
void ModuleMapParser::parseUmbrellaDirDecl(SourceLocation UmbrellaLoc) {
   if (!Tok.is(MMToken::StringLiteral)) {
     Diags.Report(Tok.getLocation(), diag::err_mmap_expected_header)
       << "umbrella";
     HadError = true;
     return;
   }
 
   std::string DirName = std::string(Tok.getString());
+  std::string DirNameAsWritten = DirName;
   SourceLocation DirNameLoc = consumeToken();
 
   // Check whether we already have an umbrella.
   if (ActiveModule->Umbrella) {
     Diags.Report(DirNameLoc, diag::err_mmap_umbrella_clash)
       << ActiveModule->getFullModuleName();
     HadError = true;
     return;
if (UsesRequiresExcludedHack.count(ActiveModule)) {
     // uncommonly used Tcl module on Darwin platforms.
     std::error_code EC;
     SmallVector<Module::Header, 6> Headers;
     llvm::vfs::FileSystem &FS =
         SourceMgr.getFileManager().getVirtualFileSystem();
     for (llvm::vfs::recursive_directory_iterator I(FS, Dir->getName(), EC), E;
          I != E && !EC; I.increment(EC)) {
       if (auto FE = SourceMgr.getFileManager().getFile(I->path())) {
-        Module::Header Header = {std::string(I->path()), *FE};
+        Module::Header Header = {"", std::string(I->path()), *FE};
         Headers.push_back(std::move(Header));
       }
     }
 
     // Sort header paths so that the pcm doesn't depend on iteration order.
     llvm::array_pod_sort(Headers.begin(), Headers.end(), compareModuleHeaders);
 
     for (auto &Header : Headers)
       Map.addHeader(ActiveModule, std::move(Header), ModuleMap::TextualHeader);
     return;
   }
 
   if (Module *OwningModule = Map.UmbrellaDirs[Dir]) {
     Diags.Report(UmbrellaLoc, diag::err_mmap_umbrella_clash)
       << OwningModule->getFullModuleName();
     HadError = true;
     return;
   }
 
   // Record this umbrella directory.
-  Map.setUmbrellaDir(ActiveModule, Dir, DirName);
+  Map.setUmbrellaDir(ActiveModule, Dir, DirNameAsWritten, DirName);
 }
 
 /// Parse a module export declaration.
 ///
 ///   export-declaration:
 ///     'export' wildcard-module-id
 ///
 ///   wildcard-module-id:

clang/lib/Serialization/ASTReader.cpp

while (d != End) {
     Module *Mod = Reader.getSubmodule(GlobalSMID);
     FileManager &FileMgr = Reader.getFileManager();
     ModuleMap &ModMap =
         Reader.getPreprocessor().getHeaderSearchInfo().getModuleMap();
 
     std::string Filename = std::string(key.Filename);
     if (key.Imported)
       Reader.ResolveImportedPath(M, Filename);
-    // FIXME: This is not always the right filename-as-written, but we're not
-    // going to use this information to rebuild the module, so it doesn't make
-    // a lot of difference.
-    Module::Header H = {std::string(key.Filename), *FileMgr.getFile(Filename)};
+    // FIXME: NameAsWritten
+    Module::Header H = {std::string(key.Filename), "",
+                        *FileMgr.getFile(Filename)};
     ModMap.addHeader(Mod, H, HeaderRole, /*Imported*/true);
     HFI.isModuleHeader |= !(HeaderRole & ModuleMap::TextualHeader);
   }
 
   // This HeaderFileInfo was externally loaded.
   HFI.External = true;
   HFI.IsValid = true;
   return HFI;
case SUBMODULE_DEFINITION: {
       break;
     }
 
     case SUBMODULE_UMBRELLA_HEADER: {
       std::string Filename = std::string(Blob);
       ResolveImportedPath(F, Filename);
       if (auto Umbrella = PP.getFileManager().getFile(Filename)) {
         if (!CurrentModule->getUmbrellaHeader())
-          ModMap.setUmbrellaHeader(CurrentModule, *Umbrella, Blob);
+          // FIXME: NameAsWritten
+          ModMap.setUmbrellaHeader(CurrentModule, *Umbrella, Blob, "");
         else if (CurrentModule->getUmbrellaHeader().Entry != *Umbrella) {
           if ((ClientLoadCapabilities & ARR_OutOfDate) == 0)
             Error("mismatched umbrella headers in submodule");
           return OutOfDate;
         }
       }
       break;
     }
case SUBMODULE_TOPHEADER:
       CurrentModule->addTopHeaderFilename(Blob);
       break;
 
     case SUBMODULE_UMBRELLA_DIR: {
       std::string Dirname = std::string(Blob);
       ResolveImportedPath(F, Dirname);
       if (auto Umbrella = PP.getFileManager().getDirectory(Dirname)) {
         if (!CurrentModule->getUmbrellaDir())
-          ModMap.setUmbrellaDir(CurrentModule, *Umbrella, Blob);
+          // FIXME: NameAsWritten
+          ModMap.setUmbrellaDir(CurrentModule, *Umbrella, Blob, "");
         else if (CurrentModule->getUmbrellaDir().Entry != *Umbrella) {
           if ((ClientLoadCapabilities & ARR_OutOfDate) == 0)
             Error("mismatched umbrella directories in submodule");
           return OutOfDate;
         }
       }
       break;
     }

clang/test/Modules/Inputs/explicit-build-inferred/frameworks/Inferred.framework/Headers/Inferred.h

+typedef int inferred;

clang/test/Modules/Inputs/explicit-build-inferred/frameworks/module.modulemap

+framework module * {}

clang/test/Modules/explicit-build-inferred.cpp

+// RUN: rm -rf %t && mkdir %t
+//
+// RUN: %clang_cc1 -fmodules -fno-implicit-modules -fimplicit-module-maps \
+// RUN:   -emit-module -x c++ %S/Inputs/explicit-build-inferred/frameworks/module.modulemap \
+// RUN:   -fmodule-name=Inferred -o %t/Inferred.pcm -F %S/Inputs/explicit-build-inferred/frameworks
+//
+// RUN: %clang_cc1 -fmodules -fno-implicit-modules -fsyntax-only %s \
+// RUN:   -fmodule-map-file=%S/Inputs/explicit-build-inferred/frameworks/module.modulemap \
+// RUN:   -fmodule-file=%t/Inferred.pcm -F %S/Inputs/explicit-build-inferred/frameworks
+
+#include <Inferred/Inferred.h>
+
+inferred a = 0;