Merge pull request #8571 from apple/jan_svoboda/cache-real-path

[clang][deps] Cache `VFS::getRealPath()` and `VFS::exists()`

Author: nkcsgexi

clang/include/clang/Tooling/DependencyScanning/DependencyScanningCASFilesystem.h

@@ -52,7 +52,7 @@ class DependencyScanningCASFilesystem : public llvm::cas::ThreadSafeFileSystem {
     return FS->setCurrentWorkingDirectory(Path);
   }
   std::error_code getRealPath(const Twine &Path,
-                              SmallVectorImpl<char> &Output) const override {
+                              SmallVectorImpl<char> &Output) override {
     return FS->getRealPath(Path, Output);
   }
   std::error_code isLocal(const Twine &Path, bool &Result) override {

clang/include/clang/Tooling/DependencyScanning/DependencyScanningFilesystem.h

@@ -154,6 +154,8 @@ class CachedFileSystemEntry {
   CachedFileContents *Contents;
 };
 
+using CachedRealPath = llvm::ErrorOr<std::string>;
+
 /// This class is a shared cache, that caches the 'stat' and 'open' calls to the
 /// underlying real file system, and the scanned preprocessor directives of
 /// files.
@@ -166,9 +168,11 @@ class DependencyScanningFilesystemSharedCache {
     /// The mutex that needs to be locked before mutation of any member.
     mutable std::mutex CacheLock;
 
-    /// Map from filenames to cached entries.
-    llvm::StringMap<const CachedFileSystemEntry *, llvm::BumpPtrAllocator>
-        EntriesByFilename;
+    /// Map from filenames to cached entries and real paths.
+    llvm::StringMap<
+        std::pair<const CachedFileSystemEntry *, const CachedRealPath *>,
+        llvm::BumpPtrAllocator>
+        CacheByFilename;
 
     /// Map from unique IDs to cached entries.
     llvm::DenseMap<llvm::sys::fs::UniqueID, const CachedFileSystemEntry *>
@@ -180,6 +184,9 @@ class DependencyScanningFilesystemSharedCache {
     /// The backing storage for cached contents.
     llvm::SpecificBumpPtrAllocator<CachedFileContents> ContentsStorage;
 
+    /// The backing storage for cached real paths.
+    llvm::SpecificBumpPtrAllocator<CachedRealPath> RealPathStorage;
+
     /// Returns entry associated with the filename or nullptr if none is found.
     const CachedFileSystemEntry *findEntryByFilename(StringRef Filename) const;
 
@@ -207,6 +214,17 @@ class DependencyScanningFilesystemSharedCache {
     const CachedFileSystemEntry &
     getOrInsertEntryForFilename(StringRef Filename,
                                 const CachedFileSystemEntry &Entry);
+
+    /// Returns the real path associated with the filename or nullptr if none is
+    /// found.
+    const CachedRealPath *findRealPathByFilename(StringRef Filename) const;
+
+    /// Returns the real path associated with the filename if there is some.
+    /// Otherwise, constructs new one with the given one, associates it with the
+    /// filename and returns the result.
+    const CachedRealPath &
+    getOrEmplaceRealPathForFilename(StringRef Filename,
+                                    llvm::ErrorOr<StringRef> RealPath);
   };
 
   DependencyScanningFilesystemSharedCache();
@@ -223,14 +241,17 @@ class DependencyScanningFilesystemSharedCache {
 /// This class is a local cache, that caches the 'stat' and 'open' calls to the
 /// underlying real file system.
 class DependencyScanningFilesystemLocalCache {
-  llvm::StringMap<const CachedFileSystemEntry *, llvm::BumpPtrAllocator> Cache;
+  llvm::StringMap<
+      std::pair<const CachedFileSystemEntry *, const CachedRealPath *>,
+      llvm::BumpPtrAllocator>
+      Cache;
 
 public:
   /// Returns entry associated with the filename or nullptr if none is found.
   const CachedFileSystemEntry *findEntryByFilename(StringRef Filename) const {
     assert(llvm::sys::path::is_absolute_gnu(Filename));
     auto It = Cache.find(Filename);
-    return It == Cache.end() ? nullptr : It->getValue();
+    return It == Cache.end() ? nullptr : It->getValue().first;
   }
 
   /// Associates the given entry with the filename and returns the given entry
@@ -239,9 +260,40 @@ class DependencyScanningFilesystemLocalCache {
   insertEntryForFilename(StringRef Filename,
                          const CachedFileSystemEntry &Entry) {
     assert(llvm::sys::path::is_absolute_gnu(Filename));
-    const auto *InsertedEntry = Cache.insert({Filename, &Entry}).first->second;
-    assert(InsertedEntry == &Entry && "entry already present");
-    return *InsertedEntry;
+    auto [It, Inserted] = Cache.insert({Filename, {&Entry, nullptr}});
+    auto &[CachedEntry, CachedRealPath] = It->getValue();
+    if (!Inserted) {
+      // The file is already present in the local cache. If we got here, it only
+      // contains the real path. Let's make sure the entry is populated too.
+      assert((!CachedEntry && CachedRealPath) && "entry already present");
+      CachedEntry = &Entry;
+    }
+    return *CachedEntry;
+  }
+
+  /// Returns real path associated with the filename or nullptr if none is
+  /// found.
+  const CachedRealPath *findRealPathByFilename(StringRef Filename) const {
+    assert(llvm::sys::path::is_absolute_gnu(Filename));
+    auto It = Cache.find(Filename);
+    return It == Cache.end() ? nullptr : It->getValue().second;
+  }
+
+  /// Associates the given real path with the filename and returns the given
+  /// entry pointer (for convenience).
+  const CachedRealPath &
+  insertRealPathForFilename(StringRef Filename,
+                            const CachedRealPath &RealPath) {
+    assert(llvm::sys::path::is_absolute_gnu(Filename));
+    auto [It, Inserted] = Cache.insert({Filename, {nullptr, &RealPath}});
+    auto &[CachedEntry, CachedRealPath] = It->getValue();
+    if (!Inserted) {
+      // The file is already present in the local cache. If we got here, it only
+      // contains the entry. Let's make sure the real path is populated too.
+      assert((!CachedRealPath && CachedEntry) && "real path already present");
+      CachedRealPath = &RealPath;
+    }
+    return *CachedRealPath;
   }
 };
 
@@ -312,6 +364,9 @@ class DependencyScanningWorkerFilesystem
   llvm::ErrorOr<std::unique_ptr<llvm::vfs::File>>
   openFileForRead(const Twine &Path) override;
 
+  std::error_code getRealPath(const Twine &Path,
+                              SmallVectorImpl<char> &Output) override;
+
   std::error_code setCurrentWorkingDirectory(const Twine &Path) override;
 
   /// Returns entry for the given filename.
@@ -326,6 +381,10 @@ class DependencyScanningWorkerFilesystem
   /// false if not (i.e. this entry is not a file or its scan fails).
   bool ensureDirectiveTokensArePopulated(EntryRef Entry);
 
+  /// Check whether \p Path exists. By default checks cached result of \c
+  /// status(), and falls back on FS if unable to do so.
+  bool exists(const Twine &Path) override;
+
 private:
   /// For a filename that's not yet associated with any entry in the caches,
   /// uses the underlying filesystem to either look up the entry based in the
@@ -421,6 +480,10 @@ class DependencyScanningWorkerFilesystem
   llvm::ErrorOr<std::string> WorkingDirForCacheLookup;
 
   void updateWorkingDirForCacheLookup();
+
+  llvm::ErrorOr<StringRef>
+  tryGetFilenameForLookup(StringRef OriginalFilename,
+                          llvm::SmallVectorImpl<char> &PathBuf) const;
 };
 
 } // end namespace dependencies

clang/lib/Tooling/DependencyScanning/DependencyScanningFilesystem.cpp

@@ -118,8 +118,8 @@ DependencyScanningFilesystemSharedCache::CacheShard::findEntryByFilename(
     StringRef Filename) const {
   assert(llvm::sys::path::is_absolute_gnu(Filename));
   std::lock_guard<std::mutex> LockGuard(CacheLock);
-  auto It = EntriesByFilename.find(Filename);
-  return It == EntriesByFilename.end() ? nullptr : It->getValue();
+  auto It = CacheByFilename.find(Filename);
+  return It == CacheByFilename.end() ? nullptr : It->getValue().first;
 }
 
 const CachedFileSystemEntry *
@@ -135,11 +135,16 @@ DependencyScanningFilesystemSharedCache::CacheShard::
     getOrEmplaceEntryForFilename(StringRef Filename,
                                  llvm::ErrorOr<llvm::vfs::Status> Stat) {
   std::lock_guard<std::mutex> LockGuard(CacheLock);
-  auto Insertion = EntriesByFilename.insert({Filename, nullptr});
-  if (Insertion.second)
-    Insertion.first->second =
+  auto [It, Inserted] = CacheByFilename.insert({Filename, {nullptr, nullptr}});
+  auto &[CachedEntry, CachedRealPath] = It->getValue();
+  if (!CachedEntry) {
+    // The entry is not present in the shared cache. Either the cache doesn't
+    // know about the file at all, or it only knows about its real path.
+    assert((Inserted || CachedRealPath) && "existing file with empty pair");
+    CachedEntry =
         new (EntryStorage.Allocate()) CachedFileSystemEntry(std::move(Stat));
-  return *Insertion.first->second;
+  }
+  return *CachedEntry;
 }
 
 const CachedFileSystemEntry &
@@ -148,24 +153,58 @@ DependencyScanningFilesystemSharedCache::CacheShard::getOrEmplaceEntryForUID(
     std::unique_ptr<llvm::MemoryBuffer> Contents,
     std::optional<cas::ObjectRef> CASContents) {
   std::lock_guard<std::mutex> LockGuard(CacheLock);
-  auto Insertion = EntriesByUID.insert({UID, nullptr});
-  if (Insertion.second) {
+  auto [It, Inserted] = EntriesByUID.insert({UID, nullptr});
+  auto &CachedEntry = It->getSecond();
+  if (Inserted) {
     CachedFileContents *StoredContents = nullptr;
     if (Contents)
       StoredContents = new (ContentsStorage.Allocate())
           CachedFileContents(std::move(Contents), std::move(CASContents));
-    Insertion.first->second = new (EntryStorage.Allocate())
+    CachedEntry = new (EntryStorage.Allocate())
         CachedFileSystemEntry(std::move(Stat), StoredContents);
   }
-  return *Insertion.first->second;
+  return *CachedEntry;
 }
 
 const CachedFileSystemEntry &
 DependencyScanningFilesystemSharedCache::CacheShard::
     getOrInsertEntryForFilename(StringRef Filename,
                                 const CachedFileSystemEntry &Entry) {
   std::lock_guard<std::mutex> LockGuard(CacheLock);
-  return *EntriesByFilename.insert({Filename, &Entry}).first->getValue();
+  auto [It, Inserted] = CacheByFilename.insert({Filename, {&Entry, nullptr}});
+  auto &[CachedEntry, CachedRealPath] = It->getValue();
+  if (!Inserted || !CachedEntry)
+    CachedEntry = &Entry;
+  return *CachedEntry;
+}
+
+const CachedRealPath *
+DependencyScanningFilesystemSharedCache::CacheShard::findRealPathByFilename(
+    StringRef Filename) const {
+  assert(llvm::sys::path::is_absolute_gnu(Filename));
+  std::lock_guard<std::mutex> LockGuard(CacheLock);
+  auto It = CacheByFilename.find(Filename);
+  return It == CacheByFilename.end() ? nullptr : It->getValue().second;
+}
+
+const CachedRealPath &DependencyScanningFilesystemSharedCache::CacheShard::
+    getOrEmplaceRealPathForFilename(StringRef Filename,
+                                    llvm::ErrorOr<llvm::StringRef> RealPath) {
+  std::lock_guard<std::mutex> LockGuard(CacheLock);
+
+  const CachedRealPath *&StoredRealPath = CacheByFilename[Filename].second;
+  if (!StoredRealPath) {
+    auto OwnedRealPath = [&]() -> CachedRealPath {
+      if (!RealPath)
+        return RealPath.getError();
+      return RealPath->str();
+    }();
+
+    StoredRealPath = new (RealPathStorage.Allocate())
+        CachedRealPath(std::move(OwnedRealPath));
+  }
+
+  return *StoredRealPath;
 }
 
 static bool shouldCacheStatFailures(StringRef Filename) {
@@ -239,24 +278,15 @@ DependencyScanningWorkerFilesystem::computeAndStoreResult(
 llvm::ErrorOr<EntryRef>
 DependencyScanningWorkerFilesystem::getOrCreateFileSystemEntry(
     StringRef OriginalFilename) {
-  StringRef FilenameForLookup;
   SmallString<256> PathBuf;
-  if (llvm::sys::path::is_absolute_gnu(OriginalFilename)) {
-    FilenameForLookup = OriginalFilename;
-  } else if (!WorkingDirForCacheLookup) {
-    return WorkingDirForCacheLookup.getError();
-  } else {
-    StringRef RelFilename = OriginalFilename;
-    RelFilename.consume_front("./");
-    PathBuf = *WorkingDirForCacheLookup;
-    llvm::sys::path::append(PathBuf, RelFilename);
-    FilenameForLookup = PathBuf.str();
-  }
-  assert(llvm::sys::path::is_absolute_gnu(FilenameForLookup));
+  auto FilenameForLookup = tryGetFilenameForLookup(OriginalFilename, PathBuf);
+  if (!FilenameForLookup)
+    return FilenameForLookup.getError();
+
   if (const auto *Entry =
-          findEntryByFilenameWithWriteThrough(FilenameForLookup))
+          findEntryByFilenameWithWriteThrough(*FilenameForLookup))
     return EntryRef(OriginalFilename, *Entry).unwrapError();
-  auto MaybeEntry = computeAndStoreResult(OriginalFilename, FilenameForLookup);
+  auto MaybeEntry = computeAndStoreResult(OriginalFilename, *FilenameForLookup);
   if (!MaybeEntry)
     return MaybeEntry.getError();
   return EntryRef(OriginalFilename, *MaybeEntry).unwrapError();
@@ -276,6 +306,17 @@ DependencyScanningWorkerFilesystem::status(const Twine &Path) {
   return Result->getStatus();
 }
 
+bool DependencyScanningWorkerFilesystem::exists(const Twine &Path) {
+  // While some VFS overlay filesystems may implement more-efficient
+  // mechanisms for `exists` queries, `DependencyScanningWorkerFilesystem`
+  // typically wraps `RealFileSystem` which does not specialize `exists`,
+  // so it is not likely to benefit from such optimizations. Instead,
+  // it is more-valuable to have this query go through the
+  // cached-`status` code-path of the `DependencyScanningWorkerFilesystem`.
+  llvm::ErrorOr<llvm::vfs::Status> Status = status(Path);
+  return Status && Status->exists();
+}
+
 namespace {
 
 /// The VFS that is used by clang consumes the \c CachedFileSystemEntry using
@@ -343,6 +384,54 @@ DependencyScanningWorkerFilesystem::openFileForRead(const Twine &Path) {
   return DepScanFile::create(Result.get());
 }
 
+std::error_code
+DependencyScanningWorkerFilesystem::getRealPath(const Twine &Path,
+                                                SmallVectorImpl<char> &Output) {
+  SmallString<256> OwnedFilename;
+  StringRef OriginalFilename = Path.toStringRef(OwnedFilename);
+
+  SmallString<256> PathBuf;
+  auto FilenameForLookup = tryGetFilenameForLookup(OriginalFilename, PathBuf);
+  if (!FilenameForLookup)
+    return FilenameForLookup.getError();
+
+  auto HandleCachedRealPath =
+      [&Output](const CachedRealPath &RealPath) -> std::error_code {
+    if (!RealPath)
+      return RealPath.getError();
+    Output.assign(RealPath->begin(), RealPath->end());
+    return {};
+  };
+
+  // If we already have the result in local cache, no work required.
+  if (const auto *RealPath =
+          LocalCache.findRealPathByFilename(*FilenameForLookup))
+    return HandleCachedRealPath(*RealPath);
+
+  // If we have the result in the shared cache, cache it locally.
+  auto &Shard = SharedCache.getShardForFilename(*FilenameForLookup);
+  if (const auto *ShardRealPath =
+          Shard.findRealPathByFilename(*FilenameForLookup)) {
+    const auto &RealPath = LocalCache.insertRealPathForFilename(
+        *FilenameForLookup, *ShardRealPath);
+    return HandleCachedRealPath(RealPath);
+  }
+
+  // If we don't know the real path, compute it...
+  std::error_code EC = getUnderlyingFS().getRealPath(OriginalFilename, Output);
+  llvm::ErrorOr<llvm::StringRef> ComputedRealPath = EC;
+  if (!EC)
+    ComputedRealPath = StringRef{Output.data(), Output.size()};
+
+  // ...and try to write it into the shared cache. In case some other thread won
+  // this race and already wrote its own result there, just adopt it. Write
+  // whatever is in the shared cache into the local one.
+  const auto &RealPath = Shard.getOrEmplaceRealPathForFilename(
+      *FilenameForLookup, ComputedRealPath);
+  return HandleCachedRealPath(
+      LocalCache.insertRealPathForFilename(*FilenameForLookup, RealPath));
+}
+
 std::error_code DependencyScanningWorkerFilesystem::setCurrentWorkingDirectory(
     const Twine &Path) {
   std::error_code EC = ProxyFileSystem::setCurrentWorkingDirectory(Path);
@@ -364,4 +453,24 @@ void DependencyScanningWorkerFilesystem::updateWorkingDirForCacheLookup() {
          llvm::sys::path::is_absolute_gnu(*WorkingDirForCacheLookup));
 }
 
+llvm::ErrorOr<StringRef>
+DependencyScanningWorkerFilesystem::tryGetFilenameForLookup(
+    StringRef OriginalFilename, llvm::SmallVectorImpl<char> &PathBuf) const {
+  StringRef FilenameForLookup;
+  if (llvm::sys::path::is_absolute_gnu(OriginalFilename)) {
+    FilenameForLookup = OriginalFilename;
+  } else if (!WorkingDirForCacheLookup) {
+    return WorkingDirForCacheLookup.getError();
+  } else {
+    StringRef RelFilename = OriginalFilename;
+    RelFilename.consume_front("./");
+    PathBuf.assign(WorkingDirForCacheLookup->begin(),
+                   WorkingDirForCacheLookup->end());
+    llvm::sys::path::append(PathBuf, RelFilename);
+    FilenameForLookup = StringRef{PathBuf.begin(), PathBuf.size()};
+  }
+  assert(llvm::sys::path::is_absolute_gnu(FilenameForLookup));
+  return FilenameForLookup;
+}
+
 const char DependencyScanningWorkerFilesystem::ID = 0;