1 //===--- FileManager.cpp - File System Probing and Caching ----------------===//
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
7 //===----------------------------------------------------------------------===//
9 // This file implements the FileManager interface.
11 //===----------------------------------------------------------------------===//
13 // TODO: This should index all interesting directories with dirent calls.
15 // opendir/readdir_r/closedir ?
17 //===----------------------------------------------------------------------===//
19 #include "clang/Basic/FileManager.h"
20 #include "clang/Basic/FileSystemStatCache.h"
21 #include "llvm/ADT/STLExtras.h"
22 #include "llvm/ADT/SmallString.h"
23 #include "llvm/ADT/Statistic.h"
24 #include "llvm/Config/llvm-config.h"
25 #include "llvm/Support/FileSystem.h"
26 #include "llvm/Support/MemoryBuffer.h"
27 #include "llvm/Support/Path.h"
28 #include "llvm/Support/raw_ostream.h"
37 using namespace clang;
39 #define DEBUG_TYPE "file-search"
41 ALWAYS_ENABLED_STATISTIC(NumDirLookups, "Number of directory lookups.");
42 ALWAYS_ENABLED_STATISTIC(NumFileLookups, "Number of file lookups.");
43 ALWAYS_ENABLED_STATISTIC(NumDirCacheMisses,
44 "Number of directory cache misses.");
45 ALWAYS_ENABLED_STATISTIC(NumFileCacheMisses, "Number of file cache misses.");
47 //===----------------------------------------------------------------------===//
49 //===----------------------------------------------------------------------===//
51 FileManager::FileManager(const FileSystemOptions &FSO,
52 IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS)
53 : FS(std::move(FS)), FileSystemOpts(FSO), SeenDirEntries(64),
54 SeenFileEntries(64), NextFileUID(0) {
55 // If the caller doesn't provide a virtual file system, just grab the real
58 this->FS = llvm::vfs::getRealFileSystem();
61 FileManager::~FileManager() = default;
63 void FileManager::setStatCache(std::unique_ptr<FileSystemStatCache> statCache) {
64 assert(statCache && "No stat cache provided?");
65 StatCache = std::move(statCache);
68 void FileManager::clearStatCache() { StatCache.reset(); }
70 /// Retrieve the directory that the given file name resides in.
71 /// Filename can point to either a real file or a virtual file.
72 static llvm::ErrorOr<const DirectoryEntry *>
73 getDirectoryFromFile(FileManager &FileMgr, StringRef Filename,
76 return std::errc::no_such_file_or_directory;
78 if (llvm::sys::path::is_separator(Filename[Filename.size() - 1]))
79 return std::errc::is_a_directory;
81 StringRef DirName = llvm::sys::path::parent_path(Filename);
82 // Use the current directory if file has no path component.
86 return FileMgr.getDirectory(DirName, CacheFailure);
89 /// Add all ancestors of the given path (pointing to either a file or
90 /// a directory) as virtual directories.
91 void FileManager::addAncestorsAsVirtualDirs(StringRef Path) {
92 StringRef DirName = llvm::sys::path::parent_path(Path);
96 auto &NamedDirEnt = *SeenDirEntries.insert(
97 {DirName, std::errc::no_such_file_or_directory}).first;
99 // When caching a virtual directory, we always cache its ancestors
100 // at the same time. Therefore, if DirName is already in the cache,
101 // we don't need to recurse as its ancestors must also already be in
102 // the cache (or it's a known non-virtual directory).
103 if (NamedDirEnt.second)
106 // Add the virtual directory to the cache.
107 auto UDE = std::make_unique<DirectoryEntry>();
108 UDE->Name = NamedDirEnt.first();
109 NamedDirEnt.second = *UDE.get();
110 VirtualDirectoryEntries.push_back(std::move(UDE));
112 // Recursively add the other ancestors.
113 addAncestorsAsVirtualDirs(DirName);
116 llvm::Expected<DirectoryEntryRef>
117 FileManager::getDirectoryRef(StringRef DirName, bool CacheFailure) {
118 // stat doesn't like trailing separators except for root directory.
119 // At least, on Win32 MSVCRT, stat() cannot strip trailing '/'.
120 // (though it can strip '\\')
121 if (DirName.size() > 1 &&
122 DirName != llvm::sys::path::root_path(DirName) &&
123 llvm::sys::path::is_separator(DirName.back()))
124 DirName = DirName.substr(0, DirName.size()-1);
126 // Fixing a problem with "clang C:test.c" on Windows.
127 // Stat("C:") does not recognize "C:" as a valid directory
128 std::string DirNameStr;
129 if (DirName.size() > 1 && DirName.back() == ':' &&
130 DirName.equals_lower(llvm::sys::path::root_name(DirName))) {
131 DirNameStr = DirName.str() + '.';
132 DirName = DirNameStr;
138 // See if there was already an entry in the map. Note that the map
139 // contains both virtual and real directories.
140 auto SeenDirInsertResult =
141 SeenDirEntries.insert({DirName, std::errc::no_such_file_or_directory});
142 if (!SeenDirInsertResult.second) {
143 if (SeenDirInsertResult.first->second)
144 return DirectoryEntryRef(&*SeenDirInsertResult.first);
145 return llvm::errorCodeToError(SeenDirInsertResult.first->second.getError());
148 // We've not seen this before. Fill it in.
150 auto &NamedDirEnt = *SeenDirInsertResult.first;
151 assert(!NamedDirEnt.second && "should be newly-created");
153 // Get the null-terminated directory name as stored as the key of the
154 // SeenDirEntries map.
155 StringRef InterndDirName = NamedDirEnt.first();
157 // Check to see if the directory exists.
158 llvm::vfs::Status Status;
159 auto statError = getStatValue(InterndDirName, Status, false,
160 nullptr /*directory lookup*/);
162 // There's no real directory at the given path.
164 NamedDirEnt.second = statError;
166 SeenDirEntries.erase(DirName);
167 return llvm::errorCodeToError(statError);
170 // It exists. See if we have already opened a directory with the
171 // same inode (this occurs on Unix-like systems when one dir is
172 // symlinked to another, for example) or the same path (on
174 DirectoryEntry &UDE = UniqueRealDirs[Status.getUniqueID()];
176 NamedDirEnt.second = UDE;
177 if (UDE.getName().empty()) {
178 // We don't have this directory yet, add it. We use the string
179 // key from the SeenDirEntries map as the string.
180 UDE.Name = InterndDirName;
183 return DirectoryEntryRef(&NamedDirEnt);
186 llvm::ErrorOr<const DirectoryEntry *>
187 FileManager::getDirectory(StringRef DirName, bool CacheFailure) {
188 auto Result = getDirectoryRef(DirName, CacheFailure);
190 return &Result->getDirEntry();
191 return llvm::errorToErrorCode(Result.takeError());
194 llvm::ErrorOr<const FileEntry *>
195 FileManager::getFile(StringRef Filename, bool openFile, bool CacheFailure) {
196 auto Result = getFileRef(Filename, openFile, CacheFailure);
198 return &Result->getFileEntry();
199 return llvm::errorToErrorCode(Result.takeError());
202 llvm::Expected<FileEntryRef>
203 FileManager::getFileRef(StringRef Filename, bool openFile, bool CacheFailure) {
206 // See if there is already an entry in the map.
207 auto SeenFileInsertResult =
208 SeenFileEntries.insert({Filename, std::errc::no_such_file_or_directory});
209 if (!SeenFileInsertResult.second) {
210 if (!SeenFileInsertResult.first->second)
211 return llvm::errorCodeToError(
212 SeenFileInsertResult.first->second.getError());
213 // Construct and return and FileEntryRef, unless it's a redirect to another
215 SeenFileEntryOrRedirect Value = *SeenFileInsertResult.first->second;
217 if (LLVM_LIKELY(FE = Value.dyn_cast<FileEntry *>()))
218 return FileEntryRef(SeenFileInsertResult.first->first(), *FE);
219 return getFileRef(*Value.get<const StringRef *>(), openFile, CacheFailure);
222 // We've not seen this before. Fill it in.
223 ++NumFileCacheMisses;
224 auto *NamedFileEnt = &*SeenFileInsertResult.first;
225 assert(!NamedFileEnt->second && "should be newly-created");
227 // Get the null-terminated file name as stored as the key of the
228 // SeenFileEntries map.
229 StringRef InterndFileName = NamedFileEnt->first();
231 // Look up the directory for the file. When looking up something like
232 // sys/foo.h we'll discover all of the search directories that have a 'sys'
233 // subdirectory. This will let us avoid having to waste time on known-to-fail
234 // searches when we go to find sys/bar.h, because all the search directories
235 // without a 'sys' subdir will get a cached failure result.
236 auto DirInfoOrErr = getDirectoryFromFile(*this, Filename, CacheFailure);
237 if (!DirInfoOrErr) { // Directory doesn't exist, file can't exist.
239 NamedFileEnt->second = DirInfoOrErr.getError();
241 SeenFileEntries.erase(Filename);
243 return llvm::errorCodeToError(DirInfoOrErr.getError());
245 const DirectoryEntry *DirInfo = *DirInfoOrErr;
247 // FIXME: Use the directory info to prune this, before doing the stat syscall.
248 // FIXME: This will reduce the # syscalls.
250 // Check to see if the file exists.
251 std::unique_ptr<llvm::vfs::File> F;
252 llvm::vfs::Status Status;
253 auto statError = getStatValue(InterndFileName, Status, true,
254 openFile ? &F : nullptr);
256 // There's no real file at the given path.
258 NamedFileEnt->second = statError;
260 SeenFileEntries.erase(Filename);
262 return llvm::errorCodeToError(statError);
265 assert((openFile || !F) && "undesired open file");
267 // It exists. See if we have already opened a file with the same inode.
268 // This occurs when one dir is symlinked to another, for example.
269 FileEntry &UFE = UniqueRealFiles[Status.getUniqueID()];
271 NamedFileEnt->second = &UFE;
273 // If the name returned by getStatValue is different than Filename, re-intern
275 if (Status.getName() != Filename) {
276 auto &NewNamedFileEnt =
277 *SeenFileEntries.insert({Status.getName(), &UFE}).first;
278 assert((*NewNamedFileEnt.second).get<FileEntry *>() == &UFE &&
279 "filename from getStatValue() refers to wrong file");
280 InterndFileName = NewNamedFileEnt.first().data();
281 // In addition to re-interning the name, construct a redirecting seen file
282 // entry, that will point to the name the filesystem actually wants to use.
283 StringRef *Redirect = new (CanonicalNameStorage) StringRef(InterndFileName);
284 auto SeenFileInsertResultIt = SeenFileEntries.find(Filename);
285 assert(SeenFileInsertResultIt != SeenFileEntries.end() &&
286 "unexpected SeenFileEntries cache miss");
287 SeenFileInsertResultIt->second = Redirect;
288 NamedFileEnt = &*SeenFileInsertResultIt;
291 if (UFE.isValid()) { // Already have an entry with this inode, return it.
293 // FIXME: this hack ensures that if we look up a file by a virtual path in
294 // the VFS that the getDir() will have the virtual path, even if we found
295 // the file by a 'real' path first. This is required in order to find a
296 // module's structure when its headers/module map are mapped in the VFS.
297 // We should remove this as soon as we can properly support a file having
299 if (DirInfo != UFE.Dir && Status.IsVFSMapped)
302 // Always update the name to use the last name by which a file was accessed.
303 // FIXME: Neither this nor always using the first name is correct; we want
304 // to switch towards a design where we return a FileName object that
305 // encapsulates both the name by which the file was accessed and the
306 // corresponding FileEntry.
307 // FIXME: The Name should be removed from FileEntry once all clients
308 // adopt FileEntryRef.
309 UFE.Name = InterndFileName;
311 return FileEntryRef(InterndFileName, UFE);
314 // Otherwise, we don't have this file yet, add it.
315 UFE.Name = InterndFileName;
316 UFE.Size = Status.getSize();
317 UFE.ModTime = llvm::sys::toTimeT(Status.getLastModificationTime());
319 UFE.UID = NextFileUID++;
320 UFE.UniqueID = Status.getUniqueID();
321 UFE.IsNamedPipe = Status.getType() == llvm::sys::fs::file_type::fifo_file;
322 UFE.File = std::move(F);
326 if (auto PathName = UFE.File->getName())
327 fillRealPathName(&UFE, *PathName);
328 } else if (!openFile) {
329 // We should still fill the path even if we aren't opening the file.
330 fillRealPathName(&UFE, InterndFileName);
332 return FileEntryRef(InterndFileName, UFE);
336 FileManager::getVirtualFile(StringRef Filename, off_t Size,
337 time_t ModificationTime) {
340 // See if there is already an entry in the map for an existing file.
341 auto &NamedFileEnt = *SeenFileEntries.insert(
342 {Filename, std::errc::no_such_file_or_directory}).first;
343 if (NamedFileEnt.second) {
344 SeenFileEntryOrRedirect Value = *NamedFileEnt.second;
346 if (LLVM_LIKELY(FE = Value.dyn_cast<FileEntry *>()))
348 return getVirtualFile(*Value.get<const StringRef *>(), Size,
352 // We've not seen this before, or the file is cached as non-existent.
353 ++NumFileCacheMisses;
354 addAncestorsAsVirtualDirs(Filename);
355 FileEntry *UFE = nullptr;
357 // Now that all ancestors of Filename are in the cache, the
358 // following call is guaranteed to find the DirectoryEntry from the
360 auto DirInfo = getDirectoryFromFile(*this, Filename, /*CacheFailure=*/true);
362 "The directory of a virtual file should already be in the cache.");
364 // Check to see if the file exists. If so, drop the virtual file
365 llvm::vfs::Status Status;
366 const char *InterndFileName = NamedFileEnt.first().data();
367 if (!getStatValue(InterndFileName, Status, true, nullptr)) {
368 UFE = &UniqueRealFiles[Status.getUniqueID()];
369 Status = llvm::vfs::Status(
370 Status.getName(), Status.getUniqueID(),
371 llvm::sys::toTimePoint(ModificationTime),
372 Status.getUser(), Status.getGroup(), Size,
373 Status.getType(), Status.getPermissions());
375 NamedFileEnt.second = UFE;
377 // If we had already opened this file, close it now so we don't
378 // leak the descriptor. We're not going to use the file
379 // descriptor anyway, since this is a virtual file.
383 // If we already have an entry with this inode, return it.
387 UFE->UniqueID = Status.getUniqueID();
388 UFE->IsNamedPipe = Status.getType() == llvm::sys::fs::file_type::fifo_file;
389 fillRealPathName(UFE, Status.getName());
391 VirtualFileEntries.push_back(std::make_unique<FileEntry>());
392 UFE = VirtualFileEntries.back().get();
393 NamedFileEnt.second = UFE;
396 UFE->Name = InterndFileName;
398 UFE->ModTime = ModificationTime;
400 UFE->UID = NextFileUID++;
406 llvm::Optional<FileEntryRef> FileManager::getBypassFile(FileEntryRef VF) {
407 // Stat of the file and return nullptr if it doesn't exist.
408 llvm::vfs::Status Status;
409 if (getStatValue(VF.getName(), Status, /*isFile=*/true, /*F=*/nullptr))
412 // Fill it in from the stat.
413 BypassFileEntries.push_back(std::make_unique<FileEntry>());
414 const FileEntry &VFE = VF.getFileEntry();
415 FileEntry &BFE = *BypassFileEntries.back();
416 BFE.Name = VFE.getName();
417 BFE.Size = Status.getSize();
419 BFE.ModTime = llvm::sys::toTimeT(Status.getLastModificationTime());
420 BFE.UID = NextFileUID++;
422 return FileEntryRef(VF.getName(), BFE);
425 bool FileManager::FixupRelativePath(SmallVectorImpl<char> &path) const {
426 StringRef pathRef(path.data(), path.size());
428 if (FileSystemOpts.WorkingDir.empty()
429 || llvm::sys::path::is_absolute(pathRef))
432 SmallString<128> NewPath(FileSystemOpts.WorkingDir);
433 llvm::sys::path::append(NewPath, pathRef);
438 bool FileManager::makeAbsolutePath(SmallVectorImpl<char> &Path) const {
439 bool Changed = FixupRelativePath(Path);
441 if (!llvm::sys::path::is_absolute(StringRef(Path.data(), Path.size()))) {
442 FS->makeAbsolute(Path);
449 void FileManager::fillRealPathName(FileEntry *UFE, llvm::StringRef FileName) {
450 llvm::SmallString<128> AbsPath(FileName);
451 // This is not the same as `VFS::getRealPath()`, which resolves symlinks
452 // but can be very expensive on real file systems.
453 // FIXME: the semantic of RealPathName is unclear, and the name might be
454 // misleading. We need to clean up the interface here.
455 makeAbsolutePath(AbsPath);
456 llvm::sys::path::remove_dots(AbsPath, /*remove_dot_dot=*/true);
457 UFE->RealPathName = AbsPath.str();
460 llvm::ErrorOr<std::unique_ptr<llvm::MemoryBuffer>>
461 FileManager::getBufferForFile(const FileEntry *Entry, bool isVolatile) {
462 uint64_t FileSize = Entry->getSize();
463 // If there's a high enough chance that the file have changed since we
464 // got its size, force a stat before opening it.
468 StringRef Filename = Entry->getName();
469 // If the file is already open, use the open file descriptor.
472 Entry->File->getBuffer(Filename, FileSize,
473 /*RequiresNullTerminator=*/true, isVolatile);
478 // Otherwise, open the file.
479 return getBufferForFileImpl(Filename, FileSize, isVolatile);
482 llvm::ErrorOr<std::unique_ptr<llvm::MemoryBuffer>>
483 FileManager::getBufferForFileImpl(StringRef Filename, int64_t FileSize,
485 if (FileSystemOpts.WorkingDir.empty())
486 return FS->getBufferForFile(Filename, FileSize,
487 /*RequiresNullTerminator=*/true, isVolatile);
489 SmallString<128> FilePath(Filename);
490 FixupRelativePath(FilePath);
491 return FS->getBufferForFile(FilePath, FileSize,
492 /*RequiresNullTerminator=*/true, isVolatile);
495 /// getStatValue - Get the 'stat' information for the specified path,
496 /// using the cache to accelerate it if possible. This returns true
497 /// if the path points to a virtual file or does not exist, or returns
498 /// false if it's an existent real file. If FileDescriptor is NULL,
499 /// do directory look-up instead of file look-up.
501 FileManager::getStatValue(StringRef Path, llvm::vfs::Status &Status,
502 bool isFile, std::unique_ptr<llvm::vfs::File> *F) {
503 // FIXME: FileSystemOpts shouldn't be passed in here, all paths should be
505 if (FileSystemOpts.WorkingDir.empty())
506 return FileSystemStatCache::get(Path, Status, isFile, F,
507 StatCache.get(), *FS);
509 SmallString<128> FilePath(Path);
510 FixupRelativePath(FilePath);
512 return FileSystemStatCache::get(FilePath.c_str(), Status, isFile, F,
513 StatCache.get(), *FS);
517 FileManager::getNoncachedStatValue(StringRef Path,
518 llvm::vfs::Status &Result) {
519 SmallString<128> FilePath(Path);
520 FixupRelativePath(FilePath);
522 llvm::ErrorOr<llvm::vfs::Status> S = FS->status(FilePath.c_str());
526 return std::error_code();
529 void FileManager::GetUniqueIDMapping(
530 SmallVectorImpl<const FileEntry *> &UIDToFiles) const {
532 UIDToFiles.resize(NextFileUID);
535 for (llvm::StringMap<llvm::ErrorOr<SeenFileEntryOrRedirect>,
536 llvm::BumpPtrAllocator>::const_iterator
537 FE = SeenFileEntries.begin(),
538 FEEnd = SeenFileEntries.end();
540 if (llvm::ErrorOr<SeenFileEntryOrRedirect> Entry = FE->getValue()) {
541 if (const auto *FE = (*Entry).dyn_cast<FileEntry *>())
542 UIDToFiles[FE->getUID()] = FE;
545 // Map virtual file entries
546 for (const auto &VFE : VirtualFileEntries)
547 UIDToFiles[VFE->getUID()] = VFE.get();
550 StringRef FileManager::getCanonicalName(const DirectoryEntry *Dir) {
551 llvm::DenseMap<const void *, llvm::StringRef>::iterator Known
552 = CanonicalNames.find(Dir);
553 if (Known != CanonicalNames.end())
554 return Known->second;
556 StringRef CanonicalName(Dir->getName());
558 SmallString<4096> CanonicalNameBuf;
559 if (!FS->getRealPath(Dir->getName(), CanonicalNameBuf))
560 CanonicalName = StringRef(CanonicalNameBuf).copy(CanonicalNameStorage);
562 CanonicalNames.insert({Dir, CanonicalName});
563 return CanonicalName;
566 StringRef FileManager::getCanonicalName(const FileEntry *File) {
567 llvm::DenseMap<const void *, llvm::StringRef>::iterator Known
568 = CanonicalNames.find(File);
569 if (Known != CanonicalNames.end())
570 return Known->second;
572 StringRef CanonicalName(File->getName());
574 SmallString<4096> CanonicalNameBuf;
575 if (!FS->getRealPath(File->getName(), CanonicalNameBuf))
576 CanonicalName = StringRef(CanonicalNameBuf).copy(CanonicalNameStorage);
578 CanonicalNames.insert({File, CanonicalName});
579 return CanonicalName;
582 void FileManager::PrintStats() const {
583 llvm::errs() << "\n*** File Manager Stats:\n";
584 llvm::errs() << UniqueRealFiles.size() << " real files found, "
585 << UniqueRealDirs.size() << " real dirs found.\n";
586 llvm::errs() << VirtualFileEntries.size() << " virtual files found, "
587 << VirtualDirectoryEntries.size() << " virtual dirs found.\n";
588 llvm::errs() << NumDirLookups << " dir lookups, "
589 << NumDirCacheMisses << " dir cache misses.\n";
590 llvm::errs() << NumFileLookups << " file lookups, "
591 << NumFileCacheMisses << " file cache misses.\n";
593 //llvm::errs() << PagesMapped << BytesOfPagesMapped << FSLookups;