LLVM  16.0.0git
FileUtilities.cpp
Go to the documentation of this file.
1 //===- Support/FileUtilities.cpp - File System Utilities ------------------===//
2 //
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
6 //
7 //===----------------------------------------------------------------------===//
8 //
9 // This file implements a family of utility functions which are useful for doing
10 // various things with files.
11 //
12 //===----------------------------------------------------------------------===//
13 
15 #include "llvm/ADT/SmallString.h"
16 #include "llvm/ADT/StringExtras.h"
17 #include "llvm/Support/Error.h"
18 #include "llvm/Support/ErrorOr.h"
20 #include "llvm/Support/Process.h"
22 #include <cmath>
23 #include <cstdint>
24 #include <cstdlib>
25 #include <cstring>
26 #include <memory>
27 #include <system_error>
28 
29 using namespace llvm;
30 
31 static bool isSignedChar(char C) {
32  return (C == '+' || C == '-');
33 }
34 
35 static bool isExponentChar(char C) {
36  switch (C) {
37  case 'D': // Strange exponential notation.
38  case 'd': // Strange exponential notation.
39  case 'e':
40  case 'E': return true;
41  default: return false;
42  }
43 }
44 
45 static bool isNumberChar(char C) {
46  switch (C) {
47  case '0': case '1': case '2': case '3': case '4':
48  case '5': case '6': case '7': case '8': case '9':
49  case '.': return true;
50  default: return isSignedChar(C) || isExponentChar(C);
51  }
52 }
53 
54 static const char *BackupNumber(const char *Pos, const char *FirstChar) {
55  // If we didn't stop in the middle of a number, don't backup.
56  if (!isNumberChar(*Pos)) return Pos;
57 
58  // Otherwise, return to the start of the number.
59  bool HasPeriod = false;
60  while (Pos > FirstChar && isNumberChar(Pos[-1])) {
61  // Backup over at most one period.
62  if (Pos[-1] == '.') {
63  if (HasPeriod)
64  break;
65  HasPeriod = true;
66  }
67 
68  --Pos;
69  if (Pos > FirstChar && isSignedChar(Pos[0]) && !isExponentChar(Pos[-1]))
70  break;
71  }
72  return Pos;
73 }
74 
75 /// EndOfNumber - Return the first character that is not part of the specified
76 /// number. This assumes that the buffer is null terminated, so it won't fall
77 /// off the end.
78 static const char *EndOfNumber(const char *Pos) {
79  while (isNumberChar(*Pos))
80  ++Pos;
81  return Pos;
82 }
83 
84 /// CompareNumbers - compare two numbers, returning true if they are different.
85 static bool CompareNumbers(const char *&F1P, const char *&F2P,
86  const char *F1End, const char *F2End,
87  double AbsTolerance, double RelTolerance,
88  std::string *ErrorMsg) {
89  const char *F1NumEnd, *F2NumEnd;
90  double V1 = 0.0, V2 = 0.0;
91 
92  // If one of the positions is at a space and the other isn't, chomp up 'til
93  // the end of the space.
94  while (isSpace(static_cast<unsigned char>(*F1P)) && F1P != F1End)
95  ++F1P;
96  while (isSpace(static_cast<unsigned char>(*F2P)) && F2P != F2End)
97  ++F2P;
98 
99  // If we stop on numbers, compare their difference.
100  if (!isNumberChar(*F1P) || !isNumberChar(*F2P)) {
101  // The diff failed.
102  F1NumEnd = F1P;
103  F2NumEnd = F2P;
104  } else {
105  // Note that some ugliness is built into this to permit support for numbers
106  // that use "D" or "d" as their exponential marker, e.g. "1.234D45". This
107  // occurs in 200.sixtrack in spec2k.
108  V1 = strtod(F1P, const_cast<char**>(&F1NumEnd));
109  V2 = strtod(F2P, const_cast<char**>(&F2NumEnd));
110 
111  if (*F1NumEnd == 'D' || *F1NumEnd == 'd') {
112  // Copy string into tmp buffer to replace the 'D' with an 'e'.
113  SmallString<200> StrTmp(F1P, EndOfNumber(F1NumEnd)+1);
114  // Strange exponential notation!
115  StrTmp[static_cast<unsigned>(F1NumEnd-F1P)] = 'e';
116 
117  V1 = strtod(&StrTmp[0], const_cast<char**>(&F1NumEnd));
118  F1NumEnd = F1P + (F1NumEnd-&StrTmp[0]);
119  }
120 
121  if (*F2NumEnd == 'D' || *F2NumEnd == 'd') {
122  // Copy string into tmp buffer to replace the 'D' with an 'e'.
123  SmallString<200> StrTmp(F2P, EndOfNumber(F2NumEnd)+1);
124  // Strange exponential notation!
125  StrTmp[static_cast<unsigned>(F2NumEnd-F2P)] = 'e';
126 
127  V2 = strtod(&StrTmp[0], const_cast<char**>(&F2NumEnd));
128  F2NumEnd = F2P + (F2NumEnd-&StrTmp[0]);
129  }
130  }
131 
132  if (F1NumEnd == F1P || F2NumEnd == F2P) {
133  if (ErrorMsg) {
134  *ErrorMsg = "FP Comparison failed, not a numeric difference between '";
135  *ErrorMsg += F1P[0];
136  *ErrorMsg += "' and '";
137  *ErrorMsg += F2P[0];
138  *ErrorMsg += "'";
139  }
140  return true;
141  }
142 
143  // Check to see if these are inside the absolute tolerance
144  if (AbsTolerance < std::abs(V1-V2)) {
145  // Nope, check the relative tolerance...
146  double Diff;
147  if (V2)
148  Diff = std::abs(V1/V2 - 1.0);
149  else if (V1)
150  Diff = std::abs(V2/V1 - 1.0);
151  else
152  Diff = 0; // Both zero.
153  if (Diff > RelTolerance) {
154  if (ErrorMsg) {
155  raw_string_ostream(*ErrorMsg)
156  << "Compared: " << V1 << " and " << V2 << '\n'
157  << "abs. diff = " << std::abs(V1-V2) << " rel.diff = " << Diff << '\n'
158  << "Out of tolerance: rel/abs: " << RelTolerance << '/'
159  << AbsTolerance;
160  }
161  return true;
162  }
163  }
164 
165  // Otherwise, advance our read pointers to the end of the numbers.
166  F1P = F1NumEnd; F2P = F2NumEnd;
167  return false;
168 }
169 
170 /// DiffFilesWithTolerance - Compare the two files specified, returning 0 if the
171 /// files match, 1 if they are different, and 2 if there is a file error. This
172 /// function differs from DiffFiles in that you can specify an absolete and
173 /// relative FP error that is allowed to exist. If you specify a string to fill
174 /// in for the error option, it will set the string to an error message if an
175 /// error occurs, allowing the caller to distinguish between a failed diff and a
176 /// file system error.
177 ///
179  StringRef NameB,
180  double AbsTol, double RelTol,
181  std::string *Error) {
182  // Now its safe to mmap the files into memory because both files
183  // have a non-zero size.
185  if (std::error_code EC = F1OrErr.getError()) {
186  if (Error)
187  *Error = EC.message();
188  return 2;
189  }
190  MemoryBuffer &F1 = *F1OrErr.get();
191 
193  if (std::error_code EC = F2OrErr.getError()) {
194  if (Error)
195  *Error = EC.message();
196  return 2;
197  }
198  MemoryBuffer &F2 = *F2OrErr.get();
199 
200  // Okay, now that we opened the files, scan them for the first difference.
201  const char *File1Start = F1.getBufferStart();
202  const char *File2Start = F2.getBufferStart();
203  const char *File1End = F1.getBufferEnd();
204  const char *File2End = F2.getBufferEnd();
205  const char *F1P = File1Start;
206  const char *F2P = File2Start;
207  uint64_t A_size = F1.getBufferSize();
208  uint64_t B_size = F2.getBufferSize();
209 
210  // Are the buffers identical? Common case: Handle this efficiently.
211  if (A_size == B_size &&
212  std::memcmp(File1Start, File2Start, A_size) == 0)
213  return 0;
214 
215  // Otherwise, we are done a tolerances are set.
216  if (AbsTol == 0 && RelTol == 0) {
217  if (Error)
218  *Error = "Files differ without tolerance allowance";
219  return 1; // Files different!
220  }
221 
222  bool CompareFailed = false;
223  while (true) {
224  // Scan for the end of file or next difference.
225  while (F1P < File1End && F2P < File2End && *F1P == *F2P) {
226  ++F1P;
227  ++F2P;
228  }
229 
230  if (F1P >= File1End || F2P >= File2End) break;
231 
232  // Okay, we must have found a difference. Backup to the start of the
233  // current number each stream is at so that we can compare from the
234  // beginning.
235  F1P = BackupNumber(F1P, File1Start);
236  F2P = BackupNumber(F2P, File2Start);
237 
238  // Now that we are at the start of the numbers, compare them, exiting if
239  // they don't match.
240  if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error)) {
241  CompareFailed = true;
242  break;
243  }
244  }
245 
246  // Okay, we reached the end of file. If both files are at the end, we
247  // succeeded.
248  bool F1AtEnd = F1P >= File1End;
249  bool F2AtEnd = F2P >= File2End;
250  if (!CompareFailed && (!F1AtEnd || !F2AtEnd)) {
251  // Else, we might have run off the end due to a number: backup and retry.
252  if (F1AtEnd && isNumberChar(F1P[-1])) --F1P;
253  if (F2AtEnd && isNumberChar(F2P[-1])) --F2P;
254  F1P = BackupNumber(F1P, File1Start);
255  F2P = BackupNumber(F2P, File2Start);
256 
257  // Now that we are at the start of the numbers, compare them, exiting if
258  // they don't match.
259  if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error))
260  CompareFailed = true;
261 
262  // If we found the end, we succeeded.
263  if (F1P < File1End || F2P < File2End)
264  CompareFailed = true;
265  }
266 
267  return CompareFailed;
268 }
269 
271  OS << "atomic_write_error: ";
272  switch (Error) {
274  OS << "failed_to_create_uniq_file";
275  return;
277  OS << "output_stream_error";
278  return;
280  OS << "failed_to_rename_temp_file";
281  return;
282  }
283  llvm_unreachable("unknown atomic_write_error value in "
284  "failed_to_rename_temp_file::log()");
285 }
286 
288  StringRef FinalPath, StringRef Buffer) {
289  return writeFileAtomically(TempPathModel, FinalPath,
290  [&Buffer](llvm::raw_ostream &OS) {
291  OS.write(Buffer.data(), Buffer.size());
292  return llvm::Error::success();
293  });
294 }
295 
297  StringRef TempPathModel, StringRef FinalPath,
299  SmallString<128> GeneratedUniqPath;
300  int TempFD;
301  if (sys::fs::createUniqueFile(TempPathModel, TempFD, GeneratedUniqPath)) {
302  return llvm::make_error<AtomicFileWriteError>(
304  }
305  llvm::FileRemover RemoveTmpFileOnFail(GeneratedUniqPath);
306 
307  raw_fd_ostream OS(TempFD, /*shouldClose=*/true);
308  if (llvm::Error Err = Writer(OS)) {
309  return Err;
310  }
311 
312  OS.close();
313  if (OS.has_error()) {
314  OS.clear_error();
315  return llvm::make_error<AtomicFileWriteError>(
317  }
318 
319  if (sys::fs::rename(/*from=*/GeneratedUniqPath, /*to=*/FinalPath)) {
320  return llvm::make_error<AtomicFileWriteError>(
322  }
323 
324  RemoveTmpFileOnFail.releaseFile();
325  return Error::success();
326 }
327 
331 
332  if (InputFilename != "-") {
333  if (auto EC = sys::fs::status(InputFilename, Status))
334  return createFileError(InputFilename, EC);
335  } else {
336  Status.permissions(static_cast<sys::fs::perms>(0777));
337  }
338 
339  return FilePermissionsApplier(InputFilename, Status);
340 }
341 
343  StringRef OutputFilename, bool CopyDates,
344  Optional<sys::fs::perms> OverwritePermissions) {
345  sys::fs::file_status Status = InputStatus;
346 
347  if (OverwritePermissions)
348  Status.permissions(*OverwritePermissions);
349 
350  int FD = 0;
351 
352  // Writing to stdout should not be treated as an error here, just
353  // do not set access/modification times or permissions.
354  if (OutputFilename == "-")
355  return Error::success();
356 
357  if (std::error_code EC = sys::fs::openFileForWrite(OutputFilename, FD,
359  return createFileError(OutputFilename, EC);
360 
361  if (CopyDates)
362  if (std::error_code EC = sys::fs::setLastAccessAndModificationTime(
363  FD, Status.getLastAccessedTime(), Status.getLastModificationTime()))
364  return createFileError(OutputFilename, EC);
365 
366  sys::fs::file_status OStat;
367  if (std::error_code EC = sys::fs::status(FD, OStat))
368  return createFileError(OutputFilename, EC);
369  if (OStat.type() == sys::fs::file_type::regular_file) {
370 #ifndef _WIN32
371  // Keep ownership if llvm-objcopy is called under root.
372  if (OutputFilename == InputFilename && OStat.getUser() == 0)
373  sys::fs::changeFileOwnership(FD, Status.getUser(), Status.getGroup());
374 #endif
375 
376  sys::fs::perms Perm = Status.permissions();
377  if (OutputFilename != InputFilename)
378  Perm = static_cast<sys::fs::perms>(Perm & ~sys::fs::getUmask() & ~06000);
379 #ifdef _WIN32
380  if (std::error_code EC = sys::fs::setPermissions(OutputFilename, Perm))
381 #else
382  if (std::error_code EC = sys::fs::setPermissions(FD, Perm))
383 #endif
384  return createFileError(OutputFilename, EC);
385  }
386 
387  if (std::error_code EC = sys::Process::SafelyCloseFileDescriptor(FD))
388  return createFileError(OutputFilename, EC);
389 
390  return Error::success();
391 }
392 
MemoryBuffer.h
BackupNumber
static const char * BackupNumber(const char *Pos, const char *FirstChar)
Definition: FileUtilities.cpp:54
llvm::sys::fs::openFileForWrite
std::error_code openFileForWrite(const Twine &Name, int &ResultFD, CreationDisposition Disp=CD_CreateAlways, OpenFlags Flags=OF_None, unsigned Mode=0666)
Opens the file with the given name in a write-only or read-write mode, returning its open file descri...
Definition: FileSystem.h:1065
llvm
This is an optimization pass for GlobalISel generic memory operations.
Definition: AddressRanges.h:18
llvm::raw_fd_ostream::close
void close()
Manually flush the stream and close the file.
Definition: raw_ostream.cpp:791
llvm::atomic_write_error::output_stream_error
@ output_stream_error
llvm::FileRemover
FileRemover - This class is a simple object meant to be stack allocated.
Definition: FileUtilities.h:42
llvm::raw_string_ostream
A raw_ostream that writes to an std::string.
Definition: raw_ostream.h:628
llvm::FileRemover::releaseFile
void releaseFile()
releaseFile - Take ownership of the file away from the FileRemover so it will not be removed when the...
Definition: FileUtilities.h:76
llvm::FilePermissionsApplier
FilePermssionsApplier helps to copy permissions from an input file to an output one.
Definition: FileUtilities.h:117
llvm::Error::success
static ErrorSuccess success()
Create a success value.
Definition: Error.h:329
memcmp
Merge contiguous icmps into a memcmp
Definition: MergeICmps.cpp:903
Error.h
llvm::AtomicFileWriteError::ID
static char ID
Definition: FileUtilities.h:92
llvm::sys::fs::createUniqueFile
std::error_code createUniqueFile(const Twine &Model, int &ResultFD, SmallVectorImpl< char > &ResultPath, OpenFlags Flags=OF_None, unsigned Mode=all_read|all_write)
Create a uniquely named file.
Definition: Path.cpp:822
llvm::FilePermissionsApplier::create
static Expected< FilePermissionsApplier > create(StringRef InputFilename)
Definition: FileUtilities.cpp:329
llvm::sys::fs::file_type::regular_file
@ regular_file
llvm::sys::fs::CD_OpenExisting
@ CD_OpenExisting
CD_OpenExisting - When opening a file:
Definition: FileSystem.h:743
llvm::Optional
Definition: APInt.h:33
llvm::Expected
Tagged union holding either a T or a Error.
Definition: APFloat.h:41
llvm::sys::fs::setLastAccessAndModificationTime
std::error_code setLastAccessAndModificationTime(int FD, TimePoint<> AccessTime, TimePoint<> ModificationTime)
Set the file modification and access time.
endif
__FakeVCSRevision h endif() endif() set(generated_files "$
Definition: CMakeLists.txt:16
llvm::MemoryBuffer
This interface provides simple read-only access to a block of memory, and provides simple methods for...
Definition: MemoryBuffer.h:50
isExponentChar
static bool isExponentChar(char C)
Definition: FileUtilities.cpp:35
InputFilename
static cl::opt< std::string > InputFilename(cl::Positional, cl::desc("<input file>"), cl::init("-"))
llvm::MemoryBuffer::getFile
static ErrorOr< std::unique_ptr< MemoryBuffer > > getFile(const Twine &Filename, bool IsText=false, bool RequiresNullTerminator=true, bool IsVolatile=false)
Open the specified file as a MemoryBuffer, returning a new MemoryBuffer if successful,...
Definition: MemoryBuffer.cpp:239
Process.h
llvm::writeFileAtomically
llvm::Error writeFileAtomically(StringRef TempPathModel, StringRef FinalPath, StringRef Buffer)
Creates a unique file with name according to the given TempPathModel, writes content of Buffer to the...
Definition: FileUtilities.cpp:287
llvm::MemoryBuffer::getBufferEnd
const char * getBufferEnd() const
Definition: MemoryBuffer.h:66
llvm::raw_fd_ostream::has_error
bool has_error() const
Return the value of the flag in this raw_fd_ostream indicating whether an output error has been encou...
Definition: raw_ostream.h:531
SmallString.h
llvm::createFileError
Error createFileError(const Twine &F, Error E)
Concatenate a source file path and/or name with an Error.
Definition: Error.h:1320
llvm::MemoryBuffer::getBufferSize
size_t getBufferSize() const
Definition: MemoryBuffer.h:67
isSignedChar
static bool isSignedChar(char C)
Definition: FileUtilities.cpp:31
C
(vector float) vec_cmpeq(*A, *B) C
Definition: README_ALTIVEC.txt:86
llvm::raw_fd_ostream::clear_error
void clear_error()
Set the flag read by has_error() to false.
Definition: raw_ostream.h:542
llvm::ErrorOr::getError
std::error_code getError() const
Definition: ErrorOr.h:153
llvm::atomic_write_error::failed_to_rename_temp_file
@ failed_to_rename_temp_file
llvm::raw_ostream::write
raw_ostream & write(unsigned char C)
Definition: raw_ostream.cpp:218
llvm::raw_ostream
This class implements an extremely fast bulk output stream that can only output to a stream.
Definition: raw_ostream.h:52
llvm::StringRef::data
const char * data() const
data - Get a pointer to the start of the string (which may not be null terminated).
Definition: StringRef.h:131
llvm::sys::fs::changeFileOwnership
std::error_code changeFileOwnership(int FD, uint32_t Owner, uint32_t Group)
Change ownership of a file.
llvm::SmallString
SmallString - A SmallString is just a SmallVector with methods and accessors that make it work better...
Definition: SmallString.h:26
llvm::sys::Process::SafelyCloseFileDescriptor
static std::error_code SafelyCloseFileDescriptor(int FD)
llvm::sys::fs::file_status
Represents the result of a call to sys::fs::status().
Definition: FileSystem.h:226
uint64_t
ErrorOr.h
StringExtras.h
llvm::FilePermissionsApplier::apply
Error apply(StringRef OutputFilename, bool CopyDates=false, Optional< sys::fs::perms > OverwritePermissions=None)
Apply stored permissions to the OutputFilename.
Definition: FileUtilities.cpp:342
function
print Print MemDeps of function
Definition: MemDepPrinter.cpp:82
llvm::sys::fs::setPermissions
std::error_code setPermissions(const Twine &Path, perms Permissions)
Set file permissions.
llvm::sys::fs::basic_file_status::type
file_type type() const
Definition: FileSystem.h:184
Status
Definition: SIModeRegister.cpp:29
llvm::sys::fs::basic_file_status::getUser
uint32_t getUser() const
Definition: FileSystem.h:203
FileUtilities.h
llvm::StringRef
StringRef - Represent a constant reference to a string, i.e.
Definition: StringRef.h:50
llvm_unreachable
#define llvm_unreachable(msg)
Marks that the current location is not supposed to be reachable.
Definition: ErrorHandling.h:143
llvm::sys::fs::rename
std::error_code rename(const Twine &from, const Twine &to)
Rename from to to.
llvm::AtomicFileWriteError::log
void log(raw_ostream &OS) const override
Print an error message to an output stream.
Definition: FileUtilities.cpp:270
llvm::atomic_write_error::failed_to_create_uniq_file
@ failed_to_create_uniq_file
llvm::raw_fd_ostream
A raw_ostream that writes to a file descriptor.
Definition: raw_ostream.h:440
llvm::StringRef::size
constexpr size_t size() const
size - Get the string size.
Definition: StringRef.h:137
llvm::NVPTX::PTXLdStInstCode::V2
@ V2
Definition: NVPTX.h:123
llvm::MemoryBuffer::getBufferStart
const char * getBufferStart() const
Definition: MemoryBuffer.h:65
llvm::Error
Lightweight error class with error context and mandatory checking.
Definition: Error.h:155
llvm::DiffFilesWithTolerance
int DiffFilesWithTolerance(StringRef FileA, StringRef FileB, double AbsTol, double RelTol, std::string *Error=nullptr)
DiffFilesWithTolerance - Compare the two files specified, returning 0 if the files match,...
Definition: FileUtilities.cpp:178
CompareNumbers
static bool CompareNumbers(const char *&F1P, const char *&F2P, const char *F1End, const char *F2End, double AbsTolerance, double RelTolerance, std::string *ErrorMsg)
CompareNumbers - compare two numbers, returning true if they are different.
Definition: FileUtilities.cpp:85
llvm::sys::fs::perms
perms
Definition: FileSystem.h:86
llvm::sys::fs::getUmask
unsigned getUmask()
Get file creation mode mask of the process.
llvm::ErrorOr::get
reference get()
Definition: ErrorOr.h:150
OutputFilename
static cl::opt< std::string > OutputFilename("o", cl::desc("Output filename"), cl::value_desc("filename"), cl::init("-"))
llvm::sys::fs::status
std::error_code status(const Twine &path, file_status &result, bool follow=true)
Get file status as if by POSIX stat().
llvm::ErrorOr
Represents either an error or a value T.
Definition: ErrorOr.h:56
isNumberChar
static bool isNumberChar(char C)
Definition: FileUtilities.cpp:45
EndOfNumber
static const char * EndOfNumber(const char *Pos)
EndOfNumber - Return the first character that is not part of the specified number.
Definition: FileUtilities.cpp:78
raw_ostream.h
llvm::abs
APFloat abs(APFloat X)
Returns the absolute value of the argument.
Definition: APFloat.h:1288