1 //===- Job.cpp - Command to Execute ---------------------------------------===//
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 #include "clang/Driver/Job.h"
10 #include "clang/Basic/LLVM.h"
11 #include "clang/Driver/Driver.h"
12 #include "clang/Driver/DriverDiagnostic.h"
13 #include "clang/Driver/InputInfo.h"
14 #include "clang/Driver/Tool.h"
15 #include "clang/Driver/ToolChain.h"
16 #include "llvm/ADT/ArrayRef.h"
17 #include "llvm/ADT/SmallString.h"
18 #include "llvm/ADT/SmallVector.h"
19 #include "llvm/ADT/StringExtras.h"
20 #include "llvm/ADT/StringRef.h"
21 #include "llvm/ADT/StringSet.h"
22 #include "llvm/ADT/StringSwitch.h"
23 #include "llvm/Support/CrashRecoveryContext.h"
24 #include "llvm/Support/FileSystem.h"
25 #include "llvm/Support/Path.h"
26 #include "llvm/Support/PrettyStackTrace.h"
27 #include "llvm/Support/Program.h"
28 #include "llvm/Support/raw_ostream.h"
29 #include <algorithm>
30 #include <cassert>
31 #include <cstddef>
32 #include <string>
33 #include <system_error>
34 #include <utility>
35 
36 using namespace clang;
37 using namespace driver;
38 
39 Command::Command(const Action &Source, const Tool &Creator,
40                  ResponseFileSupport ResponseSupport, const char *Executable,
41                  const llvm::opt::ArgStringList &Arguments,
42                  ArrayRef<InputInfo> Inputs, ArrayRef<InputInfo> Outputs,
43                  const char *PrependArg)
44     : Source(Source), Creator(Creator), ResponseSupport(ResponseSupport),
45       Executable(Executable), PrependArg(PrependArg), Arguments(Arguments) {
46   for (const auto &II : Inputs)
47     if (II.isFilename())
48       InputInfoList.push_back(II);
49   for (const auto &II : Outputs)
50     if (II.isFilename())
51       OutputFilenames.push_back(II.getFilename());
52 }
53 
54 /// Check if the compiler flag in question should be skipped when
55 /// emitting a reproducer. Also track how many arguments it has and if the
56 /// option is some kind of include path.
57 static bool skipArgs(const char *Flag, bool HaveCrashVFS, int &SkipNum,
58                      bool &IsInclude) {
59   SkipNum = 2;
60   // These flags are all of the form -Flag <Arg> and are treated as two
61   // arguments.  Therefore, we need to skip the flag and the next argument.
62   bool ShouldSkip = llvm::StringSwitch<bool>(Flag)
63     .Cases("-MF", "-MT", "-MQ", "-serialize-diagnostic-file", true)
64     .Cases("-o", "-dependency-file", true)
65     .Cases("-fdebug-compilation-dir", "-diagnostic-log-file", true)
66     .Cases("-dwarf-debug-flags", "-ivfsoverlay", true)
67     .Default(false);
68   if (ShouldSkip)
69     return true;
70 
71   // Some include flags shouldn't be skipped if we have a crash VFS
72   IsInclude = llvm::StringSwitch<bool>(Flag)
73     .Cases("-include", "-header-include-file", true)
74     .Cases("-idirafter", "-internal-isystem", "-iwithprefix", true)
75     .Cases("-internal-externc-isystem", "-iprefix", true)
76     .Cases("-iwithprefixbefore", "-isystem", "-iquote", true)
77     .Cases("-isysroot", "-I", "-F", "-resource-dir", true)
78     .Cases("-iframework", "-include-pch", true)
79     .Default(false);
80   if (IsInclude)
81     return !HaveCrashVFS;
82 
83   // The remaining flags are treated as a single argument.
84 
85   // These flags are all of the form -Flag and have no second argument.
86   ShouldSkip = llvm::StringSwitch<bool>(Flag)
87     .Cases("-M", "-MM", "-MG", "-MP", "-MD", true)
88     .Case("-MMD", true)
89     .Default(false);
90 
91   // Match found.
92   SkipNum = 1;
93   if (ShouldSkip)
94     return true;
95 
96   // These flags are treated as a single argument (e.g., -F<Dir>).
97   StringRef FlagRef(Flag);
98   IsInclude = FlagRef.startswith("-F") || FlagRef.startswith("-I");
99   if (IsInclude)
100     return !HaveCrashVFS;
101   if (FlagRef.startswith("-fmodules-cache-path="))
102     return true;
103 
104   SkipNum = 0;
105   return false;
106 }
107 
108 void Command::writeResponseFile(raw_ostream &OS) const {
109   // In a file list, we only write the set of inputs to the response file
110   if (ResponseSupport.ResponseKind == ResponseFileSupport::RF_FileList) {
111     for (const auto *Arg : InputFileList) {
112       OS << Arg << '\n';
113     }
114     return;
115   }
116 
117   // In regular response files, we send all arguments to the response file.
118   // Wrapping all arguments in double quotes ensures that both Unix tools and
119   // Windows tools understand the response file.
120   for (const auto *Arg : Arguments) {
121     OS << '"';
122 
123     for (; *Arg != '\0'; Arg++) {
124       if (*Arg == '\"' || *Arg == '\\') {
125         OS << '\\';
126       }
127       OS << *Arg;
128     }
129 
130     OS << "\" ";
131   }
132 }
133 
134 void Command::buildArgvForResponseFile(
135     llvm::SmallVectorImpl<const char *> &Out) const {
136   // When not a file list, all arguments are sent to the response file.
137   // This leaves us to set the argv to a single parameter, requesting the tool
138   // to read the response file.
139   if (ResponseSupport.ResponseKind != ResponseFileSupport::RF_FileList) {
140     Out.push_back(Executable);
141     Out.push_back(ResponseFileFlag.c_str());
142     return;
143   }
144 
145   llvm::StringSet<> Inputs;
146   for (const auto *InputName : InputFileList)
147     Inputs.insert(InputName);
148   Out.push_back(Executable);
149 
150   if (PrependArg)
151     Out.push_back(PrependArg);
152 
153   // In a file list, build args vector ignoring parameters that will go in the
154   // response file (elements of the InputFileList vector)
155   bool FirstInput = true;
156   for (const auto *Arg : Arguments) {
157     if (Inputs.count(Arg) == 0) {
158       Out.push_back(Arg);
159     } else if (FirstInput) {
160       FirstInput = false;
161       Out.push_back(ResponseSupport.ResponseFlag);
162       Out.push_back(ResponseFile);
163     }
164   }
165 }
166 
167 /// Rewrite relative include-like flag paths to absolute ones.
168 static void
169 rewriteIncludes(const llvm::ArrayRef<const char *> &Args, size_t Idx,
170                 size_t NumArgs,
171                 llvm::SmallVectorImpl<llvm::SmallString<128>> &IncFlags) {
172   using namespace llvm;
173   using namespace sys;
174 
175   auto getAbsPath = [](StringRef InInc, SmallVectorImpl<char> &OutInc) -> bool {
176     if (path::is_absolute(InInc)) // Nothing to do here...
177       return false;
178     std::error_code EC = fs::current_path(OutInc);
179     if (EC)
180       return false;
181     path::append(OutInc, InInc);
182     return true;
183   };
184 
185   SmallString<128> NewInc;
186   if (NumArgs == 1) {
187     StringRef FlagRef(Args[Idx + NumArgs - 1]);
188     assert((FlagRef.startswith("-F") || FlagRef.startswith("-I")) &&
189             "Expecting -I or -F");
190     StringRef Inc = FlagRef.slice(2, StringRef::npos);
191     if (getAbsPath(Inc, NewInc)) {
192       SmallString<128> NewArg(FlagRef.slice(0, 2));
193       NewArg += NewInc;
194       IncFlags.push_back(std::move(NewArg));
195     }
196     return;
197   }
198 
199   assert(NumArgs == 2 && "Not expecting more than two arguments");
200   StringRef Inc(Args[Idx + NumArgs - 1]);
201   if (!getAbsPath(Inc, NewInc))
202     return;
203   IncFlags.push_back(SmallString<128>(Args[Idx]));
204   IncFlags.push_back(std::move(NewInc));
205 }
206 
207 void Command::Print(raw_ostream &OS, const char *Terminator, bool Quote,
208                     CrashReportInfo *CrashInfo) const {
209   // Always quote the exe.
210   OS << ' ';
211   llvm::sys::printArg(OS, Executable, /*Quote=*/true);
212 
213   ArrayRef<const char *> Args = Arguments;
214   SmallVector<const char *, 128> ArgsRespFile;
215   if (ResponseFile != nullptr) {
216     buildArgvForResponseFile(ArgsRespFile);
217     Args = ArrayRef<const char *>(ArgsRespFile).slice(1); // no executable name
218   } else if (PrependArg) {
219     OS << ' ';
220     llvm::sys::printArg(OS, PrependArg, /*Quote=*/true);
221   }
222 
223   bool HaveCrashVFS = CrashInfo && !CrashInfo->VFSPath.empty();
224   for (size_t i = 0, e = Args.size(); i < e; ++i) {
225     const char *const Arg = Args[i];
226 
227     if (CrashInfo) {
228       int NumArgs = 0;
229       bool IsInclude = false;
230       if (skipArgs(Arg, HaveCrashVFS, NumArgs, IsInclude)) {
231         i += NumArgs - 1;
232         continue;
233       }
234 
235       // Relative includes need to be expanded to absolute paths.
236       if (HaveCrashVFS && IsInclude) {
237         SmallVector<SmallString<128>, 2> NewIncFlags;
238         rewriteIncludes(Args, i, NumArgs, NewIncFlags);
239         if (!NewIncFlags.empty()) {
240           for (auto &F : NewIncFlags) {
241             OS << ' ';
242             llvm::sys::printArg(OS, F.c_str(), Quote);
243           }
244           i += NumArgs - 1;
245           continue;
246         }
247       }
248 
249       auto Found = llvm::find_if(InputInfoList, [&Arg](const InputInfo &II) {
250         return II.getFilename() == Arg;
251       });
252       if (Found != InputInfoList.end() &&
253           (i == 0 || StringRef(Args[i - 1]) != "-main-file-name")) {
254         // Replace the input file name with the crashinfo's file name.
255         OS << ' ';
256         StringRef ShortName = llvm::sys::path::filename(CrashInfo->Filename);
257         llvm::sys::printArg(OS, ShortName.str(), Quote);
258         continue;
259       }
260     }
261 
262     OS << ' ';
263     llvm::sys::printArg(OS, Arg, Quote);
264   }
265 
266   if (CrashInfo && HaveCrashVFS) {
267     OS << ' ';
268     llvm::sys::printArg(OS, "-ivfsoverlay", Quote);
269     OS << ' ';
270     llvm::sys::printArg(OS, CrashInfo->VFSPath.str(), Quote);
271 
272     // The leftover modules from the crash are stored in
273     //  <name>.cache/vfs/modules
274     // Leave it untouched for pcm inspection and provide a clean/empty dir
275     // path to contain the future generated module cache:
276     //  <name>.cache/vfs/repro-modules
277     SmallString<128> RelModCacheDir = llvm::sys::path::parent_path(
278         llvm::sys::path::parent_path(CrashInfo->VFSPath));
279     llvm::sys::path::append(RelModCacheDir, "repro-modules");
280 
281     std::string ModCachePath = "-fmodules-cache-path=";
282     ModCachePath.append(RelModCacheDir.c_str());
283 
284     OS << ' ';
285     llvm::sys::printArg(OS, ModCachePath, Quote);
286   }
287 
288   if (ResponseFile != nullptr) {
289     OS << "\n Arguments passed via response file:\n";
290     writeResponseFile(OS);
291     // Avoiding duplicated newline terminator, since FileLists are
292     // newline-separated.
293     if (ResponseSupport.ResponseKind != ResponseFileSupport::RF_FileList)
294       OS << "\n";
295     OS << " (end of response file)";
296   }
297 
298   OS << Terminator;
299 }
300 
301 void Command::setResponseFile(const char *FileName) {
302   ResponseFile = FileName;
303   ResponseFileFlag = ResponseSupport.ResponseFlag;
304   ResponseFileFlag += FileName;
305 }
306 
307 void Command::setEnvironment(llvm::ArrayRef<const char *> NewEnvironment) {
308   Environment.reserve(NewEnvironment.size() + 1);
309   Environment.assign(NewEnvironment.begin(), NewEnvironment.end());
310   Environment.push_back(nullptr);
311 }
312 
313 void Command::setRedirectFiles(
314     const std::vector<std::optional<std::string>> &Redirects) {
315   RedirectFiles = Redirects;
316 }
317 
318 void Command::PrintFileNames() const {
319   if (PrintInputFilenames) {
320     for (const auto &Arg : InputInfoList)
321       llvm::outs() << llvm::sys::path::filename(Arg.getFilename()) << "\n";
322     llvm::outs().flush();
323   }
324 }
325 
326 int Command::Execute(ArrayRef<std::optional<StringRef>> Redirects,
327                      std::string *ErrMsg, bool *ExecutionFailed) const {
328   PrintFileNames();
329 
330   SmallVector<const char *, 128> Argv;
331   if (ResponseFile == nullptr) {
332     Argv.push_back(Executable);
333     if (PrependArg)
334       Argv.push_back(PrependArg);
335     Argv.append(Arguments.begin(), Arguments.end());
336     Argv.push_back(nullptr);
337   } else {
338     // If the command is too large, we need to put arguments in a response file.
339     std::string RespContents;
340     llvm::raw_string_ostream SS(RespContents);
341 
342     // Write file contents and build the Argv vector
343     writeResponseFile(SS);
344     buildArgvForResponseFile(Argv);
345     Argv.push_back(nullptr);
346     SS.flush();
347 
348     // Save the response file in the appropriate encoding
349     if (std::error_code EC = writeFileWithEncoding(
350             ResponseFile, RespContents, ResponseSupport.ResponseEncoding)) {
351       if (ErrMsg)
352         *ErrMsg = EC.message();
353       if (ExecutionFailed)
354         *ExecutionFailed = true;
355       // Return -1 by convention (see llvm/include/llvm/Support/Program.h) to
356       // indicate the requested executable cannot be started.
357       return -1;
358     }
359   }
360 
361   std::optional<ArrayRef<StringRef>> Env;
362   std::vector<StringRef> ArgvVectorStorage;
363   if (!Environment.empty()) {
364     assert(Environment.back() == nullptr &&
365            "Environment vector should be null-terminated by now");
366     ArgvVectorStorage = llvm::toStringRefArray(Environment.data());
367     Env = ArrayRef(ArgvVectorStorage);
368   }
369 
370   auto Args = llvm::toStringRefArray(Argv.data());
371 
372   // Use Job-specific redirect files if they are present.
373   if (!RedirectFiles.empty()) {
374     std::vector<std::optional<StringRef>> RedirectFilesOptional;
375     for (const auto &Ele : RedirectFiles)
376       if (Ele)
377         RedirectFilesOptional.push_back(std::optional<StringRef>(*Ele));
378       else
379         RedirectFilesOptional.push_back(std::nullopt);
380 
381     return llvm::sys::ExecuteAndWait(Executable, Args, Env,
382                                      ArrayRef(RedirectFilesOptional),
383                                      /*secondsToWait=*/0, /*memoryLimit=*/0,
384                                      ErrMsg, ExecutionFailed, &ProcStat);
385   }
386 
387   return llvm::sys::ExecuteAndWait(Executable, Args, Env, Redirects,
388                                    /*secondsToWait*/ 0, /*memoryLimit*/ 0,
389                                    ErrMsg, ExecutionFailed, &ProcStat);
390 }
391 
392 CC1Command::CC1Command(const Action &Source, const Tool &Creator,
393                        ResponseFileSupport ResponseSupport,
394                        const char *Executable,
395                        const llvm::opt::ArgStringList &Arguments,
396                        ArrayRef<InputInfo> Inputs, ArrayRef<InputInfo> Outputs,
397                        const char *PrependArg)
398     : Command(Source, Creator, ResponseSupport, Executable, Arguments, Inputs,
399               Outputs, PrependArg) {
400   InProcess = true;
401 }
402 
403 void CC1Command::Print(raw_ostream &OS, const char *Terminator, bool Quote,
404                        CrashReportInfo *CrashInfo) const {
405   if (InProcess)
406     OS << " (in-process)\n";
407   Command::Print(OS, Terminator, Quote, CrashInfo);
408 }
409 
410 int CC1Command::Execute(ArrayRef<std::optional<StringRef>> Redirects,
411                         std::string *ErrMsg, bool *ExecutionFailed) const {
412   // FIXME: Currently, if there're more than one job, we disable
413   // -fintegrate-cc1. If we're no longer a integrated-cc1 job, fallback to
414   // out-of-process execution. See discussion in https://reviews.llvm.org/D74447
415   if (!InProcess)
416     return Command::Execute(Redirects, ErrMsg, ExecutionFailed);
417 
418   PrintFileNames();
419 
420   SmallVector<const char *, 128> Argv;
421   Argv.push_back(getExecutable());
422   Argv.append(getArguments().begin(), getArguments().end());
423   Argv.push_back(nullptr);
424   Argv.pop_back(); // The terminating null element shall not be part of the
425                    // slice (main() behavior).
426 
427   // This flag simply indicates that the program couldn't start, which isn't
428   // applicable here.
429   if (ExecutionFailed)
430     *ExecutionFailed = false;
431 
432   llvm::CrashRecoveryContext CRC;
433   CRC.DumpStackAndCleanupOnFailure = true;
434 
435   const void *PrettyState = llvm::SavePrettyStackState();
436   const Driver &D = getCreator().getToolChain().getDriver();
437 
438   int R = 0;
439   // Enter ExecuteCC1Tool() instead of starting up a new process
440   if (!CRC.RunSafely([&]() { R = D.CC1Main(Argv); })) {
441     llvm::RestorePrettyStackState(PrettyState);
442     return CRC.RetCode;
443   }
444   return R;
445 }
446 
447 void CC1Command::setEnvironment(llvm::ArrayRef<const char *> NewEnvironment) {
448   // We don't support set a new environment when calling into ExecuteCC1Tool()
449   llvm_unreachable(
450       "The CC1Command doesn't support changing the environment vars!");
451 }
452 
453 void JobList::Print(raw_ostream &OS, const char *Terminator, bool Quote,
454                     CrashReportInfo *CrashInfo) const {
455   for (const auto &Job : *this)
456     Job.Print(OS, Terminator, Quote, CrashInfo);
457 }
458 
459 void JobList::clear() { Jobs.clear(); }
460