xref: /llvm-project/lld/MachO/LTO.cpp (revision b4e000e6005bd0f11240133aa335efcbb8424a23) 
1 //===- LTO.cpp ------------------------------------------------------------===//
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 "LTO.h"
10 #include "Config.h"
11 #include "Driver.h"
12 #include "InputFiles.h"
13 #include "Symbols.h"
14 #include "Target.h"
15 
16 #include "lld/Common/Args.h"
17 #include "lld/Common/CommonLinkerContext.h"
18 #include "lld/Common/Filesystem.h"
19 #include "lld/Common/Strings.h"
20 #include "lld/Common/TargetOptionsCommandFlags.h"
21 #include "llvm/Bitcode/BitcodeWriter.h"
22 #include "llvm/LTO/Config.h"
23 #include "llvm/LTO/LTO.h"
24 #include "llvm/Support/Caching.h"
25 #include "llvm/Support/FileSystem.h"
26 #include "llvm/Support/Path.h"
27 #include "llvm/Support/raw_ostream.h"
28 #include "llvm/Transforms/ObjCARC.h"
29 
30 using namespace lld;
31 using namespace lld::macho;
32 using namespace llvm;
33 using namespace llvm::MachO;
34 using namespace llvm::sys;
35 
36 static std::string getThinLTOOutputFile(StringRef modulePath) {
37   return lto::getThinLTOOutputFile(modulePath, config->thinLTOPrefixReplaceOld,
38                                    config->thinLTOPrefixReplaceNew);
39 }
40 
41 static lto::Config createConfig() {
42   lto::Config c;
43   c.Options = initTargetOptionsFromCodeGenFlags();
44   c.Options.EmitAddrsig = config->icfLevel == ICFLevel::safe;
45   for (StringRef C : config->mllvmOpts)
46     c.MllvmArgs.emplace_back(C.str());
47   for (StringRef pluginFn : config->passPlugins)
48     c.PassPlugins.push_back(std::string(pluginFn));
49   c.OptPipeline = std::string(config->ltoNewPmPasses);
50   c.CodeModel = getCodeModelFromCMModel();
51   c.CPU = getCPUStr();
52   c.MAttrs = getMAttrs();
53   c.DiagHandler = diagnosticHandler;
54 
55   c.AlwaysEmitRegularLTOObj = !config->ltoObjPath.empty();
56 
57   c.TimeTraceEnabled = config->timeTraceEnabled;
58   c.TimeTraceGranularity = config->timeTraceGranularity;
59   c.DebugPassManager = config->ltoDebugPassManager;
60   c.CSIRProfile = std::string(config->csProfilePath);
61   c.RunCSIRInstr = config->csProfileGenerate;
62   c.PGOWarnMismatch = config->pgoWarnMismatch;
63   c.OptLevel = config->ltoo;
64   c.CGOptLevel = config->ltoCgo;
65   if (config->saveTemps)
66     checkError(c.addSaveTemps(config->outputFile.str() + ".",
67                               /*UseInputModulePath=*/true));
68   return c;
69 }
70 
71 // If `originalPath` exists, hardlinks `path` to `originalPath`. If that fails,
72 // or `originalPath` is not set, saves `buffer` to `path`.
73 static void saveOrHardlinkBuffer(StringRef buffer, const Twine &path,
74                                  std::optional<StringRef> originalPath) {
75   if (originalPath) {
76     auto err = fs::create_hard_link(*originalPath, path);
77     if (!err)
78       return;
79   }
80   saveBuffer(buffer, path);
81 }
82 
83 BitcodeCompiler::BitcodeCompiler() {
84   // Initialize indexFile.
85   if (!config->thinLTOIndexOnlyArg.empty())
86     indexFile = openFile(config->thinLTOIndexOnlyArg);
87 
88   // Initialize ltoObj.
89   lto::ThinBackend backend;
90   auto onIndexWrite = [&](StringRef S) { thinIndices.erase(S); };
91   if (config->thinLTOIndexOnly) {
92     backend = lto::createWriteIndexesThinBackend(
93         llvm::hardware_concurrency(config->thinLTOJobs),
94         std::string(config->thinLTOPrefixReplaceOld),
95         std::string(config->thinLTOPrefixReplaceNew),
96         std::string(config->thinLTOPrefixReplaceNativeObject),
97         config->thinLTOEmitImportsFiles, indexFile.get(), onIndexWrite);
98   } else {
99     backend = lto::createInProcessThinBackend(
100         llvm::heavyweight_hardware_concurrency(config->thinLTOJobs),
101         onIndexWrite, config->thinLTOEmitIndexFiles,
102         config->thinLTOEmitImportsFiles);
103   }
104 
105   ltoObj = std::make_unique<lto::LTO>(createConfig(), backend);
106 }
107 
108 void BitcodeCompiler::add(BitcodeFile &f) {
109   lto::InputFile &obj = *f.obj;
110 
111   if (config->thinLTOEmitIndexFiles)
112     thinIndices.insert(obj.getName());
113 
114   ArrayRef<lto::InputFile::Symbol> objSyms = obj.symbols();
115   std::vector<lto::SymbolResolution> resols;
116   resols.reserve(objSyms.size());
117 
118   // Provide a resolution to the LTO API for each symbol.
119   bool exportDynamic =
120       config->outputType != MH_EXECUTE || config->exportDynamic;
121   auto symIt = f.symbols.begin();
122   for (const lto::InputFile::Symbol &objSym : objSyms) {
123     resols.emplace_back();
124     lto::SymbolResolution &r = resols.back();
125     Symbol *sym = *symIt++;
126 
127     // Ideally we shouldn't check for SF_Undefined but currently IRObjectFile
128     // reports two symbols for module ASM defined. Without this check, lld
129     // flags an undefined in IR with a definition in ASM as prevailing.
130     // Once IRObjectFile is fixed to report only one symbol this hack can
131     // be removed.
132     r.Prevailing = !objSym.isUndefined() && sym->getFile() == &f;
133 
134     if (const auto *defined = dyn_cast<Defined>(sym)) {
135       r.ExportDynamic =
136           defined->isExternal() && !defined->privateExtern && exportDynamic;
137       r.FinalDefinitionInLinkageUnit =
138           !defined->isExternalWeakDef() && !defined->interposable;
139     } else if (const auto *common = dyn_cast<CommonSymbol>(sym)) {
140       r.ExportDynamic = !common->privateExtern && exportDynamic;
141       r.FinalDefinitionInLinkageUnit = true;
142     }
143 
144     r.VisibleToRegularObj =
145         sym->isUsedInRegularObj || (r.Prevailing && r.ExportDynamic);
146 
147     // Un-define the symbol so that we don't get duplicate symbol errors when we
148     // load the ObjFile emitted by LTO compilation.
149     if (r.Prevailing)
150       replaceSymbol<Undefined>(sym, sym->getName(), sym->getFile(),
151                                RefState::Strong, /*wasBitcodeSymbol=*/true);
152 
153     // TODO: set the other resolution configs properly
154   }
155   checkError(ltoObj->add(std::move(f.obj), resols));
156   hasFiles = true;
157 }
158 
159 // If LazyObjFile has not been added to link, emit empty index files.
160 // This is needed because this is what GNU gold plugin does and we have a
161 // distributed build system that depends on that behavior.
162 static void thinLTOCreateEmptyIndexFiles() {
163   DenseSet<StringRef> linkedBitCodeFiles;
164   for (InputFile *file : inputFiles)
165     if (auto *f = dyn_cast<BitcodeFile>(file))
166       if (!f->lazy)
167         linkedBitCodeFiles.insert(f->getName());
168 
169   for (InputFile *file : inputFiles) {
170     if (auto *f = dyn_cast<BitcodeFile>(file)) {
171       if (!f->lazy)
172         continue;
173       if (linkedBitCodeFiles.contains(f->getName()))
174         continue;
175       std::string path =
176           replaceThinLTOSuffix(getThinLTOOutputFile(f->obj->getName()));
177       std::unique_ptr<raw_fd_ostream> os = openFile(path + ".thinlto.bc");
178       if (!os)
179         continue;
180 
181       ModuleSummaryIndex m(/*HaveGVs=*/false);
182       m.setSkipModuleByDistributedBackend();
183       writeIndexToFile(m, *os);
184       if (config->thinLTOEmitImportsFiles)
185         openFile(path + ".imports");
186     }
187   }
188 }
189 
190 // Merge all the bitcode files we have seen, codegen the result
191 // and return the resulting ObjectFile(s).
192 std::vector<ObjFile *> BitcodeCompiler::compile() {
193   unsigned maxTasks = ltoObj->getMaxTasks();
194   buf.resize(maxTasks);
195   files.resize(maxTasks);
196 
197   // The -cache_path_lto option specifies the path to a directory in which
198   // to cache native object files for ThinLTO incremental builds. If a path was
199   // specified, configure LTO to use it as the cache directory.
200   FileCache cache;
201   if (!config->thinLTOCacheDir.empty())
202     cache = check(localCache("ThinLTO", "Thin", config->thinLTOCacheDir,
203                              [&](size_t task, const Twine &moduleName,
204                                  std::unique_ptr<MemoryBuffer> mb) {
205                                files[task] = std::move(mb);
206                              }));
207 
208   if (hasFiles)
209     checkError(ltoObj->run(
210         [&](size_t task, const Twine &moduleName) {
211           return std::make_unique<CachedFileStream>(
212               std::make_unique<raw_svector_ostream>(buf[task]));
213         },
214         cache));
215 
216   // Emit empty index files for non-indexed files
217   for (StringRef s : thinIndices) {
218     std::string path = getThinLTOOutputFile(s);
219     openFile(path + ".thinlto.bc");
220     if (config->thinLTOEmitImportsFiles)
221       openFile(path + ".imports");
222   }
223 
224   if (config->thinLTOEmitIndexFiles)
225     thinLTOCreateEmptyIndexFiles();
226 
227   // In ThinLTO mode, Clang passes a temporary directory in -object_path_lto,
228   // while the argument is a single file in FullLTO mode.
229   bool objPathIsDir = true;
230   if (!config->ltoObjPath.empty()) {
231     if (std::error_code ec = fs::create_directories(config->ltoObjPath))
232       fatal("cannot create LTO object path " + config->ltoObjPath + ": " +
233             ec.message());
234 
235     if (!fs::is_directory(config->ltoObjPath)) {
236       objPathIsDir = false;
237       unsigned objCount =
238           count_if(buf, [](const SmallString<0> &b) { return !b.empty(); });
239       if (objCount > 1)
240         fatal("-object_path_lto must specify a directory when using ThinLTO");
241     }
242   }
243 
244   auto outputFilePath = [objPathIsDir](int i) {
245     SmallString<261> filePath("/tmp/lto.tmp");
246     if (!config->ltoObjPath.empty()) {
247       filePath = config->ltoObjPath;
248       if (objPathIsDir)
249         path::append(filePath, Twine(i) + "." +
250                                    getArchitectureName(config->arch()) +
251                                    ".lto.o");
252     }
253     return filePath;
254   };
255 
256   // ThinLTO with index only option is required to generate only the index
257   // files. After that, we exit from linker and ThinLTO backend runs in a
258   // distributed environment.
259   if (config->thinLTOIndexOnly) {
260     if (!config->ltoObjPath.empty())
261       saveBuffer(buf[0], outputFilePath(0));
262     if (indexFile)
263       indexFile->close();
264     return {};
265   }
266 
267   if (!config->thinLTOCacheDir.empty())
268     pruneCache(config->thinLTOCacheDir, config->thinLTOCachePolicy, files);
269 
270   std::vector<ObjFile *> ret;
271   for (unsigned i = 0; i < maxTasks; ++i) {
272     // Get the native object contents either from the cache or from memory.  Do
273     // not use the cached MemoryBuffer directly to ensure dsymutil does not
274     // race with the cache pruner.
275     StringRef objBuf;
276     std::optional<StringRef> cachePath;
277     if (files[i]) {
278       objBuf = files[i]->getBuffer();
279       cachePath = files[i]->getBufferIdentifier();
280     } else {
281       objBuf = buf[i];
282     }
283     if (objBuf.empty())
284       continue;
285 
286     // FIXME: should `saveTemps` and `ltoObjPath` use the same file name?
287     if (config->saveTemps)
288       saveBuffer(objBuf,
289                  config->outputFile + ((i == 0) ? "" : Twine(i)) + ".lto.o");
290 
291     auto filePath = outputFilePath(i);
292     uint32_t modTime = 0;
293     if (!config->ltoObjPath.empty()) {
294       saveOrHardlinkBuffer(objBuf, filePath, cachePath);
295       modTime = getModTime(filePath);
296     }
297     ret.push_back(make<ObjFile>(
298         MemoryBufferRef(objBuf, saver().save(filePath.str())), modTime,
299         /*archiveName=*/"", /*lazy=*/false,
300         /*forceHidden=*/false, /*compatArch=*/true, /*builtFromBitcode=*/true));
301   }
302 
303   return ret;
304 }
305