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 "InputFiles.h"
12 #include "LinkerScript.h"
13 #include "SymbolTable.h"
14 #include "Symbols.h"
15 #include "lld/Common/Args.h"
16 #include "lld/Common/ErrorHandler.h"
17 #include "lld/Common/TargetOptionsCommandFlags.h"
18 #include "llvm/ADT/STLExtras.h"
19 #include "llvm/ADT/SmallString.h"
20 #include "llvm/ADT/StringRef.h"
21 #include "llvm/ADT/Twine.h"
22 #include "llvm/BinaryFormat/ELF.h"
23 #include "llvm/Bitcode/BitcodeReader.h"
24 #include "llvm/Bitcode/BitcodeWriter.h"
25 #include "llvm/IR/DiagnosticPrinter.h"
26 #include "llvm/LTO/Caching.h"
27 #include "llvm/LTO/Config.h"
28 #include "llvm/LTO/LTO.h"
29 #include "llvm/Object/SymbolicFile.h"
30 #include "llvm/Support/CodeGen.h"
31 #include "llvm/Support/Error.h"
32 #include "llvm/Support/FileSystem.h"
33 #include "llvm/Support/MemoryBuffer.h"
34 #include <algorithm>
35 #include <cstddef>
36 #include <memory>
37 #include <string>
38 #include <system_error>
39 #include <vector>
40 
41 using namespace llvm;
42 using namespace llvm::object;
43 using namespace llvm::ELF;
44 using namespace lld;
45 using namespace lld::elf;
46 
47 // Creates an empty file to store a list of object files for final
48 // linking of distributed ThinLTO.
openFile(StringRef file)49 static std::unique_ptr<raw_fd_ostream> openFile(StringRef file) {
50   std::error_code ec;
51   auto ret =
52       std::make_unique<raw_fd_ostream>(file, ec, sys::fs::OpenFlags::OF_None);
53   if (ec) {
54     error("cannot open " + file + ": " + ec.message());
55     return nullptr;
56   }
57   return ret;
58 }
59 
60 // The merged bitcode after LTO is large. Try opening a file stream that
61 // supports reading, seeking and writing. Such a file allows BitcodeWriter to
62 // flush buffered data to reduce memory consumption. If this fails, open a file
63 // stream that supports only write.
openLTOOutputFile(StringRef file)64 static std::unique_ptr<raw_fd_ostream> openLTOOutputFile(StringRef file) {
65   std::error_code ec;
66   std::unique_ptr<raw_fd_ostream> fs =
67       std::make_unique<raw_fd_stream>(file, ec);
68   if (!ec)
69     return fs;
70   return openFile(file);
71 }
72 
getThinLTOOutputFile(StringRef modulePath)73 static std::string getThinLTOOutputFile(StringRef modulePath) {
74   return lto::getThinLTOOutputFile(
75       std::string(modulePath), std::string(config->thinLTOPrefixReplace.first),
76       std::string(config->thinLTOPrefixReplace.second));
77 }
78 
createConfig()79 static lto::Config createConfig() {
80   lto::Config c;
81 
82   // LLD supports the new relocations and address-significance tables.
83   c.Options = initTargetOptionsFromCodeGenFlags();
84   c.Options.RelaxELFRelocations = true;
85   c.Options.EmitAddrsig = true;
86 
87   // Always emit a section per function/datum with LTO.
88   c.Options.FunctionSections = true;
89   c.Options.DataSections = true;
90 
91   // Check if basic block sections must be used.
92   // Allowed values for --lto-basic-block-sections are "all", "labels",
93   // "<file name specifying basic block ids>", or none.  This is the equivalent
94   // of -fbasic-block-sections= flag in clang.
95   if (!config->ltoBasicBlockSections.empty()) {
96     if (config->ltoBasicBlockSections == "all") {
97       c.Options.BBSections = BasicBlockSection::All;
98     } else if (config->ltoBasicBlockSections == "labels") {
99       c.Options.BBSections = BasicBlockSection::Labels;
100     } else if (config->ltoBasicBlockSections == "none") {
101       c.Options.BBSections = BasicBlockSection::None;
102     } else {
103       ErrorOr<std::unique_ptr<MemoryBuffer>> MBOrErr =
104           MemoryBuffer::getFile(config->ltoBasicBlockSections.str());
105       if (!MBOrErr) {
106         error("cannot open " + config->ltoBasicBlockSections + ":" +
107               MBOrErr.getError().message());
108       } else {
109         c.Options.BBSectionsFuncListBuf = std::move(*MBOrErr);
110       }
111       c.Options.BBSections = BasicBlockSection::List;
112     }
113   }
114 
115   c.Options.PseudoProbeForProfiling = config->ltoPseudoProbeForProfiling;
116   c.Options.UniqueBasicBlockSectionNames =
117       config->ltoUniqueBasicBlockSectionNames;
118 
119   if (auto relocModel = getRelocModelFromCMModel())
120     c.RelocModel = *relocModel;
121   else if (config->relocatable)
122     c.RelocModel = None;
123   else if (config->isPic)
124     c.RelocModel = Reloc::PIC_;
125   else
126     c.RelocModel = Reloc::Static;
127 
128   c.CodeModel = getCodeModelFromCMModel();
129   c.DisableVerify = config->disableVerify;
130   c.DiagHandler = diagnosticHandler;
131   c.OptLevel = config->ltoo;
132   c.CPU = getCPUStr();
133   c.MAttrs = getMAttrs();
134   c.CGOptLevel = args::getCGOptLevel(config->ltoo);
135 
136   c.PTO.LoopVectorization = c.OptLevel > 1;
137   c.PTO.SLPVectorization = c.OptLevel > 1;
138 
139   // Set up a custom pipeline if we've been asked to.
140   c.OptPipeline = std::string(config->ltoNewPmPasses);
141   c.AAPipeline = std::string(config->ltoAAPipeline);
142 
143   // Set up optimization remarks if we've been asked to.
144   c.RemarksFilename = std::string(config->optRemarksFilename);
145   c.RemarksPasses = std::string(config->optRemarksPasses);
146   c.RemarksWithHotness = config->optRemarksWithHotness;
147   c.RemarksHotnessThreshold = config->optRemarksHotnessThreshold;
148   c.RemarksFormat = std::string(config->optRemarksFormat);
149 
150   c.SampleProfile = std::string(config->ltoSampleProfile);
151   c.UseNewPM = config->ltoNewPassManager;
152   c.DebugPassManager = config->ltoDebugPassManager;
153   c.DwoDir = std::string(config->dwoDir);
154 
155   c.HasWholeProgramVisibility = config->ltoWholeProgramVisibility;
156   c.AlwaysEmitRegularLTOObj = !config->ltoObjPath.empty();
157 
158   for (const llvm::StringRef &name : config->thinLTOModulesToCompile)
159     c.ThinLTOModulesToCompile.emplace_back(name);
160 
161   c.TimeTraceEnabled = config->timeTraceEnabled;
162   c.TimeTraceGranularity = config->timeTraceGranularity;
163 
164   c.CSIRProfile = std::string(config->ltoCSProfileFile);
165   c.RunCSIRInstr = config->ltoCSProfileGenerate;
166 
167   if (config->emitLLVM) {
168     c.PostInternalizeModuleHook = [](size_t task, const Module &m) {
169       if (std::unique_ptr<raw_fd_ostream> os =
170               openLTOOutputFile(config->outputFile))
171         WriteBitcodeToFile(m, *os, false);
172       return false;
173     };
174   }
175 
176   if (config->ltoEmitAsm)
177     c.CGFileType = CGFT_AssemblyFile;
178 
179   if (config->saveTemps)
180     checkError(c.addSaveTemps(config->outputFile.str() + ".",
181                               /*UseInputModulePath*/ true));
182   return c;
183 }
184 
BitcodeCompiler()185 BitcodeCompiler::BitcodeCompiler() {
186   // Initialize indexFile.
187   if (!config->thinLTOIndexOnlyArg.empty())
188     indexFile = openFile(config->thinLTOIndexOnlyArg);
189 
190   // Initialize ltoObj.
191   lto::ThinBackend backend;
192   if (config->thinLTOIndexOnly) {
193     auto onIndexWrite = [&](StringRef s) { thinIndices.erase(s); };
194     backend = lto::createWriteIndexesThinBackend(
195         std::string(config->thinLTOPrefixReplace.first),
196         std::string(config->thinLTOPrefixReplace.second),
197         config->thinLTOEmitImportsFiles, indexFile.get(), onIndexWrite);
198   } else {
199     backend = lto::createInProcessThinBackend(
200         llvm::heavyweight_hardware_concurrency(config->thinLTOJobs));
201   }
202 
203   ltoObj = std::make_unique<lto::LTO>(createConfig(), backend,
204                                        config->ltoPartitions);
205 
206   // Initialize usedStartStop.
207   for (Symbol *sym : symtab->symbols()) {
208     StringRef s = sym->getName();
209     for (StringRef prefix : {"__start_", "__stop_"})
210       if (s.startswith(prefix))
211         usedStartStop.insert(s.substr(prefix.size()));
212   }
213 }
214 
215 BitcodeCompiler::~BitcodeCompiler() = default;
216 
add(BitcodeFile & f)217 void BitcodeCompiler::add(BitcodeFile &f) {
218   lto::InputFile &obj = *f.obj;
219   bool isExec = !config->shared && !config->relocatable;
220 
221   if (config->thinLTOIndexOnly)
222     thinIndices.insert(obj.getName());
223 
224   ArrayRef<Symbol *> syms = f.getSymbols();
225   ArrayRef<lto::InputFile::Symbol> objSyms = obj.symbols();
226   std::vector<lto::SymbolResolution> resols(syms.size());
227 
228   // Provide a resolution to the LTO API for each symbol.
229   for (size_t i = 0, e = syms.size(); i != e; ++i) {
230     Symbol *sym = syms[i];
231     const lto::InputFile::Symbol &objSym = objSyms[i];
232     lto::SymbolResolution &r = resols[i];
233 
234     // Ideally we shouldn't check for SF_Undefined but currently IRObjectFile
235     // reports two symbols for module ASM defined. Without this check, lld
236     // flags an undefined in IR with a definition in ASM as prevailing.
237     // Once IRObjectFile is fixed to report only one symbol this hack can
238     // be removed.
239     r.Prevailing = !objSym.isUndefined() && sym->file == &f;
240 
241     // We ask LTO to preserve following global symbols:
242     // 1) All symbols when doing relocatable link, so that them can be used
243     //    for doing final link.
244     // 2) Symbols that are used in regular objects.
245     // 3) C named sections if we have corresponding __start_/__stop_ symbol.
246     // 4) Symbols that are defined in bitcode files and used for dynamic linking.
247     r.VisibleToRegularObj = config->relocatable || sym->isUsedInRegularObj ||
248                             (r.Prevailing && sym->includeInDynsym()) ||
249                             usedStartStop.count(objSym.getSectionName());
250     // Identify symbols exported dynamically, and that therefore could be
251     // referenced by a shared library not visible to the linker.
252     r.ExportDynamic = sym->computeBinding() != STB_LOCAL &&
253                       (sym->isExportDynamic(sym->kind(), sym->visibility) ||
254                        sym->exportDynamic || sym->inDynamicList);
255     const auto *dr = dyn_cast<Defined>(sym);
256     r.FinalDefinitionInLinkageUnit =
257         (isExec || sym->visibility != STV_DEFAULT) && dr &&
258         // Skip absolute symbols from ELF objects, otherwise PC-rel relocations
259         // will be generated by for them, triggering linker errors.
260         // Symbol section is always null for bitcode symbols, hence the check
261         // for isElf(). Skip linker script defined symbols as well: they have
262         // no File defined.
263         !(dr->section == nullptr && (!sym->file || sym->file->isElf()));
264 
265     if (r.Prevailing)
266       sym->replace(Undefined{nullptr, sym->getName(), STB_GLOBAL, STV_DEFAULT,
267                              sym->type});
268 
269     // We tell LTO to not apply interprocedural optimization for wrapped
270     // (with --wrap) symbols because otherwise LTO would inline them while
271     // their values are still not final.
272     r.LinkerRedefined = !sym->canInline;
273   }
274   checkError(ltoObj->add(std::move(f.obj), resols));
275 }
276 
277 // If LazyObjFile has not been added to link, emit empty index files.
278 // This is needed because this is what GNU gold plugin does and we have a
279 // distributed build system that depends on that behavior.
thinLTOCreateEmptyIndexFiles()280 static void thinLTOCreateEmptyIndexFiles() {
281   for (LazyObjFile *f : lazyObjFiles) {
282     if (f->fetched || !isBitcode(f->mb))
283       continue;
284     std::string path = replaceThinLTOSuffix(getThinLTOOutputFile(f->getName()));
285     std::unique_ptr<raw_fd_ostream> os = openFile(path + ".thinlto.bc");
286     if (!os)
287       continue;
288 
289     ModuleSummaryIndex m(/*HaveGVs*/ false);
290     m.setSkipModuleByDistributedBackend();
291     WriteIndexToFile(m, *os);
292     if (config->thinLTOEmitImportsFiles)
293       openFile(path + ".imports");
294   }
295 }
296 
297 // Merge all the bitcode files we have seen, codegen the result
298 // and return the resulting ObjectFile(s).
compile()299 std::vector<InputFile *> BitcodeCompiler::compile() {
300   unsigned maxTasks = ltoObj->getMaxTasks();
301   buf.resize(maxTasks);
302   files.resize(maxTasks);
303 
304   // The --thinlto-cache-dir option specifies the path to a directory in which
305   // to cache native object files for ThinLTO incremental builds. If a path was
306   // specified, configure LTO to use it as the cache directory.
307   lto::NativeObjectCache cache;
308   if (!config->thinLTOCacheDir.empty())
309     cache = check(
310         lto::localCache(config->thinLTOCacheDir,
311                         [&](size_t task, std::unique_ptr<MemoryBuffer> mb) {
312                           files[task] = std::move(mb);
313                         }));
314 
315   if (!bitcodeFiles.empty())
316     checkError(ltoObj->run(
317         [&](size_t task) {
318           return std::make_unique<lto::NativeObjectStream>(
319               std::make_unique<raw_svector_ostream>(buf[task]));
320         },
321         cache));
322 
323   // Emit empty index files for non-indexed files but not in single-module mode.
324   if (config->thinLTOModulesToCompile.empty()) {
325     for (StringRef s : thinIndices) {
326       std::string path = getThinLTOOutputFile(s);
327       openFile(path + ".thinlto.bc");
328       if (config->thinLTOEmitImportsFiles)
329         openFile(path + ".imports");
330     }
331   }
332 
333   if (config->thinLTOIndexOnly) {
334     thinLTOCreateEmptyIndexFiles();
335 
336     if (!config->ltoObjPath.empty())
337       saveBuffer(buf[0], config->ltoObjPath);
338 
339     // ThinLTO with index only option is required to generate only the index
340     // files. After that, we exit from linker and ThinLTO backend runs in a
341     // distributed environment.
342     if (indexFile)
343       indexFile->close();
344     return {};
345   }
346 
347   if (!config->thinLTOCacheDir.empty())
348     pruneCache(config->thinLTOCacheDir, config->thinLTOCachePolicy);
349 
350   if (!config->ltoObjPath.empty()) {
351     saveBuffer(buf[0], config->ltoObjPath);
352     for (unsigned i = 1; i != maxTasks; ++i)
353       saveBuffer(buf[i], config->ltoObjPath + Twine(i));
354   }
355 
356   if (config->saveTemps) {
357     if (!buf[0].empty())
358       saveBuffer(buf[0], config->outputFile + ".lto.o");
359     for (unsigned i = 1; i != maxTasks; ++i)
360       saveBuffer(buf[i], config->outputFile + Twine(i) + ".lto.o");
361   }
362 
363   if (config->ltoEmitAsm) {
364     saveBuffer(buf[0], config->outputFile);
365     for (unsigned i = 1; i != maxTasks; ++i)
366       saveBuffer(buf[i], config->outputFile + Twine(i));
367     return {};
368   }
369 
370   std::vector<InputFile *> ret;
371   for (unsigned i = 0; i != maxTasks; ++i)
372     if (!buf[i].empty())
373       ret.push_back(createObjectFile(MemoryBufferRef(buf[i], "lto.tmp")));
374 
375   for (std::unique_ptr<MemoryBuffer> &file : files)
376     if (file)
377       ret.push_back(createObjectFile(*file));
378   return ret;
379 }
380