1 //===-- gold-plugin.cpp - Plugin to gold for Link Time Optimization  ------===//
2 //
3 //                     The LLVM Compiler Infrastructure
4 //
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
7 //
8 //===----------------------------------------------------------------------===//
9 //
10 // This is a gold plugin for LLVM. It provides an LLVM implementation of the
11 // interface described in http://gcc.gnu.org/wiki/whopr/driver .
12 //
13 //===----------------------------------------------------------------------===//
14 
15 #include "llvm/Config/config.h" // plugin-api.h requires HAVE_STDINT_H
16 #include "llvm-c/lto.h"
17 #include "llvm/ADT/StringSet.h"
18 #include "llvm/CodeGen/CommandFlags.h"
19 #include "llvm/LTO/LTOCodeGenerator.h"
20 #include "llvm/LTO/LTOModule.h"
21 #include "llvm/Support/MemoryBuffer.h"
22 #include "llvm/Support/TargetSelect.h"
23 #include <list>
24 #include <plugin-api.h>
25 #include <system_error>
26 #include <vector>
27 
28 #ifndef LDPO_PIE
29 // FIXME: remove this declaration when we stop maintaining Ubuntu Quantal and
30 // Precise and Debian Wheezy (binutils 2.23 is required)
31 # define LDPO_PIE 3
32 #endif
33 
34 using namespace llvm;
35 
36 namespace {
37 struct claimed_file {
38   void *handle;
39   std::vector<ld_plugin_symbol> syms;
40 };
41 }
42 
43 static ld_plugin_status discard_message(int level, const char *format, ...) {
44   // Die loudly. Recent versions of Gold pass ld_plugin_message as the first
45   // callback in the transfer vector. This should never be called.
46   abort();
47 }
48 
49 static ld_plugin_add_symbols add_symbols = nullptr;
50 static ld_plugin_get_symbols get_symbols = nullptr;
51 static ld_plugin_add_input_file add_input_file = nullptr;
52 static ld_plugin_set_extra_library_path set_extra_library_path = nullptr;
53 static ld_plugin_get_view get_view = nullptr;
54 static ld_plugin_message message = discard_message;
55 static lto_codegen_model output_type = LTO_CODEGEN_PIC_MODEL_STATIC;
56 static std::string output_name = "";
57 static std::list<claimed_file> Modules;
58 static std::vector<std::string> Cleanup;
59 static LTOCodeGenerator *CodeGen = nullptr;
60 static StringSet<> CannotBeHidden;
61 static llvm::TargetOptions TargetOpts;
62 
63 namespace options {
64   enum generate_bc { BC_NO, BC_ALSO, BC_ONLY };
65   static bool generate_api_file = false;
66   static generate_bc generate_bc_file = BC_NO;
67   static std::string bc_path;
68   static std::string obj_path;
69   static std::string extra_library_path;
70   static std::string triple;
71   static std::string mcpu;
72   // Additional options to pass into the code generator.
73   // Note: This array will contain all plugin options which are not claimed
74   // as plugin exclusive to pass to the code generator.
75   // For example, "generate-api-file" and "as"options are for the plugin
76   // use only and will not be passed.
77   static std::vector<const char *> extra;
78 
79   static void process_plugin_option(const char* opt_)
80   {
81     if (opt_ == nullptr)
82       return;
83     llvm::StringRef opt = opt_;
84 
85     if (opt == "generate-api-file") {
86       generate_api_file = true;
87     } else if (opt.startswith("mcpu=")) {
88       mcpu = opt.substr(strlen("mcpu="));
89     } else if (opt.startswith("extra-library-path=")) {
90       extra_library_path = opt.substr(strlen("extra_library_path="));
91     } else if (opt.startswith("mtriple=")) {
92       triple = opt.substr(strlen("mtriple="));
93     } else if (opt.startswith("obj-path=")) {
94       obj_path = opt.substr(strlen("obj-path="));
95     } else if (opt == "emit-llvm") {
96       generate_bc_file = BC_ONLY;
97     } else if (opt == "also-emit-llvm") {
98       generate_bc_file = BC_ALSO;
99     } else if (opt.startswith("also-emit-llvm=")) {
100       llvm::StringRef path = opt.substr(strlen("also-emit-llvm="));
101       generate_bc_file = BC_ALSO;
102       if (!bc_path.empty()) {
103         message(LDPL_WARNING, "Path to the output IL file specified twice. "
104                               "Discarding %s",
105                 opt_);
106       } else {
107         bc_path = path;
108       }
109     } else {
110       // Save this option to pass to the code generator.
111       extra.push_back(opt_);
112     }
113   }
114 }
115 
116 static ld_plugin_status claim_file_hook(const ld_plugin_input_file *file,
117                                         int *claimed);
118 static ld_plugin_status all_symbols_read_hook(void);
119 static ld_plugin_status cleanup_hook(void);
120 
121 extern "C" ld_plugin_status onload(ld_plugin_tv *tv);
122 ld_plugin_status onload(ld_plugin_tv *tv) {
123   InitializeAllTargetInfos();
124   InitializeAllTargets();
125   InitializeAllTargetMCs();
126   InitializeAllAsmParsers();
127   InitializeAllAsmPrinters();
128 
129   // We're given a pointer to the first transfer vector. We read through them
130   // until we find one where tv_tag == LDPT_NULL. The REGISTER_* tagged values
131   // contain pointers to functions that we need to call to register our own
132   // hooks. The others are addresses of functions we can use to call into gold
133   // for services.
134 
135   bool registeredClaimFile = false;
136   bool RegisteredAllSymbolsRead = false;
137 
138   for (; tv->tv_tag != LDPT_NULL; ++tv) {
139     switch (tv->tv_tag) {
140       case LDPT_OUTPUT_NAME:
141         output_name = tv->tv_u.tv_string;
142         break;
143       case LDPT_LINKER_OUTPUT:
144         switch (tv->tv_u.tv_val) {
145           case LDPO_REL:  // .o
146           case LDPO_DYN:  // .so
147           case LDPO_PIE:  // position independent executable
148             output_type = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
149             break;
150           case LDPO_EXEC:  // .exe
151             output_type = LTO_CODEGEN_PIC_MODEL_STATIC;
152             break;
153           default:
154             message(LDPL_ERROR, "Unknown output file type %d", tv->tv_u.tv_val);
155             return LDPS_ERR;
156         }
157         break;
158       case LDPT_OPTION:
159         options::process_plugin_option(tv->tv_u.tv_string);
160         break;
161       case LDPT_REGISTER_CLAIM_FILE_HOOK: {
162         ld_plugin_register_claim_file callback;
163         callback = tv->tv_u.tv_register_claim_file;
164 
165         if (callback(claim_file_hook) != LDPS_OK)
166           return LDPS_ERR;
167 
168         registeredClaimFile = true;
169       } break;
170       case LDPT_REGISTER_ALL_SYMBOLS_READ_HOOK: {
171         ld_plugin_register_all_symbols_read callback;
172         callback = tv->tv_u.tv_register_all_symbols_read;
173 
174         if (callback(all_symbols_read_hook) != LDPS_OK)
175           return LDPS_ERR;
176 
177         RegisteredAllSymbolsRead = true;
178       } break;
179       case LDPT_REGISTER_CLEANUP_HOOK: {
180         ld_plugin_register_cleanup callback;
181         callback = tv->tv_u.tv_register_cleanup;
182 
183         if (callback(cleanup_hook) != LDPS_OK)
184           return LDPS_ERR;
185       } break;
186       case LDPT_ADD_SYMBOLS:
187         add_symbols = tv->tv_u.tv_add_symbols;
188         break;
189       case LDPT_GET_SYMBOLS_V2:
190         get_symbols = tv->tv_u.tv_get_symbols;
191         break;
192       case LDPT_ADD_INPUT_FILE:
193         add_input_file = tv->tv_u.tv_add_input_file;
194         break;
195       case LDPT_SET_EXTRA_LIBRARY_PATH:
196         set_extra_library_path = tv->tv_u.tv_set_extra_library_path;
197         break;
198       case LDPT_GET_VIEW:
199         get_view = tv->tv_u.tv_get_view;
200         break;
201       case LDPT_MESSAGE:
202         message = tv->tv_u.tv_message;
203         break;
204       default:
205         break;
206     }
207   }
208 
209   if (!registeredClaimFile) {
210     message(LDPL_ERROR, "register_claim_file not passed to LLVMgold.");
211     return LDPS_ERR;
212   }
213   if (!add_symbols) {
214     message(LDPL_ERROR, "add_symbols not passed to LLVMgold.");
215     return LDPS_ERR;
216   }
217 
218   if (!RegisteredAllSymbolsRead)
219     return LDPS_OK;
220 
221   CodeGen = new LTOCodeGenerator();
222 
223   // Pass through extra options to the code generator.
224   if (!options::extra.empty()) {
225     for (const char *Opt : options::extra)
226       CodeGen->setCodeGenDebugOptions(Opt);
227   }
228 
229   CodeGen->parseCodeGenDebugOptions();
230   if (MAttrs.size()) {
231     std::string Attrs;
232     for (unsigned I = 0; I < MAttrs.size(); ++I) {
233       if (I > 0)
234         Attrs.append(",");
235       Attrs.append(MAttrs[I]);
236     }
237     CodeGen->setAttr(Attrs.c_str());
238   }
239 
240   TargetOpts = InitTargetOptionsFromCodeGenFlags();
241   CodeGen->setTargetOptions(TargetOpts);
242 
243   return LDPS_OK;
244 }
245 
246 /// Called by gold to see whether this file is one that our plugin can handle.
247 /// We'll try to open it and register all the symbols with add_symbol if
248 /// possible.
249 static ld_plugin_status claim_file_hook(const ld_plugin_input_file *file,
250                                         int *claimed) {
251   const void *view;
252   std::unique_ptr<MemoryBuffer> buffer;
253   if (get_view) {
254     if (get_view(file->handle, &view) != LDPS_OK) {
255       message(LDPL_ERROR, "Failed to get a view of %s", file->name);
256       return LDPS_ERR;
257     }
258   } else {
259     int64_t offset = 0;
260     // Gold has found what might be IR part-way inside of a file, such as
261     // an .a archive.
262     if (file->offset) {
263       offset = file->offset;
264     }
265     ErrorOr<std::unique_ptr<MemoryBuffer>> BufferOrErr =
266         MemoryBuffer::getOpenFileSlice(file->fd, file->name, file->filesize,
267                                        offset);
268     if (std::error_code EC = BufferOrErr.getError()) {
269       message(LDPL_ERROR, EC.message().c_str());
270       return LDPS_ERR;
271     }
272     buffer = std::move(BufferOrErr.get());
273     view = buffer->getBufferStart();
274   }
275 
276   if (!LTOModule::isBitcodeFile(view, file->filesize))
277     return LDPS_OK;
278 
279   *claimed = 1;
280 
281   std::string Error;
282   LTOModule *M =
283       LTOModule::createFromBuffer(view, file->filesize, TargetOpts, Error);
284   if (!M) {
285     message(LDPL_ERROR, "LLVM gold plugin has failed to create LTO module: %s",
286             Error.c_str());
287     return LDPS_ERR;
288   }
289 
290   Modules.resize(Modules.size() + 1);
291   claimed_file &cf = Modules.back();
292 
293   if (!options::triple.empty())
294     M->setTargetTriple(options::triple.c_str());
295 
296   cf.handle = file->handle;
297   unsigned sym_count = M->getSymbolCount();
298   cf.syms.reserve(sym_count);
299 
300   for (unsigned i = 0; i != sym_count; ++i) {
301     lto_symbol_attributes attrs = M->getSymbolAttributes(i);
302     if ((attrs & LTO_SYMBOL_SCOPE_MASK) == LTO_SYMBOL_SCOPE_INTERNAL)
303       continue;
304 
305     cf.syms.push_back(ld_plugin_symbol());
306     ld_plugin_symbol &sym = cf.syms.back();
307     sym.name = strdup(M->getSymbolName(i));
308     sym.version = nullptr;
309 
310     int scope = attrs & LTO_SYMBOL_SCOPE_MASK;
311     bool CanBeHidden = scope == LTO_SYMBOL_SCOPE_DEFAULT_CAN_BE_HIDDEN;
312     if (!CanBeHidden)
313       CannotBeHidden.insert(sym.name);
314     switch (scope) {
315       case LTO_SYMBOL_SCOPE_HIDDEN:
316         sym.visibility = LDPV_HIDDEN;
317         break;
318       case LTO_SYMBOL_SCOPE_PROTECTED:
319         sym.visibility = LDPV_PROTECTED;
320         break;
321       case 0: // extern
322       case LTO_SYMBOL_SCOPE_DEFAULT:
323       case LTO_SYMBOL_SCOPE_DEFAULT_CAN_BE_HIDDEN:
324         sym.visibility = LDPV_DEFAULT;
325         break;
326       default:
327         message(LDPL_ERROR, "Unknown scope attribute: %d", scope);
328         return LDPS_ERR;
329     }
330 
331     int definition = attrs & LTO_SYMBOL_DEFINITION_MASK;
332     sym.comdat_key = nullptr;
333     switch (definition) {
334       case LTO_SYMBOL_DEFINITION_REGULAR:
335         sym.def = LDPK_DEF;
336         break;
337       case LTO_SYMBOL_DEFINITION_UNDEFINED:
338         sym.def = LDPK_UNDEF;
339         break;
340       case LTO_SYMBOL_DEFINITION_TENTATIVE:
341         sym.def = LDPK_COMMON;
342         break;
343       case LTO_SYMBOL_DEFINITION_WEAK:
344         sym.comdat_key = sym.name;
345         sym.def = LDPK_WEAKDEF;
346         break;
347       case LTO_SYMBOL_DEFINITION_WEAKUNDEF:
348         sym.def = LDPK_WEAKUNDEF;
349         break;
350       default:
351         message(LDPL_ERROR, "Unknown definition attribute: %d", definition);
352         return LDPS_ERR;
353     }
354 
355     sym.size = 0;
356 
357     sym.resolution = LDPR_UNKNOWN;
358   }
359 
360   cf.syms.reserve(cf.syms.size());
361 
362   if (!cf.syms.empty()) {
363     if (add_symbols(cf.handle, cf.syms.size(), &cf.syms[0]) != LDPS_OK) {
364       message(LDPL_ERROR, "Unable to add symbols!");
365       return LDPS_ERR;
366     }
367   }
368 
369   if (CodeGen) {
370     std::string Error;
371     if (!CodeGen->addModule(M, Error)) {
372       message(LDPL_ERROR, "Error linking module: %s", Error.c_str());
373       return LDPS_ERR;
374     }
375   }
376 
377   delete M;
378 
379   return LDPS_OK;
380 }
381 
382 static bool mustPreserve(ld_plugin_symbol &Sym) {
383   if (Sym.resolution == LDPR_PREVAILING_DEF)
384     return true;
385   if (Sym.resolution == LDPR_PREVAILING_DEF_IRONLY_EXP)
386     return CannotBeHidden.count(Sym.name);
387   return false;
388 }
389 
390 /// gold informs us that all symbols have been read. At this point, we use
391 /// get_symbols to see if any of our definitions have been overridden by a
392 /// native object file. Then, perform optimization and codegen.
393 static ld_plugin_status allSymbolsReadHook(raw_fd_ostream *apiFile) {
394   assert(CodeGen);
395 
396   for (claimed_file &F : Modules) {
397     if (F.syms.empty())
398       continue;
399     get_symbols(F.handle, F.syms.size(), &F.syms[0]);
400     for (ld_plugin_symbol &Sym : F.syms) {
401       if (mustPreserve(Sym)) {
402         CodeGen->addMustPreserveSymbol(Sym.name);
403 
404         if (options::generate_api_file)
405           (*apiFile) << Sym.name << "\n";
406       }
407     }
408   }
409 
410   CodeGen->setCodePICModel(output_type);
411   CodeGen->setDebugInfo(LTO_DEBUG_MODEL_DWARF);
412   if (!options::mcpu.empty())
413     CodeGen->setCpu(options::mcpu.c_str());
414 
415   if (options::generate_bc_file != options::BC_NO) {
416     std::string path;
417     if (options::generate_bc_file == options::BC_ONLY)
418       path = output_name;
419     else if (!options::bc_path.empty())
420       path = options::bc_path;
421     else
422       path = output_name + ".bc";
423     std::string Error;
424     if (!CodeGen->writeMergedModules(path.c_str(), Error))
425       message(LDPL_FATAL, "Failed to write the output file.");
426     if (options::generate_bc_file == options::BC_ONLY)
427       return LDPS_OK;
428   }
429 
430   std::string ObjPath;
431   {
432     const char *Temp;
433     std::string Error;
434     if (!CodeGen->compile_to_file(&Temp, /*DisableOpt*/ false, /*DisableInline*/
435                                   false, /*DisableGVNLoadPRE*/ false, Error))
436       message(LDPL_ERROR, "Could not produce a combined object file\n");
437     ObjPath = Temp;
438   }
439 
440   for (claimed_file &F : Modules) {
441     for (ld_plugin_symbol &Sym : F.syms)
442       free(Sym.name);
443   }
444 
445   if (add_input_file(ObjPath.c_str()) != LDPS_OK) {
446     message(LDPL_ERROR, "Unable to add .o file to the link.");
447     message(LDPL_ERROR, "File left behind in: %s", ObjPath.c_str());
448     return LDPS_ERR;
449   }
450 
451   if (!options::extra_library_path.empty() &&
452       set_extra_library_path(options::extra_library_path.c_str()) != LDPS_OK) {
453     message(LDPL_ERROR, "Unable to set the extra library path.");
454     return LDPS_ERR;
455   }
456 
457   if (options::obj_path.empty())
458     Cleanup.push_back(ObjPath);
459 
460   return LDPS_OK;
461 }
462 
463 static ld_plugin_status all_symbols_read_hook(void) {
464   ld_plugin_status Ret;
465   if (!options::generate_api_file) {
466     Ret = allSymbolsReadHook(nullptr);
467   } else {
468     std::string Error;
469     raw_fd_ostream apiFile("apifile.txt", Error, sys::fs::F_None);
470     if (!Error.empty())
471       message(LDPL_FATAL, "Unable to open apifile.txt for writing: %s",
472               Error.c_str());
473     Ret = allSymbolsReadHook(&apiFile);
474   }
475 
476   delete CodeGen;
477 
478   if (options::generate_bc_file == options::BC_ONLY)
479     exit(0);
480 
481   return Ret;
482 }
483 
484 static ld_plugin_status cleanup_hook(void) {
485   for (std::string &Name : Cleanup) {
486     std::error_code EC = sys::fs::remove(Name);
487     if (EC)
488       message(LDPL_ERROR, "Failed to delete '%s': %s", Name.c_str(),
489               EC.message().c_str());
490   }
491 
492   return LDPS_OK;
493 }
494