1 //===-- gold-plugin.cpp - Plugin to gold for Link Time Optimization ------===//
3 // The LLVM Compiler Infrastructure
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
8 //===----------------------------------------------------------------------===//
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 .
13 //===----------------------------------------------------------------------===//
15 #include "llvm/Config/config.h"
16 #include "plugin-api.h"
18 #include "llvm-c/lto.h"
20 #include "llvm/Support/ToolOutputFile.h"
21 #include "llvm/Support/Errno.h"
22 #include "llvm/Support/Path.h"
23 #include "llvm/Support/Program.h"
32 // Support Windows/MinGW crazyness.
42 ld_plugin_status discard_message(int level, const char *format, ...) {
43 // Die loudly. Recent versions of Gold pass ld_plugin_message as the first
44 // callback in the transfer vector. This should never be called.
48 ld_plugin_add_symbols add_symbols = NULL;
49 ld_plugin_get_symbols get_symbols = NULL;
50 ld_plugin_add_input_file add_input_file = NULL;
51 ld_plugin_add_input_library add_input_library = NULL;
52 ld_plugin_set_extra_library_path set_extra_library_path = NULL;
53 ld_plugin_message message = discard_message;
61 std::vector<ld_plugin_symbol> syms;
64 lto_codegen_model output_type = LTO_CODEGEN_PIC_MODEL_STATIC;
65 std::string output_name = "";
66 std::list<claimed_file> Modules;
67 std::vector<sys::Path> Cleanup;
71 enum generate_bc { BC_NO, BC_ALSO, BC_ONLY };
72 static bool generate_api_file = false;
73 static generate_bc generate_bc_file = BC_NO;
74 static std::string bc_path;
75 static std::string as_path;
76 static std::vector<std::string> as_args;
77 static std::vector<std::string> pass_through;
78 static std::string extra_library_path;
79 static std::string triple;
80 static std::string mcpu;
81 // Additional options to pass into the code generator.
82 // Note: This array will contain all plugin options which are not claimed
83 // as plugin exclusive to pass to the code generator.
84 // For example, "generate-api-file" and "as"options are for the plugin
85 // use only and will not be passed.
86 static std::vector<std::string> extra;
88 static void process_plugin_option(const char* opt_)
92 llvm::StringRef opt = opt_;
94 if (opt == "generate-api-file") {
95 generate_api_file = true;
96 } else if (opt.startswith("mcpu=")) {
97 mcpu = opt.substr(strlen("mcpu="));
98 } else if (opt.startswith("as=")) {
99 if (!as_path.empty()) {
100 (*message)(LDPL_WARNING, "Path to as specified twice. "
101 "Discarding %s", opt_);
103 as_path = opt.substr(strlen("as="));
105 } else if (opt.startswith("as-arg=")) {
106 llvm::StringRef item = opt.substr(strlen("as-arg="));
107 as_args.push_back(item.str());
108 } else if (opt.startswith("extra-library-path=")) {
109 extra_library_path = opt.substr(strlen("extra_library_path="));
110 } else if (opt.startswith("pass-through=")) {
111 llvm::StringRef item = opt.substr(strlen("pass-through="));
112 pass_through.push_back(item.str());
113 } else if (opt.startswith("mtriple=")) {
114 triple = opt.substr(strlen("mtriple="));
115 } else if (opt == "emit-llvm") {
116 generate_bc_file = BC_ONLY;
117 } else if (opt == "also-emit-llvm") {
118 generate_bc_file = BC_ALSO;
119 } else if (opt.startswith("also-emit-llvm=")) {
120 llvm::StringRef path = opt.substr(strlen("also-emit-llvm="));
121 generate_bc_file = BC_ALSO;
122 if (!bc_path.empty()) {
123 (*message)(LDPL_WARNING, "Path to the output IL file specified twice. "
124 "Discarding %s", opt_);
129 // Save this option to pass to the code generator.
130 extra.push_back(opt);
135 static ld_plugin_status claim_file_hook(const ld_plugin_input_file *file,
137 static ld_plugin_status all_symbols_read_hook(void);
138 static ld_plugin_status cleanup_hook(void);
140 extern "C" ld_plugin_status onload(ld_plugin_tv *tv);
141 ld_plugin_status onload(ld_plugin_tv *tv) {
142 // We're given a pointer to the first transfer vector. We read through them
143 // until we find one where tv_tag == LDPT_NULL. The REGISTER_* tagged values
144 // contain pointers to functions that we need to call to register our own
145 // hooks. The others are addresses of functions we can use to call into gold
148 bool registeredClaimFile = false;
150 for (; tv->tv_tag != LDPT_NULL; ++tv) {
151 switch (tv->tv_tag) {
152 case LDPT_API_VERSION:
153 api_version = tv->tv_u.tv_val;
155 case LDPT_GOLD_VERSION: // major * 100 + minor
156 gold_version = tv->tv_u.tv_val;
158 case LDPT_OUTPUT_NAME:
159 output_name = tv->tv_u.tv_string;
161 case LDPT_LINKER_OUTPUT:
162 switch (tv->tv_u.tv_val) {
164 case LDPO_DYN: // .so
165 output_type = LTO_CODEGEN_PIC_MODEL_DYNAMIC;
167 case LDPO_EXEC: // .exe
168 output_type = LTO_CODEGEN_PIC_MODEL_STATIC;
171 (*message)(LDPL_ERROR, "Unknown output file type %d",
175 // TODO: add an option to disable PIC.
176 //output_type = LTO_CODEGEN_PIC_MODEL_DYNAMIC_NO_PIC;
179 options::process_plugin_option(tv->tv_u.tv_string);
181 case LDPT_REGISTER_CLAIM_FILE_HOOK: {
182 ld_plugin_register_claim_file callback;
183 callback = tv->tv_u.tv_register_claim_file;
185 if ((*callback)(claim_file_hook) != LDPS_OK)
188 registeredClaimFile = true;
190 case LDPT_REGISTER_ALL_SYMBOLS_READ_HOOK: {
191 ld_plugin_register_all_symbols_read callback;
192 callback = tv->tv_u.tv_register_all_symbols_read;
194 if ((*callback)(all_symbols_read_hook) != LDPS_OK)
197 case LDPT_REGISTER_CLEANUP_HOOK: {
198 ld_plugin_register_cleanup callback;
199 callback = tv->tv_u.tv_register_cleanup;
201 if ((*callback)(cleanup_hook) != LDPS_OK)
204 case LDPT_ADD_SYMBOLS:
205 add_symbols = tv->tv_u.tv_add_symbols;
207 case LDPT_GET_SYMBOLS:
208 get_symbols = tv->tv_u.tv_get_symbols;
210 case LDPT_ADD_INPUT_FILE:
211 add_input_file = tv->tv_u.tv_add_input_file;
213 case LDPT_ADD_INPUT_LIBRARY:
214 add_input_library = tv->tv_u.tv_add_input_file;
216 case LDPT_SET_EXTRA_LIBRARY_PATH:
217 set_extra_library_path = tv->tv_u.tv_set_extra_library_path;
220 message = tv->tv_u.tv_message;
227 if (!registeredClaimFile) {
228 (*message)(LDPL_ERROR, "register_claim_file not passed to LLVMgold.");
232 (*message)(LDPL_ERROR, "add_symbols not passed to LLVMgold.");
239 /// claim_file_hook - called by gold to see whether this file is one that
240 /// our plugin can handle. We'll try to open it and register all the symbols
241 /// with add_symbol if possible.
242 static ld_plugin_status claim_file_hook(const ld_plugin_input_file *file,
247 // Gold has found what might be IR part-way inside of a file, such as
249 if (lseek(file->fd, file->offset, SEEK_SET) == -1) {
250 (*message)(LDPL_ERROR,
251 "Failed to seek to archive member of %s at offset %d: %s\n",
253 file->offset, sys::StrError(errno).c_str());
256 void *buf = malloc(file->filesize);
258 (*message)(LDPL_ERROR,
259 "Failed to allocate buffer for archive member of size: %d\n",
263 if (read(file->fd, buf, file->filesize) != file->filesize) {
264 (*message)(LDPL_ERROR,
265 "Failed to read archive member of %s at offset %d: %s\n",
268 sys::StrError(errno).c_str());
272 if (!lto_module_is_object_file_in_memory(buf, file->filesize)) {
276 M = lto_module_create_from_memory(buf, file->filesize);
279 // FIXME: We should not need to pass -1 as the file size, but there
280 // is a bug in BFD that causes it to pass 0 to us. Remove this once
282 off_t size = file->filesize ? file->filesize : -1;
284 // FIXME: We should not need to reset the position in the file, but there
285 // is a bug in BFD. Remove this once that is fixed.
286 off_t old_pos = lseek(file->fd, 0, SEEK_CUR);
288 lseek(file->fd, 0, SEEK_SET);
289 M = lto_module_create_from_fd(file->fd, file->name, size);
291 lseek(file->fd, old_pos, SEEK_SET);
297 Modules.resize(Modules.size() + 1);
298 claimed_file &cf = Modules.back();
302 (*message)(LDPL_ERROR, "Failed to create LLVM module: %s",
303 lto_get_error_message());
307 if (!options::triple.empty())
308 lto_module_set_target_triple(cf.M, options::triple.c_str());
310 cf.handle = file->handle;
311 unsigned sym_count = lto_module_get_num_symbols(cf.M);
312 cf.syms.reserve(sym_count);
314 for (unsigned i = 0; i != sym_count; ++i) {
315 lto_symbol_attributes attrs = lto_module_get_symbol_attribute(cf.M, i);
316 if ((attrs & LTO_SYMBOL_SCOPE_MASK) == LTO_SYMBOL_SCOPE_INTERNAL)
319 cf.syms.push_back(ld_plugin_symbol());
320 ld_plugin_symbol &sym = cf.syms.back();
321 sym.name = const_cast<char *>(lto_module_get_symbol_name(cf.M, i));
324 int scope = attrs & LTO_SYMBOL_SCOPE_MASK;
326 case LTO_SYMBOL_SCOPE_HIDDEN:
327 sym.visibility = LDPV_HIDDEN;
329 case LTO_SYMBOL_SCOPE_PROTECTED:
330 sym.visibility = LDPV_PROTECTED;
333 case LTO_SYMBOL_SCOPE_DEFAULT:
334 sym.visibility = LDPV_DEFAULT;
337 (*message)(LDPL_ERROR, "Unknown scope attribute: %d", scope);
341 int definition = attrs & LTO_SYMBOL_DEFINITION_MASK;
342 sym.comdat_key = NULL;
343 switch (definition) {
344 case LTO_SYMBOL_DEFINITION_REGULAR:
347 case LTO_SYMBOL_DEFINITION_UNDEFINED:
348 sym.def = LDPK_UNDEF;
350 case LTO_SYMBOL_DEFINITION_TENTATIVE:
351 sym.def = LDPK_COMMON;
353 case LTO_SYMBOL_DEFINITION_WEAK:
354 sym.comdat_key = sym.name;
355 sym.def = LDPK_WEAKDEF;
357 case LTO_SYMBOL_DEFINITION_WEAKUNDEF:
358 sym.def = LDPK_WEAKUNDEF;
361 (*message)(LDPL_ERROR, "Unknown definition attribute: %d", definition);
367 sym.resolution = LDPR_UNKNOWN;
370 cf.syms.reserve(cf.syms.size());
372 if (!cf.syms.empty()) {
373 if ((*add_symbols)(cf.handle, cf.syms.size(), &cf.syms[0]) != LDPS_OK) {
374 (*message)(LDPL_ERROR, "Unable to add symbols!");
382 /// all_symbols_read_hook - gold informs us that all symbols have been read.
383 /// At this point, we use get_symbols to see if any of our definitions have
384 /// been overridden by a native object file. Then, perform optimization and
386 static ld_plugin_status all_symbols_read_hook(void) {
387 lto_code_gen_t cg = lto_codegen_create();
389 for (std::list<claimed_file>::iterator I = Modules.begin(),
390 E = Modules.end(); I != E; ++I)
391 lto_codegen_add_module(cg, I->M);
393 std::ofstream api_file;
394 if (options::generate_api_file) {
395 api_file.open("apifile.txt", std::ofstream::out | std::ofstream::trunc);
396 if (!api_file.is_open()) {
397 (*message)(LDPL_FATAL, "Unable to open apifile.txt for writing.");
402 // If we don't preserve any symbols, libLTO will assume that all symbols are
403 // needed. Keep all symbols unless we're producing a final executable.
404 bool anySymbolsPreserved = false;
405 for (std::list<claimed_file>::iterator I = Modules.begin(),
406 E = Modules.end(); I != E; ++I) {
407 (*get_symbols)(I->handle, I->syms.size(), &I->syms[0]);
408 for (unsigned i = 0, e = I->syms.size(); i != e; i++) {
409 if (I->syms[i].resolution == LDPR_PREVAILING_DEF) {
410 lto_codegen_add_must_preserve_symbol(cg, I->syms[i].name);
411 anySymbolsPreserved = true;
413 if (options::generate_api_file)
414 api_file << I->syms[i].name << "\n";
419 if (options::generate_api_file)
422 if (!anySymbolsPreserved) {
423 // All of the IL is unnecessary!
424 lto_codegen_dispose(cg);
428 lto_codegen_set_pic_model(cg, output_type);
429 lto_codegen_set_debug_model(cg, LTO_DEBUG_MODEL_DWARF);
430 if (!options::as_path.empty()) {
431 sys::Path p = sys::Program::FindProgramByName(options::as_path);
432 lto_codegen_set_assembler_path(cg, p.c_str());
434 if (!options::as_args.empty()) {
435 std::vector<const char *> as_args_p;
436 for (std::vector<std::string>::iterator I = options::as_args.begin(),
437 E = options::as_args.end(); I != E; ++I) {
438 as_args_p.push_back(I->c_str());
440 lto_codegen_set_assembler_args(cg, &as_args_p[0], as_args_p.size());
442 if (!options::mcpu.empty())
443 lto_codegen_set_cpu(cg, options::mcpu.c_str());
445 // Pass through extra options to the code generator.
446 if (!options::extra.empty()) {
447 for (std::vector<std::string>::iterator it = options::extra.begin();
448 it != options::extra.end(); ++it) {
449 lto_codegen_debug_options(cg, (*it).c_str());
454 if (options::generate_bc_file != options::BC_NO) {
456 if (options::generate_bc_file == options::BC_ONLY)
458 else if (!options::bc_path.empty())
459 path = options::bc_path;
461 path = output_name + ".bc";
462 bool err = lto_codegen_write_merged_modules(cg, path.c_str());
464 (*message)(LDPL_FATAL, "Failed to write the output file.");
465 if (options::generate_bc_file == options::BC_ONLY)
469 const char *buffer = static_cast<const char *>(lto_codegen_compile(cg,
474 sys::Path uniqueObjPath("/tmp/llvmgold.o");
475 if (uniqueObjPath.createTemporaryFileOnDisk(true, &ErrMsg)) {
476 (*message)(LDPL_ERROR, "%s", ErrMsg.c_str());
479 tool_output_file objFile(uniqueObjPath.c_str(), ErrMsg,
480 raw_fd_ostream::F_Binary);
481 if (!ErrMsg.empty()) {
482 (*message)(LDPL_ERROR, "%s", ErrMsg.c_str());
486 objFile.os().write(buffer, bufsize);
487 objFile.os().close();
488 if (objFile.os().has_error()) {
489 (*message)(LDPL_ERROR, "Error writing output file '%s'",
490 uniqueObjPath.c_str());
491 objFile.os().clear_error();
496 lto_codegen_dispose(cg);
498 if ((*add_input_file)(uniqueObjPath.c_str()) != LDPS_OK) {
499 (*message)(LDPL_ERROR, "Unable to add .o file to the link.");
500 (*message)(LDPL_ERROR, "File left behind in: %s", uniqueObjPath.c_str());
504 if (!options::extra_library_path.empty() &&
505 set_extra_library_path(options::extra_library_path.c_str()) != LDPS_OK) {
506 (*message)(LDPL_ERROR, "Unable to set the extra library path.");
510 for (std::vector<std::string>::iterator i = options::pass_through.begin(),
511 e = options::pass_through.end();
513 std::string &item = *i;
514 const char *item_p = item.c_str();
515 if (llvm::StringRef(item).startswith("-l")) {
516 if (add_input_library(item_p + 2) != LDPS_OK) {
517 (*message)(LDPL_ERROR, "Unable to add library to the link.");
521 if (add_input_file(item_p) != LDPS_OK) {
522 (*message)(LDPL_ERROR, "Unable to add .o file to the link.");
528 Cleanup.push_back(uniqueObjPath);
533 static ld_plugin_status cleanup_hook(void) {
536 for (int i = 0, e = Cleanup.size(); i != e; ++i)
537 if (Cleanup[i].eraseFromDisk(false, &ErrMsg))
538 (*message)(LDPL_ERROR, "Failed to delete '%s': %s", Cleanup[i].c_str(),