/* Linker command language support. Copyright (C) 1991-2023 Free Software Foundation, Inc. This file is part of the GNU Binutils. This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 3 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin Street - Fifth Floor, Boston, MA 02110-1301, USA. */ #include "sysdep.h" #include #include "bfd.h" #include "libiberty.h" #include "filenames.h" #include "safe-ctype.h" #include "obstack.h" #include "bfdlink.h" #include "ctf-api.h" #include "ld.h" #include "ldmain.h" #include "ldexp.h" #include "ldlang.h" #include #include "ldlex.h" #include "ldmisc.h" #include "ldctor.h" #include "ldfile.h" #include "ldemul.h" #include "fnmatch.h" #include "demangle.h" #include "hashtab.h" #include "elf-bfd.h" #if BFD_SUPPORTS_PLUGINS #include "plugin.h" #endif /* BFD_SUPPORTS_PLUGINS */ #ifndef offsetof #define offsetof(TYPE, MEMBER) ((size_t) & (((TYPE*) 0)->MEMBER)) #endif /* Convert between addresses in bytes and sizes in octets. For currently supported targets, octets_per_byte is always a power of two, so we can use shifts. */ #define TO_ADDR(X) ((X) >> opb_shift) #define TO_SIZE(X) ((X) << opb_shift) /* Local variables. */ static struct obstack stat_obstack; static struct obstack map_obstack; static struct obstack pt_obstack; #define obstack_chunk_alloc xmalloc #define obstack_chunk_free free static const char *entry_symbol_default = "start"; static bool map_head_is_link_order = false; static lang_output_section_statement_type *default_common_section; static bool map_option_f; static bfd_vma print_dot; static lang_input_statement_type *first_file; static const char *current_target; static lang_statement_list_type *stat_save[10]; static lang_statement_list_type **stat_save_ptr = &stat_save[0]; static struct unique_sections *unique_section_list; static struct asneeded_minfo *asneeded_list_head; static unsigned int opb_shift = 0; /* Forward declarations. */ static void exp_init_os (etree_type *); static lang_input_statement_type *lookup_name (const char *); static void insert_undefined (const char *); static bool sort_def_symbol (struct bfd_link_hash_entry *, void *); static lang_statement_union_type *new_statement (enum statement_enum type, size_t size, lang_statement_list_type *list); static void print_statement (lang_statement_union_type *, lang_output_section_statement_type *); static void print_statement_list (lang_statement_union_type *, lang_output_section_statement_type *); static void print_statements (void); static void print_input_section (asection *, bool); static bool lang_one_common (struct bfd_link_hash_entry *, void *); static void lang_record_phdrs (void); static void lang_do_version_exports_section (void); static void lang_finalize_version_expr_head (struct bfd_elf_version_expr_head *); static void lang_do_memory_regions (bool); /* Exported variables. */ const char *output_target; lang_output_section_statement_type *abs_output_section; /* Header for list of statements corresponding to any files involved in the link, either specified from the command-line or added implicitely (eg. archive member used to resolved undefined symbol, wildcard statement from linker script, etc.). Next pointer is in next field of a lang_statement_header_type (reached via header field in a lang_statement_union). */ lang_statement_list_type statement_list; lang_statement_list_type lang_os_list; lang_statement_list_type *stat_ptr = &statement_list; /* Header for list of statements corresponding to files used in the final executable. This can be either object file specified on the command-line or library member resolving an undefined reference. Next pointer is in next field of a lang_input_statement_type (reached via input_statement field in a lang_statement_union). */ lang_statement_list_type file_chain = { NULL, NULL }; /* Header for list of statements corresponding to files specified on the command-line for linking. It thus contains real object files and archive but not archive members. Next pointer is in next_real_file field of a lang_input_statement_type statement (reached via input_statement field in a lang_statement_union). */ lang_statement_list_type input_file_chain; static const char *current_input_file; struct bfd_elf_dynamic_list **current_dynamic_list_p; struct bfd_sym_chain entry_symbol = { NULL, NULL }; const char *entry_section = ".text"; struct lang_input_statement_flags input_flags; bool entry_from_cmdline; bool lang_has_input_file = false; bool had_output_filename = false; bool lang_float_flag = false; bool delete_output_file_on_failure = false; struct lang_phdr *lang_phdr_list; struct lang_nocrossrefs *nocrossref_list; struct asneeded_minfo **asneeded_list_tail; #ifdef ENABLE_LIBCTF static ctf_dict_t *ctf_output; #endif /* Functions that traverse the linker script and might evaluate DEFINED() need to increment this at the start of the traversal. */ int lang_statement_iteration = 0; /* Count times through one_lang_size_sections_pass after mark phase. */ static int lang_sizing_iteration = 0; /* Return TRUE if the PATTERN argument is a wildcard pattern. Although backslashes are treated specially if a pattern contains wildcards, we do not consider the mere presence of a backslash to be enough to cause the pattern to be treated as a wildcard. That lets us handle DOS filenames more naturally. */ #define wildcardp(pattern) (strpbrk ((pattern), "?*[") != NULL) #define new_stat(x, y) \ (x##_type *) new_statement (x##_enum, sizeof (x##_type), y) #define outside_section_address(q) \ ((q)->output_offset + (q)->output_section->vma) #define outside_symbol_address(q) \ ((q)->value + outside_section_address (q->section)) /* CTF sections smaller than this are not compressed: compression of dictionaries this small doesn't gain much, and this lets consumers mmap the sections directly out of the ELF file and use them with no decompression overhead if they want to. */ #define CTF_COMPRESSION_THRESHOLD 4096 void * stat_alloc (size_t size) { return obstack_alloc (&stat_obstack, size); } /* Code for handling simple wildcards without going through fnmatch, which can be expensive because of charset translations etc. */ /* A simple wild is a literal string followed by a single '*', where the literal part is at least 4 characters long. */ static bool is_simple_wild (const char *name) { size_t len = strcspn (name, "*?["); return len >= 4 && name[len] == '*' && name[len + 1] == '\0'; } static bool match_simple_wild (const char *pattern, const char *name) { /* The first four characters of the pattern are guaranteed valid non-wildcard characters. So we can go faster. */ if (pattern[0] != name[0] || pattern[1] != name[1] || pattern[2] != name[2] || pattern[3] != name[3]) return false; pattern += 4; name += 4; while (*pattern != '*') if (*name++ != *pattern++) return false; return true; } static int name_match (const char *pattern, const char *name) { if (is_simple_wild (pattern)) return !match_simple_wild (pattern, name); if (wildcardp (pattern)) return fnmatch (pattern, name, 0); return strcmp (pattern, name); } /* Given an analyzed wildcard_spec SPEC, match it against NAME, returns zero on a match, non-zero if there's no match. */ static int spec_match (const struct wildcard_spec *spec, const char *name) { size_t nl = spec->namelen; size_t pl = spec->prefixlen; size_t sl = spec->suffixlen; size_t inputlen = strlen (name); int r; if (pl) { if (inputlen < pl) return 1; r = memcmp (spec->name, name, pl); if (r) return r; } if (sl) { if (inputlen < sl) return 1; r = memcmp (spec->name + nl - sl, name + inputlen - sl, sl); if (r) return r; } if (nl == pl + sl + 1 && spec->name[pl] == '*') return 0; if (nl > pl) return fnmatch (spec->name + pl, name + pl, 0); if (inputlen >= nl) return name[nl]; return 0; } static char * ldirname (const char *name) { const char *base = lbasename (name); char *dirname; while (base > name && IS_DIR_SEPARATOR (base[-1])) --base; if (base == name) return strdup ("."); dirname = strdup (name); dirname[base - name] = '\0'; return dirname; } /* If PATTERN is of the form archive:file, return a pointer to the separator. If not, return NULL. */ static char * archive_path (const char *pattern) { char *p = NULL; if (link_info.path_separator == 0) return p; p = strchr (pattern, link_info.path_separator); #ifdef HAVE_DOS_BASED_FILE_SYSTEM if (p == NULL || link_info.path_separator != ':') return p; /* Assume a match on the second char is part of drive specifier, as in "c:\silly.dos". */ if (p == pattern + 1 && ISALPHA (*pattern)) p = strchr (p + 1, link_info.path_separator); #endif return p; } /* Given that FILE_SPEC results in a non-NULL SEP result from archive_path, return whether F matches FILE_SPEC. */ static bool input_statement_is_archive_path (const char *file_spec, char *sep, lang_input_statement_type *f) { bool match = false; if ((*(sep + 1) == 0 || name_match (sep + 1, f->filename) == 0) && ((sep != file_spec) == (f->the_bfd != NULL && f->the_bfd->my_archive != NULL))) { match = true; if (sep != file_spec) { const char *aname = bfd_get_filename (f->the_bfd->my_archive); *sep = 0; match = name_match (file_spec, aname) == 0; *sep = link_info.path_separator; } } return match; } static bool unique_section_p (const asection *sec, const lang_output_section_statement_type *os) { struct unique_sections *unam; const char *secnam; if (!link_info.resolve_section_groups && sec->owner != NULL && bfd_is_group_section (sec->owner, sec)) return !(os != NULL && strcmp (os->name, DISCARD_SECTION_NAME) == 0); secnam = sec->name; for (unam = unique_section_list; unam; unam = unam->next) if (name_match (unam->name, secnam) == 0) return true; return false; } /* Generic traversal routines for finding matching sections. */ /* Return true if FILE matches a pattern in EXCLUDE_LIST, otherwise return false. */ static bool walk_wild_file_in_exclude_list (struct name_list *exclude_list, lang_input_statement_type *file) { struct name_list *list_tmp; for (list_tmp = exclude_list; list_tmp; list_tmp = list_tmp->next) { char *p = archive_path (list_tmp->name); if (p != NULL) { if (input_statement_is_archive_path (list_tmp->name, p, file)) return true; } else if (name_match (list_tmp->name, file->filename) == 0) return true; /* FIXME: Perhaps remove the following at some stage? Matching unadorned archives like this was never documented and has been superceded by the archive:path syntax. */ else if (file->the_bfd != NULL && file->the_bfd->my_archive != NULL && name_match (list_tmp->name, bfd_get_filename (file->the_bfd->my_archive)) == 0) return true; } return false; } /* Add SECTION (from input FILE) to the list of matching sections within PTR (the matching wildcard is SEC). */ static void add_matching_section (lang_wild_statement_type *ptr, struct wildcard_list *sec, asection *section, lang_input_statement_type *file) { lang_input_matcher_type *new_section; /* Add a section reference to the list. */ new_section = new_stat (lang_input_matcher, &ptr->matching_sections); new_section->section = section; new_section->pattern = sec; new_section->input_stmt = file; } /* Process section S (from input file FILE) in relation to wildcard statement PTR. We already know that a prefix of the name of S matches some wildcard in PTR's wildcard list. Here we check if the filename matches as well (if it's specified) and if any of the wildcards in fact does match. */ static void walk_wild_section_match (lang_wild_statement_type *ptr, lang_input_statement_type *file, asection *s) { struct wildcard_list *sec; const char *file_spec = ptr->filename; char *p; /* Check if filenames match. */ if (file_spec == NULL) ; else if ((p = archive_path (file_spec)) != NULL) { if (!input_statement_is_archive_path (file_spec, p, file)) return; } else if (wildcardp (file_spec)) { if (fnmatch (file_spec, file->filename, 0) != 0) return; } else { lang_input_statement_type *f; /* Perform the iteration over a single file. */ f = lookup_name (file_spec); if (f != file) return; } /* If filename is excluded we're done. */ if (walk_wild_file_in_exclude_list (ptr->exclude_name_list, file)) return; /* Check section name against each wildcard spec. If there's no wildcard all sections match. */ sec = ptr->section_list; if (sec == NULL) add_matching_section (ptr, sec, s, file); else { const char *sname = bfd_section_name (s); for (; sec != NULL; sec = sec->next) { if (sec->spec.name != NULL && spec_match (&sec->spec, sname) != 0) continue; /* Don't process sections from files which were excluded. */ if (!walk_wild_file_in_exclude_list (sec->spec.exclude_name_list, file)) add_matching_section (ptr, sec, s, file); } } } /* Return the numerical value of the init_priority attribute from section name NAME. */ static int get_init_priority (const asection *sec) { const char *name = bfd_section_name (sec); const char *dot; /* GCC uses the following section names for the init_priority attribute with numerical values 101 to 65535 inclusive. A lower value means a higher priority. 1: .init_array.NNNNN/.fini_array.NNNNN: Where NNNNN is the decimal numerical value of the init_priority attribute. The order of execution in .init_array is forward and .fini_array is backward. 2: .ctors.NNNNN/.dtors.NNNNN: Where NNNNN is 65535 minus the decimal numerical value of the init_priority attribute. The order of execution in .ctors is backward and .dtors is forward. .init_array.NNNNN sections would normally be placed in an output .init_array section, .fini_array.NNNNN in .fini_array, .ctors.NNNNN in .ctors, and .dtors.NNNNN in .dtors. This means we should sort by increasing number (and could just use SORT_BY_NAME in scripts). However if .ctors.NNNNN sections are being placed in .init_array (which may also contain .init_array.NNNNN sections) or .dtors.NNNNN sections are being placed in .fini_array then we need to extract the init_priority attribute and sort on that. */ dot = strrchr (name, '.'); if (dot != NULL && ISDIGIT (dot[1])) { char *end; unsigned long init_priority = strtoul (dot + 1, &end, 10); if (*end == 0) { if (dot == name + 6 && (strncmp (name, ".ctors", 6) == 0 || strncmp (name, ".dtors", 6) == 0)) init_priority = 65535 - init_priority; if (init_priority <= INT_MAX) return init_priority; } } return -1; } /* Compare sections ASEC and BSEC according to SORT. */ static int compare_section (sort_type sort, asection *asec, asection *bsec) { int ret; int a_priority, b_priority; switch (sort) { default: abort (); case by_init_priority: a_priority = get_init_priority (asec); b_priority = get_init_priority (bsec); if (a_priority < 0 || b_priority < 0) goto sort_by_name; ret = a_priority - b_priority; if (ret) break; else goto sort_by_name; case by_alignment_name: ret = bfd_section_alignment (bsec) - bfd_section_alignment (asec); if (ret) break; /* Fall through. */ case by_name: sort_by_name: ret = strcmp (bfd_section_name (asec), bfd_section_name (bsec)); break; case by_name_alignment: ret = strcmp (bfd_section_name (asec), bfd_section_name (bsec)); if (ret) break; /* Fall through. */ case by_alignment: ret = bfd_section_alignment (bsec) - bfd_section_alignment (asec); break; } return ret; } /* PE puts the sort key in the input statement. */ static const char * sort_filename (bfd *abfd) { lang_input_statement_type *is = bfd_usrdata (abfd); if (is->sort_key) return is->sort_key; return bfd_get_filename (abfd); } /* Handle wildcard sorting. This returns the place in a binary search tree where this FILE:SECTION should be inserted for wild statement WILD where the spec SEC was the matching one. The tree is later linearized. */ static lang_section_bst_type ** wild_sort (lang_wild_statement_type *wild, struct wildcard_list *sec, lang_input_statement_type *file, asection *section) { lang_section_bst_type **tree; if (!wild->filenames_sorted && (sec == NULL || sec->spec.sorted == none || sec->spec.sorted == by_none)) { /* We might be called even if _this_ spec doesn't need sorting, in which case we simply append at the right end of tree. */ return wild->rightmost; } tree = &wild->tree; while (*tree) { /* Sorting by filename takes precedence over sorting by section name. */ if (wild->filenames_sorted) { const char *fn, *ln; bool fa, la; int i; asection *lsec = (*tree)->section; /* The PE support for the .idata section as generated by dlltool assumes that files will be sorted by the name of the archive and then the name of the file within the archive. */ fa = file->the_bfd->my_archive != NULL; if (fa) fn = sort_filename (file->the_bfd->my_archive); else fn = sort_filename (file->the_bfd); la = lsec->owner->my_archive != NULL; if (la) ln = sort_filename (lsec->owner->my_archive); else ln = sort_filename (lsec->owner); i = filename_cmp (fn, ln); if (i > 0) { tree = &((*tree)->right); continue; } else if (i < 0) { tree = &((*tree)->left); continue; } if (fa || la) { if (fa) fn = sort_filename (file->the_bfd); if (la) ln = sort_filename (lsec->owner); i = filename_cmp (fn, ln); if (i > 0) { tree = &((*tree)->right); continue; } else if (i < 0) { tree = &((*tree)->left); continue; } } } /* Here either the files are not sorted by name, or we are looking at the sections for this file. */ /* Find the correct node to append this section. */ if (sec && sec->spec.sorted != none && sec->spec.sorted != by_none && compare_section (sec->spec.sorted, section, (*tree)->section) < 0) tree = &((*tree)->left); else tree = &((*tree)->right); } return tree; } /* Use wild_sort to build a BST to sort sections. */ static void output_section_callback_sort (lang_wild_statement_type *ptr, struct wildcard_list *sec, asection *section, lang_input_statement_type *file, void *output) { lang_section_bst_type *node; lang_section_bst_type **tree; lang_output_section_statement_type *os; os = (lang_output_section_statement_type *) output; if (unique_section_p (section, os)) return; node = (lang_section_bst_type *) xmalloc (sizeof (lang_section_bst_type)); node->left = 0; node->right = 0; node->section = section; node->pattern = ptr->section_list; tree = wild_sort (ptr, sec, file, section); if (tree != NULL) { *tree = node; if (tree == ptr->rightmost) ptr->rightmost = &node->right; } } /* Convert a sorted sections' BST back to list form. */ static void output_section_callback_tree_to_list (lang_wild_statement_type *ptr, lang_section_bst_type *tree, void *output) { if (tree->left) output_section_callback_tree_to_list (ptr, tree->left, output); lang_add_section (&ptr->children, tree->section, tree->pattern, ptr->section_flag_list, (lang_output_section_statement_type *) output); if (tree->right) output_section_callback_tree_to_list (ptr, tree->right, output); free (tree); } /* Sections are matched against wildcard statements via a prefix tree. The prefix tree holds prefixes of all matching patterns (up to the first wildcard character), and the wild statement from which those patterns came. When matching a section name against the tree we're walking through the tree character by character. Each statement we hit is one that potentially matches. This is checked by actually going through the (glob) matching routines. When the section name turns out to actually match we record that section in the wild statements list of matching sections. */ /* A prefix can be matched by multiple statement, so we need a list of them. */ struct wild_stmt_list { lang_wild_statement_type *stmt; struct wild_stmt_list *next; }; /* The prefix tree itself. */ struct prefixtree { /* The list of all children (linked via .next). */ struct prefixtree *child; struct prefixtree *next; /* This tree node is responsible for the prefix of parent plus 'c'. */ char c; /* The statements that potentially can match this prefix. */ struct wild_stmt_list *stmt; }; /* We always have a root node in the prefix tree. It corresponds to the empty prefix. E.g. a glob like "*" would sit in this root. */ static struct prefixtree the_root, *ptroot = &the_root; /* Given a prefix tree in *TREE, corresponding to prefix P, find or INSERT the tree node corresponding to prefix P+C. */ static struct prefixtree * get_prefix_tree (struct prefixtree **tree, char c, bool insert) { struct prefixtree *t; for (t = *tree; t; t = t->next) if (t->c == c) return t; if (!insert) return NULL; t = (struct prefixtree *) obstack_alloc (&pt_obstack, sizeof *t); t->child = NULL; t->next = *tree; t->c = c; t->stmt = NULL; *tree = t; return t; } /* Add STMT to the set of statements that can be matched by the prefix corresponding to prefix tree T. */ static void pt_add_stmt (struct prefixtree *t, lang_wild_statement_type *stmt) { struct wild_stmt_list *sl, **psl; sl = (struct wild_stmt_list *) obstack_alloc (&pt_obstack, sizeof *sl); sl->stmt = stmt; sl->next = NULL; psl = &t->stmt; while (*psl) psl = &(*psl)->next; *psl = sl; } /* Insert STMT into the global prefix tree. */ static void insert_prefix_tree (lang_wild_statement_type *stmt) { struct wildcard_list *sec; struct prefixtree *t; if (!stmt->section_list) { /* If we have no section_list (no wildcards in the wild STMT), then every section name will match, so add this to the root. */ pt_add_stmt (ptroot, stmt); return; } for (sec = stmt->section_list; sec; sec = sec->next) { const char *name = sec->spec.name ? sec->spec.name : "*"; char c; t = ptroot; for (; (c = *name); name++) { if (c == '*' || c == '[' || c == '?') break; t = get_prefix_tree (&t->child, c, true); } /* If we hit a glob character, the matching prefix is what we saw until now. If we hit the end of pattern (hence it's no glob) then we can do better: we only need to record a match when a section name completely matches, not merely a prefix, so record the trailing 0 as well. */ if (!c) t = get_prefix_tree (&t->child, 0, true); pt_add_stmt (t, stmt); } } /* Dump T indented by INDENT spaces. */ static void debug_prefix_tree_rec (struct prefixtree *t, int indent) { for (; t; t = t->next) { struct wild_stmt_list *sl; printf ("%*s %c", indent, "", t->c); for (sl = t->stmt; sl; sl = sl->next) { struct wildcard_list *curr; printf (" %p ", sl->stmt); for (curr = sl->stmt->section_list; curr; curr = curr->next) printf ("%s ", curr->spec.name ? curr->spec.name : "*"); } printf ("\n"); debug_prefix_tree_rec (t->child, indent + 2); } } /* Dump the global prefix tree. */ static void debug_prefix_tree (void) { debug_prefix_tree_rec (ptroot, 2); } /* Like strcspn() but start to look from the end to beginning of S. Returns the length of the suffix of S consisting entirely of characters not in REJECT. */ static size_t rstrcspn (const char *s, const char *reject) { size_t len = strlen (s), sufflen = 0; while (len--) { char c = s[len]; if (strchr (reject, c) != 0) break; sufflen++; } return sufflen; } /* Analyze the wildcards in wild statement PTR to setup various things for quick matching. */ static void analyze_walk_wild_section_handler (lang_wild_statement_type *ptr) { struct wildcard_list *sec; ptr->tree = NULL; ptr->rightmost = &ptr->tree; for (sec = ptr->section_list; sec != NULL; sec = sec->next) { if (sec->spec.name) { sec->spec.namelen = strlen (sec->spec.name); sec->spec.prefixlen = strcspn (sec->spec.name, "?*["); sec->spec.suffixlen = rstrcspn (sec->spec.name + sec->spec.prefixlen, "?*]"); } else sec->spec.namelen = sec->spec.prefixlen = sec->spec.suffixlen = 0; } insert_prefix_tree (ptr); } /* Match all sections from FILE against the global prefix tree, and record them into each wild statement that has a match. */ static void resolve_wild_sections (lang_input_statement_type *file) { asection *s; if (file->flags.just_syms) return; for (s = file->the_bfd->sections; s != NULL; s = s->next) { const char *sname = bfd_section_name (s); char c = 1; struct prefixtree *t = ptroot; //printf (" YYY consider %s of %s\n", sname, file->the_bfd->filename); do { if (t->stmt) { struct wild_stmt_list *sl; for (sl = t->stmt; sl; sl = sl->next) { walk_wild_section_match (sl->stmt, file, s); //printf (" ZZZ maybe place into %p\n", sl->stmt); } } if (!c) break; c = *sname++; t = get_prefix_tree (&t->child, c, false); } while (t); } } /* Match all sections from all input files against the global prefix tree. */ static void resolve_wilds (void) { LANG_FOR_EACH_INPUT_STATEMENT (f) { //printf("XXX %s\n", f->filename); if (f->the_bfd == NULL || !bfd_check_format (f->the_bfd, bfd_archive)) resolve_wild_sections (f); else { bfd *member; /* This is an archive file. We must map each member of the archive separately. */ member = bfd_openr_next_archived_file (f->the_bfd, NULL); while (member != NULL) { /* When lookup_name is called, it will call the add_symbols entry point for the archive. For each element of the archive which is included, BFD will call ldlang_add_file, which will set the usrdata field of the member to the lang_input_statement. */ if (bfd_usrdata (member) != NULL) resolve_wild_sections (bfd_usrdata (member)); member = bfd_openr_next_archived_file (f->the_bfd, member); } } } } /* For each input section that matches wild statement S calls CALLBACK with DATA. */ static void walk_wild (lang_wild_statement_type *s, callback_t callback, void *data) { lang_statement_union_type *l; for (l = s->matching_sections.head; l; l = l->header.next) { (*callback) (s, l->input_matcher.pattern, l->input_matcher.section, l->input_matcher.input_stmt, data); } } /* lang_for_each_statement walks the parse tree and calls the provided function for each node, except those inside output section statements with constraint set to -1. */ void lang_for_each_statement_worker (void (*func) (lang_statement_union_type *), lang_statement_union_type *s) { for (; s != NULL; s = s->header.next) { func (s); switch (s->header.type) { case lang_constructors_statement_enum: lang_for_each_statement_worker (func, constructor_list.head); break; case lang_output_section_statement_enum: if (s->output_section_statement.constraint != -1) lang_for_each_statement_worker (func, s->output_section_statement.children.head); break; case lang_wild_statement_enum: lang_for_each_statement_worker (func, s->wild_statement.children.head); break; case lang_group_statement_enum: lang_for_each_statement_worker (func, s->group_statement.children.head); break; case lang_data_statement_enum: case lang_reloc_statement_enum: case lang_object_symbols_statement_enum: case lang_output_statement_enum: case lang_target_statement_enum: case lang_input_section_enum: case lang_input_statement_enum: case lang_assignment_statement_enum: case lang_padding_statement_enum: case lang_address_statement_enum: case lang_fill_statement_enum: case lang_insert_statement_enum: break; default: FAIL (); break; } } } void lang_for_each_statement (void (*func) (lang_statement_union_type *)) { lang_for_each_statement_worker (func, statement_list.head); } /*----------------------------------------------------------------------*/ void lang_list_init (lang_statement_list_type *list) { list->head = NULL; list->tail = &list->head; } static void lang_statement_append (lang_statement_list_type *list, void *element, void *field) { *(list->tail) = element; list->tail = field; } void push_stat_ptr (lang_statement_list_type *new_ptr) { if (stat_save_ptr >= stat_save + sizeof (stat_save) / sizeof (stat_save[0])) abort (); *stat_save_ptr++ = stat_ptr; stat_ptr = new_ptr; } void pop_stat_ptr (void) { if (stat_save_ptr <= stat_save) abort (); stat_ptr = *--stat_save_ptr; } /* Build a new statement node for the parse tree. */ static lang_statement_union_type * new_statement (enum statement_enum type, size_t size, lang_statement_list_type *list) { lang_statement_union_type *new_stmt; new_stmt = stat_alloc (size); new_stmt->header.type = type; new_stmt->header.next = NULL; lang_statement_append (list, new_stmt, &new_stmt->header.next); return new_stmt; } /* Build a new input file node for the language. There are several ways in which we treat an input file, eg, we only look at symbols, or prefix it with a -l etc. We can be supplied with requests for input files more than once; they may, for example be split over several lines like foo.o(.text) foo.o(.data) etc, so when asked for a file we check that we haven't got it already so we don't duplicate the bfd. */ static lang_input_statement_type * new_afile (const char *name, lang_input_file_enum_type file_type, const char *target, const char *from_filename) { lang_input_statement_type *p; lang_has_input_file = true; p = new_stat (lang_input_statement, stat_ptr); memset (&p->the_bfd, 0, sizeof (*p) - offsetof (lang_input_statement_type, the_bfd)); p->extra_search_path = NULL; p->target = target; p->flags.dynamic = input_flags.dynamic; p->flags.add_DT_NEEDED_for_dynamic = input_flags.add_DT_NEEDED_for_dynamic; p->flags.add_DT_NEEDED_for_regular = input_flags.add_DT_NEEDED_for_regular; p->flags.whole_archive = input_flags.whole_archive; p->flags.sysrooted = input_flags.sysrooted; p->sort_key = NULL; switch (file_type) { case lang_input_file_is_symbols_only_enum: p->filename = name; p->local_sym_name = name; p->flags.real = true; p->flags.just_syms = true; break; case lang_input_file_is_fake_enum: p->filename = name; p->local_sym_name = name; break; case lang_input_file_is_l_enum: if (name[0] == ':' && name[1] != '\0') { p->filename = name + 1; p->flags.full_name_provided = true; } else p->filename = name; p->local_sym_name = concat ("-l", name, (const char *) NULL); p->flags.maybe_archive = true; p->flags.real = true; p->flags.search_dirs = true; break; case lang_input_file_is_marker_enum: p->filename = name; p->local_sym_name = name; p->flags.search_dirs = true; break; case lang_input_file_is_search_file_enum: p->filename = name; p->local_sym_name = name; /* If name is a relative path, search the directory of the current linker script first. */ if (from_filename && !IS_ABSOLUTE_PATH (name)) p->extra_search_path = ldirname (from_filename); p->flags.real = true; p->flags.search_dirs = true; break; case lang_input_file_is_file_enum: p->filename = name; p->local_sym_name = name; p->flags.real = true; break; default: FAIL (); } lang_statement_append (&input_file_chain, p, &p->next_real_file); return p; } lang_input_statement_type * lang_add_input_file (const char *name, lang_input_file_enum_type file_type, const char *target) { if (name != NULL && (*name == '=' || startswith (name, "$SYSROOT"))) { lang_input_statement_type *ret; char *sysrooted_name = concat (ld_sysroot, name + (*name == '=' ? 1 : strlen ("$SYSROOT")), (const char *) NULL); /* We've now forcibly prepended the sysroot, making the input file independent of the context. Therefore, temporarily force a non-sysrooted context for this statement, so it won't get the sysroot prepended again when opened. (N.B. if it's a script, any child nodes with input files starting with "/" will be handled as "sysrooted" as they'll be found to be within the sysroot subdirectory.) */ unsigned int outer_sysrooted = input_flags.sysrooted; input_flags.sysrooted = 0; ret = new_afile (sysrooted_name, file_type, target, NULL); input_flags.sysrooted = outer_sysrooted; return ret; } return new_afile (name, file_type, target, current_input_file); } struct out_section_hash_entry { struct bfd_hash_entry root; lang_statement_union_type s; }; /* The hash table. */ static struct bfd_hash_table output_section_statement_table; /* Support routines for the hash table used by lang_output_section_find, initialize the table, fill in an entry and remove the table. */ static struct bfd_hash_entry * output_section_statement_newfunc (struct bfd_hash_entry *entry, struct bfd_hash_table *table, const char *string) { lang_output_section_statement_type **nextp; struct out_section_hash_entry *ret; if (entry == NULL) { entry = (struct bfd_hash_entry *) bfd_hash_allocate (table, sizeof (*ret)); if (entry == NULL) return entry; } entry = bfd_hash_newfunc (entry, table, string); if (entry == NULL) return entry; ret = (struct out_section_hash_entry *) entry; memset (&ret->s, 0, sizeof (ret->s)); ret->s.header.type = lang_output_section_statement_enum; ret->s.output_section_statement.subsection_alignment = NULL; ret->s.output_section_statement.section_alignment = NULL; ret->s.output_section_statement.block_value = 1; lang_list_init (&ret->s.output_section_statement.children); lang_statement_append (stat_ptr, &ret->s, &ret->s.header.next); /* For every output section statement added to the list, except the first one, lang_os_list.tail points to the "next" field of the last element of the list. */ if (lang_os_list.head != NULL) ret->s.output_section_statement.prev = ((lang_output_section_statement_type *) ((char *) lang_os_list.tail - offsetof (lang_output_section_statement_type, next))); /* GCC's strict aliasing rules prevent us from just casting the address, so we store the pointer in a variable and cast that instead. */ nextp = &ret->s.output_section_statement.next; lang_statement_append (&lang_os_list, &ret->s, nextp); return &ret->root; } static void output_section_statement_table_init (void) { if (!bfd_hash_table_init_n (&output_section_statement_table, output_section_statement_newfunc, sizeof (struct out_section_hash_entry), 61)) einfo (_("%F%P: can not create hash table: %E\n")); } static void output_section_statement_table_free (void) { bfd_hash_table_free (&output_section_statement_table); } /* Build enough state so that the parser can build its tree. */ void lang_init (void) { obstack_begin (&stat_obstack, 1000); obstack_init (&pt_obstack); stat_ptr = &statement_list; output_section_statement_table_init (); lang_list_init (stat_ptr); lang_list_init (&input_file_chain); lang_list_init (&lang_os_list); lang_list_init (&file_chain); first_file = lang_add_input_file (NULL, lang_input_file_is_marker_enum, NULL); abs_output_section = lang_output_section_statement_lookup (BFD_ABS_SECTION_NAME, 0, 1); abs_output_section->bfd_section = bfd_abs_section_ptr; asneeded_list_head = NULL; asneeded_list_tail = &asneeded_list_head; } void lang_finish (void) { output_section_statement_table_free (); } /*---------------------------------------------------------------------- A region is an area of memory declared with the MEMORY { name:org=exp, len=exp ... } syntax. We maintain a list of all the regions here. If no regions are specified in the script, then the default is used which is created when looked up to be the entire data space. If create is true we are creating a region inside a MEMORY block. In this case it is probably an error to create a region that has already been created. If we are not inside a MEMORY block it is dubious to use an undeclared region name (except DEFAULT_MEMORY_REGION) and so we issue a warning. Each region has at least one name. The first name is either DEFAULT_MEMORY_REGION or the name given in the MEMORY block. You can add alias names to an existing region within a script with REGION_ALIAS (alias, region_name). Each name corresponds to at most one region. */ static lang_memory_region_type *lang_memory_region_list; static lang_memory_region_type **lang_memory_region_list_tail = &lang_memory_region_list; lang_memory_region_type * lang_memory_region_lookup (const char *const name, bool create) { lang_memory_region_name *n; lang_memory_region_type *r; lang_memory_region_type *new_region; /* NAME is NULL for LMA memspecs if no region was specified. */ if (name == NULL) return NULL; for (r = lang_memory_region_list; r != NULL; r = r->next) for (n = &r->name_list; n != NULL; n = n->next) if (strcmp (n->name, name) == 0) { if (create) einfo (_("%P:%pS: warning: redeclaration of memory region `%s'\n"), NULL, name); return r; } if (!create && strcmp (name, DEFAULT_MEMORY_REGION)) einfo (_("%P:%pS: warning: memory region `%s' not declared\n"), NULL, name); new_region = stat_alloc (sizeof (lang_memory_region_type)); new_region->name_list.name = xstrdup (name); new_region->name_list.next = NULL; new_region->next = NULL; new_region->origin_exp = NULL; new_region->origin = 0; new_region->length_exp = NULL; new_region->length = ~(bfd_size_type) 0; new_region->current = 0; new_region->last_os = NULL; new_region->flags = 0; new_region->not_flags = 0; new_region->had_full_message = false; *lang_memory_region_list_tail = new_region; lang_memory_region_list_tail = &new_region->next; return new_region; } void lang_memory_region_alias (const char *alias, const char *region_name) { lang_memory_region_name *n; lang_memory_region_type *r; lang_memory_region_type *region; /* The default region must be unique. This ensures that it is not necessary to iterate through the name list if someone wants the check if a region is the default memory region. */ if (strcmp (region_name, DEFAULT_MEMORY_REGION) == 0 || strcmp (alias, DEFAULT_MEMORY_REGION) == 0) einfo (_("%F%P:%pS: error: alias for default memory region\n"), NULL); /* Look for the target region and check if the alias is not already in use. */ region = NULL; for (r = lang_memory_region_list; r != NULL; r = r->next) for (n = &r->name_list; n != NULL; n = n->next) { if (region == NULL && strcmp (n->name, region_name) == 0) region = r; if (strcmp (n->name, alias) == 0) einfo (_("%F%P:%pS: error: redefinition of memory region " "alias `%s'\n"), NULL, alias); } /* Check if the target region exists. */ if (region == NULL) einfo (_("%F%P:%pS: error: memory region `%s' " "for alias `%s' does not exist\n"), NULL, region_name, alias); /* Add alias to region name list. */ n = stat_alloc (sizeof (lang_memory_region_name)); n->name = xstrdup (alias); n->next = region->name_list.next; region->name_list.next = n; } static lang_memory_region_type * lang_memory_default (asection *section) { lang_memory_region_type *p; flagword sec_flags = section->flags; /* Override SEC_DATA to mean a writable section. */ if ((sec_flags & (SEC_ALLOC | SEC_READONLY | SEC_CODE)) == SEC_ALLOC) sec_flags |= SEC_DATA; for (p = lang_memory_region_list; p != NULL; p = p->next) { if ((p->flags & sec_flags) != 0 && (p->not_flags & sec_flags) == 0) { return p; } } return lang_memory_region_lookup (DEFAULT_MEMORY_REGION, false); } /* Get the output section statement directly from the userdata. */ lang_output_section_statement_type * lang_output_section_get (const asection *output_section) { return bfd_section_userdata (output_section); } /* Find or create an output_section_statement with the given NAME. If CONSTRAINT is non-zero match one with that constraint, otherwise match any non-negative constraint. If CREATE is 0 return NULL when no match exists. If CREATE is 1, create an output_section_statement when no match exists or if CONSTRAINT is SPECIAL. If CREATE is 2, always make a new output_section_statement. */ lang_output_section_statement_type * lang_output_section_statement_lookup (const char *name, int constraint, int create) { struct out_section_hash_entry *entry; entry = ((struct out_section_hash_entry *) bfd_hash_lookup (&output_section_statement_table, name, create != 0, false)); if (entry == NULL) { if (create) einfo (_("%F%P: failed creating section `%s': %E\n"), name); return NULL; } if (entry->s.output_section_statement.name != NULL) { /* We have a section of this name, but it might not have the correct constraint. */ struct out_section_hash_entry *last_ent; name = entry->s.output_section_statement.name; do { if (create != 2 && !(create && constraint == SPECIAL) && (constraint == entry->s.output_section_statement.constraint || (constraint == 0 && entry->s.output_section_statement.constraint >= 0))) return &entry->s.output_section_statement; last_ent = entry; entry = (struct out_section_hash_entry *) entry->root.next; } while (entry != NULL && name == entry->s.output_section_statement.name); if (!create) return NULL; entry = ((struct out_section_hash_entry *) output_section_statement_newfunc (NULL, &output_section_statement_table, name)); if (entry == NULL) { einfo (_("%F%P: failed creating section `%s': %E\n"), name); return NULL; } entry->root = last_ent->root; last_ent->root.next = &entry->root; } entry->s.output_section_statement.name = name; entry->s.output_section_statement.constraint = constraint; entry->s.output_section_statement.dup_output = (create == 2 || constraint == SPECIAL); return &entry->s.output_section_statement; } /* Find the next output_section_statement with the same name as OS. If CONSTRAINT is non-zero, find one with that constraint otherwise match any non-negative constraint. */ lang_output_section_statement_type * next_matching_output_section_statement (lang_output_section_statement_type *os, int constraint) { /* All output_section_statements are actually part of a struct out_section_hash_entry. */ struct out_section_hash_entry *entry = (struct out_section_hash_entry *) ((char *) os - offsetof (struct out_section_hash_entry, s.output_section_statement)); const char *name = os->name; ASSERT (name == entry->root.string); do { entry = (struct out_section_hash_entry *) entry->root.next; if (entry == NULL || name != entry->s.output_section_statement.name) return NULL; } while (constraint != entry->s.output_section_statement.constraint && (constraint != 0 || entry->s.output_section_statement.constraint < 0)); return &entry->s.output_section_statement; } /* A variant of lang_output_section_find used by place_orphan. Returns the output statement that should precede a new output statement for SEC. If an exact match is found on certain flags, sets *EXACT too. */ lang_output_section_statement_type * lang_output_section_find_by_flags (const asection *sec, flagword sec_flags, lang_output_section_statement_type **exact, lang_match_sec_type_func match_type) { lang_output_section_statement_type *first, *look, *found; flagword look_flags, differ; /* We know the first statement on this list is *ABS*. May as well skip it. */ first = (void *) lang_os_list.head; first = first->next; /* First try for an exact match. */ found = NULL; for (look = first; look; look = look->next) { look_flags = look->flags; if (look->bfd_section != NULL) { look_flags = look->bfd_section->flags; if (match_type && !match_type (link_info.output_bfd, look->bfd_section, sec->owner, sec)) continue; } differ = look_flags ^ sec_flags; if (!(differ & (SEC_HAS_CONTENTS | SEC_ALLOC | SEC_LOAD | SEC_READONLY | SEC_CODE | SEC_SMALL_DATA | SEC_THREAD_LOCAL))) found = look; } if (found != NULL) { if (exact != NULL) *exact = found; return found; } if ((sec_flags & SEC_CODE) != 0 && (sec_flags & SEC_ALLOC) != 0) { /* Try for a rw code section. */ for (look = first; look; look = look->next) { look_flags = look->flags; if (look->bfd_section != NULL) { look_flags = look->bfd_section->flags; if (match_type && !match_type (link_info.output_bfd, look->bfd_section, sec->owner, sec)) continue; } differ = look_flags ^ sec_flags; if (!(differ & (SEC_HAS_CONTENTS | SEC_ALLOC | SEC_LOAD | SEC_CODE | SEC_SMALL_DATA | SEC_THREAD_LOCAL))) found = look; } } else if ((sec_flags & SEC_READONLY) != 0 && (sec_flags & SEC_ALLOC) != 0) { /* .rodata can go after .text, .sdata2 after .rodata. */ for (look = first; look; look = look->next) { look_flags = look->flags; if (look->bfd_section != NULL) { look_flags = look->bfd_section->flags; if (match_type && !match_type (link_info.output_bfd, look->bfd_section, sec->owner, sec)) continue; } differ = look_flags ^ sec_flags; if (!(differ & (SEC_HAS_CONTENTS | SEC_ALLOC | SEC_LOAD | SEC_READONLY | SEC_SMALL_DATA)) || (!(differ & (SEC_HAS_CONTENTS | SEC_ALLOC | SEC_LOAD | SEC_READONLY)) && !(look_flags & SEC_SMALL_DATA))) found = look; } } else if ((sec_flags & SEC_THREAD_LOCAL) != 0 && (sec_flags & SEC_ALLOC) != 0) { /* .tdata can go after .data, .tbss after .tdata. Treat .tbss as if it were a loaded section, and don't use match_type. */ bool seen_thread_local = false; match_type = NULL; for (look = first; look; look = look->next) { look_flags = look->flags; if (look->bfd_section != NULL) look_flags = look->bfd_section->flags; differ = look_flags ^ (sec_flags | SEC_LOAD | SEC_HAS_CONTENTS); if (!(differ & (SEC_THREAD_LOCAL | SEC_ALLOC))) { /* .tdata and .tbss must be adjacent and in that order. */ if (!(look_flags & SEC_LOAD) && (sec_flags & SEC_LOAD)) /* ..so if we're at a .tbss section and we're placing a .tdata section stop looking and return the previous section. */ break; found = look; seen_thread_local = true; } else if (seen_thread_local) break; else if (!(differ & (SEC_HAS_CONTENTS | SEC_ALLOC | SEC_LOAD))) found = look; } } else if ((sec_flags & SEC_SMALL_DATA) != 0 && (sec_flags & SEC_ALLOC) != 0) { /* .sdata goes after .data, .sbss after .sdata. */ for (look = first; look; look = look->next) { look_flags = look->flags; if (look->bfd_section != NULL) { look_flags = look->bfd_section->flags; if (match_type && !match_type (link_info.output_bfd, look->bfd_section, sec->owner, sec)) continue; } differ = look_flags ^ sec_flags; if (!(differ & (SEC_HAS_CONTENTS | SEC_ALLOC | SEC_LOAD | SEC_THREAD_LOCAL)) || ((look_flags & SEC_SMALL_DATA) && !(sec_flags & SEC_HAS_CONTENTS))) found = look; } } else if ((sec_flags & SEC_HAS_CONTENTS) != 0 && (sec_flags & SEC_ALLOC) != 0) { /* .data goes after .rodata. */ for (look = first; look; look = look->next) { look_flags = look->flags; if (look->bfd_section != NULL) { look_flags = look->bfd_section->flags; if (match_type && !match_type (link_info.output_bfd, look->bfd_section, sec->owner, sec)) continue; } differ = look_flags ^ sec_flags; if (!(differ & (SEC_HAS_CONTENTS | SEC_ALLOC | SEC_LOAD | SEC_SMALL_DATA | SEC_THREAD_LOCAL))) found = look; } } else if ((sec_flags & SEC_ALLOC) != 0) { /* .bss goes after any other alloc section. */ for (look = first; look; look = look->next) { look_flags = look->flags; if (look->bfd_section != NULL) { look_flags = look->bfd_section->flags; if (match_type && !match_type (link_info.output_bfd, look->bfd_section, sec->owner, sec)) continue; } differ = look_flags ^ sec_flags; if (!(differ & SEC_ALLOC)) found = look; } } else { /* non-alloc go last. */ for (look = first; look; look = look->next) { look_flags = look->flags; if (look->bfd_section != NULL) look_flags = look->bfd_section->flags; differ = look_flags ^ sec_flags; if (!(differ & SEC_DEBUGGING)) found = look; } return found; } if (found || !match_type) return found; return lang_output_section_find_by_flags (sec, sec_flags, NULL, NULL); } /* Find the last output section before given output statement. Used by place_orphan. */ static asection * output_prev_sec_find (lang_output_section_statement_type *os) { lang_output_section_statement_type *lookup; for (lookup = os->prev; lookup != NULL; lookup = lookup->prev) { if (lookup->constraint < 0) continue; if (lookup->bfd_section != NULL && lookup->bfd_section->owner != NULL) return lookup->bfd_section; } return NULL; } /* Look for a suitable place for a new output section statement. The idea is to skip over anything that might be inside a SECTIONS {} statement in a script, before we find another output section statement. Assignments to "dot" before an output section statement are assumed to belong to it, except in two cases; The first assignment to dot, and assignments before non-alloc sections. Otherwise we might put an orphan before . = . + SIZEOF_HEADERS or similar assignments that set the initial address, or we might insert non-alloc note sections among assignments setting end of image symbols. */ static lang_statement_union_type ** insert_os_after (lang_output_section_statement_type *after) { lang_statement_union_type **where; lang_statement_union_type **assign = NULL; bool ignore_first; ignore_first = after == (void *) lang_os_list.head; for (where = &after->header.next; *where != NULL; where = &(*where)->header.next) { switch ((*where)->header.type) { case lang_assignment_statement_enum: if (assign == NULL) { lang_assignment_statement_type *ass; ass = &(*where)->assignment_statement; if (ass->exp->type.node_class != etree_assert && ass->exp->assign.dst[0] == '.' && ass->exp->assign.dst[1] == 0) { if (!ignore_first) assign = where; ignore_first = false; } } continue; case lang_wild_statement_enum: case lang_input_section_enum: case lang_object_symbols_statement_enum: case lang_fill_statement_enum: case lang_data_statement_enum: case lang_reloc_statement_enum: case lang_padding_statement_enum: case lang_constructors_statement_enum: assign = NULL; ignore_first = false; continue; case lang_output_section_statement_enum: if (assign != NULL) { asection *s = (*where)->output_section_statement.bfd_section; if (s == NULL || s->map_head.s == NULL || (s->flags & SEC_ALLOC) != 0) where = assign; } break; case lang_input_statement_enum: case lang_address_statement_enum: case lang_target_statement_enum: case lang_output_statement_enum: case lang_group_statement_enum: case lang_insert_statement_enum: continue; case lang_input_matcher_enum: FAIL (); } break; } return where; } lang_output_section_statement_type * lang_insert_orphan (asection *s, const char *secname, int constraint, lang_output_section_statement_type *after, struct orphan_save *place, etree_type *address, lang_statement_list_type *add_child) { lang_statement_list_type add; lang_output_section_statement_type *os; lang_output_section_statement_type **os_tail; /* If we have found an appropriate place for the output section statements for this orphan, add them to our own private list, inserting them later into the global statement list. */ if (after != NULL) { lang_list_init (&add); push_stat_ptr (&add); } if (bfd_link_relocatable (&link_info) || (s->flags & (SEC_LOAD | SEC_ALLOC)) == 0) address = exp_intop (0); os_tail = (lang_output_section_statement_type **) lang_os_list.tail; os = lang_enter_output_section_statement ( secname, address, normal_section, 0, NULL, NULL, NULL, constraint, 0); if (add_child == NULL) add_child = &os->children; lang_add_section (add_child, s, NULL, NULL, os); if (after && (s->flags & (SEC_LOAD | SEC_ALLOC)) != 0) { const char *region = (after->region ? after->region->name_list.name : DEFAULT_MEMORY_REGION); const char *lma_region = (after->lma_region ? after->lma_region->name_list.name : NULL); lang_leave_output_section_statement (NULL, region, after->phdrs, lma_region); } else lang_leave_output_section_statement (NULL, DEFAULT_MEMORY_REGION, NULL, NULL); /* Restore the global list pointer. */ if (after != NULL) pop_stat_ptr (); if (after != NULL && os->bfd_section != NULL) { asection *snew, *as; bool place_after = place->stmt == NULL; bool insert_after = true; snew = os->bfd_section; /* Shuffle the bfd section list to make the output file look neater. This is really only cosmetic. */ if (place->section == NULL && after != (void *) lang_os_list.head) { asection *bfd_section = after->bfd_section; /* If the output statement hasn't been used to place any input sections (and thus doesn't have an output bfd_section), look for the closest prior output statement having an output section. */ if (bfd_section == NULL) bfd_section = output_prev_sec_find (after); if (bfd_section != NULL && bfd_section != snew) place->section = &bfd_section->next; } if (place->section == NULL) place->section = &link_info.output_bfd->sections; as = *place->section; if (!as) { /* Put the section at the end of the list. */ /* Unlink the section. */ bfd_section_list_remove (link_info.output_bfd, snew); /* Now tack it back on in the right place. */ bfd_section_list_append (link_info.output_bfd, snew); } else if ((bfd_get_flavour (link_info.output_bfd) == bfd_target_elf_flavour) && (bfd_get_flavour (s->owner) == bfd_target_elf_flavour) && ((elf_section_type (s) == SHT_NOTE && (s->flags & SEC_LOAD) != 0) || (elf_section_type (as) == SHT_NOTE && (as->flags & SEC_LOAD) != 0))) { /* Make sure that output note sections are grouped and sorted by alignments when inserting a note section or insert a section after a note section, */ asection *sec; /* A specific section after which the output note section should be placed. */ asection *after_sec; /* True if we need to insert the orphan section after a specific section to maintain output note section order. */ bool after_sec_note = false; static asection *first_orphan_note = NULL; /* Group and sort output note section by alignments in ascending order. */ after_sec = NULL; if (elf_section_type (s) == SHT_NOTE && (s->flags & SEC_LOAD) != 0) { /* Search from the beginning for the last output note section with equal or larger alignments. NB: Don't place orphan note section after non-note sections. */ first_orphan_note = NULL; for (sec = link_info.output_bfd->sections; (sec != NULL && !bfd_is_abs_section (sec)); sec = sec->next) if (sec != snew && elf_section_type (sec) == SHT_NOTE && (sec->flags & SEC_LOAD) != 0) { if (!first_orphan_note) first_orphan_note = sec; if (sec->alignment_power >= s->alignment_power) after_sec = sec; } else if (first_orphan_note) { /* Stop if there is non-note section after the first orphan note section. */ break; } /* If this will be the first orphan note section, it can be placed at the default location. */ after_sec_note = first_orphan_note != NULL; if (after_sec == NULL && after_sec_note) { /* If all output note sections have smaller alignments, place the section before all output orphan note sections. */ after_sec = first_orphan_note; insert_after = false; } } else if (first_orphan_note) { /* Don't place non-note sections in the middle of orphan note sections. */ after_sec_note = true; after_sec = as; for (sec = as->next; (sec != NULL && !bfd_is_abs_section (sec)); sec = sec->next) if (elf_section_type (sec) == SHT_NOTE && (sec->flags & SEC_LOAD) != 0) after_sec = sec; } if (after_sec_note) { if (after_sec) { /* Search forward to insert OS after AFTER_SEC output statement. */ lang_output_section_statement_type *stmt, *next; bool found = false; for (stmt = after; stmt != NULL; stmt = next) { next = stmt->next; if (insert_after) { if (stmt->bfd_section == after_sec) { place_after = true; found = true; after = stmt; break; } } else { /* If INSERT_AFTER is FALSE, place OS before AFTER_SEC output statement. */ if (next && next->bfd_section == after_sec) { place_after = true; found = true; after = stmt; break; } } } /* Search backward to insert OS after AFTER_SEC output statement. */ if (!found) for (stmt = after; stmt != NULL; stmt = stmt->prev) { if (insert_after) { if (stmt->bfd_section == after_sec) { place_after = true; after = stmt; break; } } else { /* If INSERT_AFTER is FALSE, place OS before AFTER_SEC output statement. */ if (stmt->next->bfd_section == after_sec) { place_after = true; after = stmt; break; } } } } if (after_sec == NULL || (insert_after && after_sec->next != snew) || (!insert_after && after_sec->prev != snew)) { /* Unlink the section. */ bfd_section_list_remove (link_info.output_bfd, snew); /* Place SNEW after AFTER_SEC. If AFTER_SEC is NULL, prepend SNEW. */ if (after_sec) { if (insert_after) bfd_section_list_insert_after (link_info.output_bfd, after_sec, snew); else bfd_section_list_insert_before (link_info.output_bfd, after_sec, snew); } else bfd_section_list_prepend (link_info.output_bfd, snew); } } else if (as != snew && as->prev != snew) { /* Unlink the section. */ bfd_section_list_remove (link_info.output_bfd, snew); /* Now tack it back on in the right place. */ bfd_section_list_insert_before (link_info.output_bfd, as, snew); } } else if (as != snew && as->prev != snew) { /* Unlink the section. */ bfd_section_list_remove (link_info.output_bfd, snew); /* Now tack it back on in the right place. */ bfd_section_list_insert_before (link_info.output_bfd, as, snew); } /* Save the end of this list. Further ophans of this type will follow the one we've just added. */ place->section = &snew->next; /* The following is non-cosmetic. We try to put the output statements in some sort of reasonable order here, because they determine the final load addresses of the orphan sections. In addition, placing output statements in the wrong order may require extra segments. For instance, given a typical situation of all read-only sections placed in one segment and following that a segment containing all the read-write sections, we wouldn't want to place an orphan read/write section before or amongst the read-only ones. */ if (add.head != NULL) { lang_output_section_statement_type *newly_added_os; /* Place OS after AFTER if AFTER_NOTE is TRUE. */ if (place_after) { lang_statement_union_type **where = insert_os_after (after); *add.tail = *where; *where = add.head; place->os_tail = &after->next; } else { /* Put it after the last orphan statement we added. */ *add.tail = *place->stmt; *place->stmt = add.head; } /* Fix the global list pointer if we happened to tack our new list at the tail. */ if (*stat_ptr->tail == add.head) stat_ptr->tail = add.tail; /* Save the end of this list. */ place->stmt = add.tail; /* Do the same for the list of output section statements. */ newly_added_os = *os_tail; *os_tail = NULL; newly_added_os->prev = (lang_output_section_statement_type *) ((char *) place->os_tail - offsetof (lang_output_section_statement_type, next)); newly_added_os->next = *place->os_tail; if (newly_added_os->next != NULL) newly_added_os->next->prev = newly_added_os; *place->os_tail = newly_added_os; place->os_tail = &newly_added_os->next; /* Fixing the global list pointer here is a little different. We added to the list in lang_enter_output_section_statement, trimmed off the new output_section_statment above when assigning *os_tail = NULL, but possibly added it back in the same place when assigning *place->os_tail. */ if (*os_tail == NULL) lang_os_list.tail = (lang_statement_union_type **) os_tail; } } return os; } static void lang_print_asneeded (void) { struct asneeded_minfo *m; if (asneeded_list_head == NULL) return; minfo (_("\nAs-needed library included to satisfy reference by file (symbol)\n\n")); for (m = asneeded_list_head; m != NULL; m = m->next) { int len; minfo ("%s", m->soname); len = strlen (m->soname); if (len >= 29) { print_nl (); len = 0; } print_spaces (30 - len); if (m->ref != NULL) minfo ("%pB ", m->ref); minfo ("(%pT)\n", m->name); } } static void lang_map_flags (flagword flag) { if (flag & SEC_ALLOC) minfo ("a"); if (flag & SEC_CODE) minfo ("x"); if (flag & SEC_READONLY) minfo ("r"); if (flag & SEC_DATA) minfo ("w"); if (flag & SEC_LOAD) minfo ("l"); } void lang_map (void) { lang_memory_region_type *m; bool dis_header_printed = false; LANG_FOR_EACH_INPUT_STATEMENT (file) { asection *s; if ((file->the_bfd->flags & (BFD_LINKER_CREATED | DYNAMIC)) != 0 || file->flags.just_syms) continue; if (config.print_map_discarded) for (s = file->the_bfd->sections; s != NULL; s = s->next) if ((s->output_section == NULL || s->output_section->owner != link_info.output_bfd) && (s->flags & (SEC_LINKER_CREATED | SEC_KEEP)) == 0) { if (! dis_header_printed) { fprintf (config.map_file, _("\nDiscarded input sections\n\n")); dis_header_printed = true; } print_input_section (s, true); } } minfo (_("\nMemory Configuration\n\n")); fprintf (config.map_file, "%-16s %-18s %-18s %s\n", _("Name"), _("Origin"), _("Length"), _("Attributes")); for (m = lang_memory_region_list; m != NULL; m = m->next) { fprintf (config.map_file, "%-16s", m->name_list.name); char buf[32]; bfd_sprintf_vma (link_info.output_bfd, buf, m->origin); fprintf (config.map_file, " 0x%-16s", buf); bfd_sprintf_vma (link_info.output_bfd, buf, m->length); fprintf (config.map_file, " 0x%*s", m->flags || m->not_flags ? -17 : 0, buf); if (m->flags) lang_map_flags (m->flags); if (m->not_flags) { minfo ("!"); lang_map_flags (m->not_flags); } print_nl (); } fprintf (config.map_file, _("\nLinker script and memory map\n\n")); if (!link_info.reduce_memory_overheads) { obstack_begin (&map_obstack, 1000); bfd_link_hash_traverse (link_info.hash, sort_def_symbol, 0); } expld.phase = lang_fixed_phase_enum; lang_statement_iteration++; print_statements (); ldemul_extra_map_file_text (link_info.output_bfd, &link_info, config.map_file); } static bool sort_def_symbol (struct bfd_link_hash_entry *hash_entry, void *info ATTRIBUTE_UNUSED) { if ((hash_entry->type == bfd_link_hash_defined || hash_entry->type == bfd_link_hash_defweak) && hash_entry->u.def.section->owner != link_info.output_bfd && hash_entry->u.def.section->owner != NULL) { input_section_userdata_type *ud; struct map_symbol_def *def; ud = bfd_section_userdata (hash_entry->u.def.section); if (!ud) { ud = stat_alloc (sizeof (*ud)); bfd_set_section_userdata (hash_entry->u.def.section, ud); ud->map_symbol_def_tail = &ud->map_symbol_def_head; ud->map_symbol_def_count = 0; } else if (!ud->map_symbol_def_tail) ud->map_symbol_def_tail = &ud->map_symbol_def_head; def = (struct map_symbol_def *) obstack_alloc (&map_obstack, sizeof *def); def->entry = hash_entry; *(ud->map_symbol_def_tail) = def; ud->map_symbol_def_tail = &def->next; ud->map_symbol_def_count++; } return true; } /* Initialize an output section. */ static void init_os (lang_output_section_statement_type *s, flagword flags) { if (strcmp (s->name, DISCARD_SECTION_NAME) == 0) einfo (_("%F%P: illegal use of `%s' section\n"), DISCARD_SECTION_NAME); if (!s->dup_output) s->bfd_section = bfd_get_section_by_name (link_info.output_bfd, s->name); if (s->bfd_section == NULL) s->bfd_section = bfd_make_section_anyway_with_flags (link_info.output_bfd, s->name, flags); if (s->bfd_section == NULL) { einfo (_("%F%P: output format %s cannot represent section" " called %s: %E\n"), link_info.output_bfd->xvec->name, s->name); } s->bfd_section->output_section = s->bfd_section; s->bfd_section->output_offset = 0; /* Set the userdata of the output section to the output section statement to avoid lookup. */ bfd_set_section_userdata (s->bfd_section, s); /* If there is a base address, make sure that any sections it might mention are initialized. */ if (s->addr_tree != NULL) exp_init_os (s->addr_tree); if (s->load_base != NULL) exp_init_os (s->load_base); /* If supplied an alignment, set it. */ if (s->section_alignment != NULL) s->bfd_section->alignment_power = exp_get_power (s->section_alignment, "section alignment"); } /* Make sure that all output sections mentioned in an expression are initialized. */ static void exp_init_os (etree_type *exp) { switch (exp->type.node_class) { case etree_assign: case etree_provide: case etree_provided: exp_init_os (exp->assign.src); break; case etree_binary: exp_init_os (exp->binary.lhs); exp_init_os (exp->binary.rhs); break; case etree_trinary: exp_init_os (exp->trinary.cond); exp_init_os (exp->trinary.lhs); exp_init_os (exp->trinary.rhs); break; case etree_assert: exp_init_os (exp->assert_s.child); break; case etree_unary: exp_init_os (exp->unary.child); break; case etree_name: switch (exp->type.node_code) { case ADDR: case LOADADDR: { lang_output_section_statement_type *os; os = lang_output_section_find (exp->name.name); if (os != NULL && os->bfd_section == NULL) init_os (os, 0); } } break; default: break; } } static void section_already_linked (bfd *abfd, asection *sec, void *data) { lang_input_statement_type *entry = (lang_input_statement_type *) data; /* If we are only reading symbols from this object, then we want to discard all sections. */ if (entry->flags.just_syms) { bfd_link_just_syms (abfd, sec, &link_info); return; } /* Deal with SHF_EXCLUDE ELF sections. */ if (!bfd_link_relocatable (&link_info) && (abfd->flags & BFD_PLUGIN) == 0 && (sec->flags & (SEC_GROUP | SEC_KEEP | SEC_EXCLUDE)) == SEC_EXCLUDE) sec->output_section = bfd_abs_section_ptr; if (!(abfd->flags & DYNAMIC)) bfd_section_already_linked (abfd, sec, &link_info); } /* Returns true if SECTION is one we know will be discarded based on its section flags, otherwise returns false. */ static bool lang_discard_section_p (asection *section) { bool discard; flagword flags = section->flags; /* Discard sections marked with SEC_EXCLUDE. */ discard = (flags & SEC_EXCLUDE) != 0; /* Discard the group descriptor sections when we're finally placing the sections from within the group. */ if ((flags & SEC_GROUP) != 0 && link_info.resolve_section_groups) discard = true; /* Discard debugging sections if we are stripping debugging information. */ if ((link_info.strip == strip_debugger || link_info.strip == strip_all) && (flags & SEC_DEBUGGING) != 0) discard = true; return discard; } /* The wild routines. These expand statements like *(.text) and foo.o to a list of explicit actions, like foo.o(.text), bar.o(.text) and foo.o(.text, .data). */ /* Add SECTION to the output section OUTPUT. Do this by creating a lang_input_section statement which is placed at PTR. */ void lang_add_section (lang_statement_list_type *ptr, asection *section, struct wildcard_list *pattern, struct flag_info *sflag_info, lang_output_section_statement_type *output) { flagword flags = section->flags; bool discard; lang_input_section_type *new_section; bfd *abfd = link_info.output_bfd; /* Is this section one we know should be discarded? */ discard = lang_discard_section_p (section); /* Discard input sections which are assigned to a section named DISCARD_SECTION_NAME. */ if (strcmp (output->name, DISCARD_SECTION_NAME) == 0) discard = true; if (discard) { if (section->output_section == NULL) { /* This prevents future calls from assigning this section. */ section->output_section = bfd_abs_section_ptr; } else if (link_info.non_contiguous_regions_warnings) einfo (_("%P:%pS: warning: --enable-non-contiguous-regions makes " "section `%pA' from `%pB' match /DISCARD/ clause.\n"), NULL, section, section->owner); return; } if (sflag_info) { bool keep; keep = bfd_lookup_section_flags (&link_info, sflag_info, section); if (!keep) return; } if (section->output_section != NULL) { if (!link_info.non_contiguous_regions) return; /* SECTION has already been handled in a special way (eg. LINK_ONCE): skip it. */ if (bfd_is_abs_section (section->output_section)) return; /* Already assigned to the same output section, do not process it again, to avoid creating loops between duplicate sections later. */ if (section->output_section == output->bfd_section) return; if (link_info.non_contiguous_regions_warnings && output->bfd_section) einfo (_("%P:%pS: warning: --enable-non-contiguous-regions may " "change behaviour for section `%pA' from `%pB' (assigned to " "%pA, but additional match: %pA)\n"), NULL, section, section->owner, section->output_section, output->bfd_section); /* SECTION has already been assigned to an output section, but the user allows it to be mapped to another one in case it overflows. We'll later update the actual output section in size_input_section as appropriate. */ } /* We don't copy the SEC_NEVER_LOAD flag from an input section to an output section, because we want to be able to include a SEC_NEVER_LOAD section in the middle of an otherwise loaded section (I don't know why we want to do this, but we do). build_link_order in ldwrite.c handles this case by turning the embedded SEC_NEVER_LOAD section into a fill. */ flags &= ~ SEC_NEVER_LOAD; /* If final link, don't copy the SEC_LINK_ONCE flags, they've already been processed. One reason to do this is that on pe format targets, .text$foo sections go into .text and it's odd to see .text with SEC_LINK_ONCE set. */ if ((flags & (SEC_LINK_ONCE | SEC_GROUP)) == (SEC_LINK_ONCE | SEC_GROUP)) { if (link_info.resolve_section_groups) flags &= ~(SEC_LINK_ONCE | SEC_LINK_DUPLICATES | SEC_RELOC); else flags &= ~(SEC_LINK_DUPLICATES | SEC_RELOC); } else if (!bfd_link_relocatable (&link_info)) flags &= ~(SEC_LINK_ONCE | SEC_LINK_DUPLICATES | SEC_RELOC); switch (output->sectype) { case normal_section: case overlay_section: case first_overlay_section: case type_section: break; case noalloc_section: flags &= ~SEC_ALLOC; break; case typed_readonly_section: case readonly_section: flags |= SEC_READONLY; break; case noload_section: flags &= ~SEC_LOAD; flags |= SEC_NEVER_LOAD; /* Unfortunately GNU ld has managed to evolve two different meanings to NOLOAD in scripts. ELF gets a .bss style noload, alloc, no contents section. All others get a noload, noalloc section. */ if (bfd_get_flavour (link_info.output_bfd) == bfd_target_elf_flavour) flags &= ~SEC_HAS_CONTENTS; else flags &= ~SEC_ALLOC; break; } if (output->bfd_section == NULL) init_os (output, flags); /* If SEC_READONLY is not set in the input section, then clear it from the output section. */ output->bfd_section->flags &= flags | ~SEC_READONLY; if (output->bfd_section->linker_has_input) { /* Only set SEC_READONLY flag on the first input section. */ flags &= ~ SEC_READONLY; /* Keep SEC_MERGE and SEC_STRINGS only if they are the same. */ if ((output->bfd_section->flags & (SEC_MERGE | SEC_STRINGS)) != (flags & (SEC_MERGE | SEC_STRINGS)) || ((flags & SEC_MERGE) != 0 && output->bfd_section->entsize != section->entsize)) { output->bfd_section->flags &= ~ (SEC_MERGE | SEC_STRINGS); flags &= ~ (SEC_MERGE | SEC_STRINGS); } } output->bfd_section->flags |= flags; if (!output->bfd_section->linker_has_input) { output->bfd_section->linker_has_input = 1; /* This must happen after flags have been updated. The output section may have been created before we saw its first input section, eg. for a data statement. */ bfd_init_private_section_data (section->owner, section, link_info.output_bfd, output->bfd_section, &link_info); if ((flags & SEC_MERGE) != 0) output->bfd_section->entsize = section->entsize; } if ((flags & SEC_TIC54X_BLOCK) != 0 && bfd_get_arch (section->owner) == bfd_arch_tic54x) { /* FIXME: This value should really be obtained from the bfd... */ output->block_value = 128; } /* When a .ctors section is placed in .init_array it must be copied in reverse order. Similarly for .dtors. Set that up. */ if (bfd_get_flavour (link_info.output_bfd) == bfd_target_elf_flavour && ((startswith (section->name, ".ctors") && strcmp (output->bfd_section->name, ".init_array") == 0) || (startswith (section->name, ".dtors") && strcmp (output->bfd_section->name, ".fini_array") == 0)) && (section->name[6] == 0 || section->name[6] == '.')) section->flags |= SEC_ELF_REVERSE_COPY; if (section->alignment_power > output->bfd_section->alignment_power) output->bfd_section->alignment_power = section->alignment_power; section->output_section = output->bfd_section; if (!map_head_is_link_order) { asection *s = output->bfd_section->map_tail.s; output->bfd_section->map_tail.s = section; section->map_head.s = NULL; section->map_tail.s = s; if (s != NULL) s->map_head.s = section; else output->bfd_section->map_head.s = section; } /* Add a section reference to the list. */ new_section = new_stat (lang_input_section, ptr); new_section->section = section; new_section->pattern = pattern; } /* Expand a wild statement for a particular FILE. SECTION may be NULL, in which case it is a wild card. This assumes that the wild statement doesn't need any sorting (of filenames or sections). */ static void output_section_callback_nosort (lang_wild_statement_type *ptr, struct wildcard_list *sec ATTRIBUTE_UNUSED, asection *section, lang_input_statement_type *file ATTRIBUTE_UNUSED, void *output) { lang_output_section_statement_type *os; os = (lang_output_section_statement_type *) output; /* Exclude sections that match UNIQUE_SECTION_LIST. */ if (unique_section_p (section, os)) return; lang_add_section (&ptr->children, section, ptr->section_list, ptr->section_flag_list, os); } /* Check if all sections in a wild statement for a particular FILE are readonly. */ static void check_section_callback (lang_wild_statement_type *ptr ATTRIBUTE_UNUSED, struct wildcard_list *sec ATTRIBUTE_UNUSED, asection *section, lang_input_statement_type *file ATTRIBUTE_UNUSED, void *output) { lang_output_section_statement_type *os; os = (lang_output_section_statement_type *) output; /* Exclude sections that match UNIQUE_SECTION_LIST. */ if (unique_section_p (section, os)) return; if (section->output_section == NULL && (section->flags & SEC_READONLY) == 0) os->all_input_readonly = false; } /* This is passed a file name which must have been seen already and added to the statement tree. We will see if it has been opened already and had its symbols read. If not then we'll read it. */ static lang_input_statement_type * lookup_name (const char *name) { lang_input_statement_type *search; for (search = (void *) input_file_chain.head; search != NULL; search = search->next_real_file) { /* Use the local_sym_name as the name of the file that has already been loaded as filename might have been transformed via the search directory lookup mechanism. */ const char *filename = search->local_sym_name; if (filename != NULL && filename_cmp (filename, name) == 0) break; } if (search == NULL) { /* Arrange to splice the input statement added by new_afile into statement_list after the current input_file_chain tail. We know input_file_chain is not an empty list, and that lookup_name was called via open_input_bfds. Later calls to lookup_name should always match an existing input_statement. */ lang_statement_union_type **tail = stat_ptr->tail; lang_statement_union_type **after = (void *) ((char *) input_file_chain.tail - offsetof (lang_input_statement_type, next_real_file) + offsetof (lang_input_statement_type, header.next)); lang_statement_union_type *rest = *after; stat_ptr->tail = after; search = new_afile (name, lang_input_file_is_search_file_enum, default_target, NULL); *stat_ptr->tail = rest; if (*tail == NULL) stat_ptr->tail = tail; } /* If we have already added this file, or this file is not real don't add this file. */ if (search->flags.loaded || !search->flags.real) return search; if (!load_symbols (search, NULL)) return NULL; return search; } /* Save LIST as a list of libraries whose symbols should not be exported. */ struct excluded_lib { char *name; struct excluded_lib *next; }; static struct excluded_lib *excluded_libs; void add_excluded_libs (const char *list) { const char *p = list, *end; while (*p != '\0') { struct excluded_lib *entry; end = strpbrk (p, ",:"); if (end == NULL) end = p + strlen (p); entry = (struct excluded_lib *) xmalloc (sizeof (*entry)); entry->next = excluded_libs; entry->name = (char *) xmalloc (end - p + 1); memcpy (entry->name, p, end - p); entry->name[end - p] = '\0'; excluded_libs = entry; if (*end == '\0') break; p = end + 1; } } static void check_excluded_libs (bfd *abfd) { struct excluded_lib *lib = excluded_libs; while (lib) { int len = strlen (lib->name); const char *filename = lbasename (bfd_get_filename (abfd)); if (strcmp (lib->name, "ALL") == 0) { abfd->no_export = true; return; } if (filename_ncmp (lib->name, filename, len) == 0 && (filename[len] == '\0' || (filename[len] == '.' && filename[len + 1] == 'a' && filename[len + 2] == '\0'))) { abfd->no_export = true; return; } lib = lib->next; } } /* Get the symbols for an input file. */ bool load_symbols (lang_input_statement_type *entry, lang_statement_list_type *place) { char **matching; if (entry->flags.loaded) return true; ldfile_open_file (entry); /* Do not process further if the file was missing. */ if (entry->flags.missing_file) return true; if (trace_files || verbose) info_msg ("%pI\n", entry); if (!bfd_check_format (entry->the_bfd, bfd_archive) && !bfd_check_format_matches (entry->the_bfd, bfd_object, &matching)) { bfd_error_type err; struct lang_input_statement_flags save_flags; extern FILE *yyin; err = bfd_get_error (); /* See if the emulation has some special knowledge. */ if (ldemul_unrecognized_file (entry)) { if (err == bfd_error_file_ambiguously_recognized) free (matching); return true; } if (err == bfd_error_file_ambiguously_recognized) { char **p; einfo (_("%P: %pB: file not recognized: %E;" " matching formats:"), entry->the_bfd); for (p = matching; *p != NULL; p++) einfo (" %s", *p); free (matching); einfo ("%F\n"); } else if (err != bfd_error_file_not_recognized || place == NULL) einfo (_("%F%P: %pB: file not recognized: %E\n"), entry->the_bfd); bfd_close (entry->the_bfd); entry->the_bfd = NULL; /* Try to interpret the file as a linker script. */ save_flags = input_flags; ldfile_open_command_file (entry->filename); push_stat_ptr (place); input_flags.add_DT_NEEDED_for_regular = entry->flags.add_DT_NEEDED_for_regular; input_flags.add_DT_NEEDED_for_dynamic = entry->flags.add_DT_NEEDED_for_dynamic; input_flags.whole_archive = entry->flags.whole_archive; input_flags.dynamic = entry->flags.dynamic; ldfile_assumed_script = true; parser_input = input_script; current_input_file = entry->filename; yyparse (); current_input_file = NULL; ldfile_assumed_script = false; /* missing_file is sticky. sysrooted will already have been restored when seeing EOF in yyparse, but no harm to restore again. */ save_flags.missing_file |= input_flags.missing_file; input_flags = save_flags; pop_stat_ptr (); fclose (yyin); yyin = NULL; entry->flags.loaded = true; return true; } if (ldemul_recognized_file (entry)) return true; /* We don't call ldlang_add_file for an archive. Instead, the add_symbols entry point will call ldlang_add_file, via the add_archive_element callback, for each element of the archive which is used. */ switch (bfd_get_format (entry->the_bfd)) { default: break; case bfd_object: if (!entry->flags.reload) ldlang_add_file (entry); break; case bfd_archive: check_excluded_libs (entry->the_bfd); bfd_set_usrdata (entry->the_bfd, entry); if (entry->flags.whole_archive) { bfd *member = NULL; bool loaded = true; for (;;) { bfd *subsbfd; member = bfd_openr_next_archived_file (entry->the_bfd, member); if (member == NULL) break; if (!bfd_check_format (member, bfd_object)) { einfo (_("%F%P: %pB: member %pB in archive is not an object\n"), entry->the_bfd, member); loaded = false; } subsbfd = member; if (!(*link_info.callbacks ->add_archive_element) (&link_info, member, "--whole-archive", &subsbfd)) abort (); /* Potentially, the add_archive_element hook may have set a substitute BFD for us. */ if (!bfd_link_add_symbols (subsbfd, &link_info)) { einfo (_("%F%P: %pB: error adding symbols: %E\n"), member); loaded = false; } } entry->flags.loaded = loaded; return loaded; } break; } if (bfd_link_add_symbols (entry->the_bfd, &link_info)) entry->flags.loaded = true; else einfo (_("%F%P: %pB: error adding symbols: %E\n"), entry->the_bfd); return entry->flags.loaded; } /* Handle a wild statement. S->FILENAME or S->SECTION_LIST or both may be NULL, indicating that it is a wildcard. Separate lang_input_section statements are created for each part of the expansion; they are added after the wild statement S. OUTPUT is the output section. */ static void wild (lang_wild_statement_type *s, const char *target ATTRIBUTE_UNUSED, lang_output_section_statement_type *output) { struct wildcard_list *sec; if (s->filenames_sorted || s->any_specs_sorted) { lang_section_bst_type *tree; walk_wild (s, output_section_callback_sort, output); tree = s->tree; if (tree) { output_section_callback_tree_to_list (s, tree, output); s->tree = NULL; s->rightmost = &s->tree; } } else walk_wild (s, output_section_callback_nosort, output); if (default_common_section == NULL) for (sec = s->section_list; sec != NULL; sec = sec->next) if (sec->spec.name != NULL && strcmp (sec->spec.name, "COMMON") == 0) { /* Remember the section that common is going to in case we later get something which doesn't know where to put it. */ default_common_section = output; break; } } /* Return TRUE iff target is the sought target. */ static int get_target (const bfd_target *target, void *data) { const char *sought = (const char *) data; return strcmp (target->name, sought) == 0; } /* Like strcpy() but convert to lower case as well. */ static void stricpy (char *dest, const char *src) { char c; while ((c = *src++) != 0) *dest++ = TOLOWER (c); *dest = 0; } /* Remove the first occurrence of needle (if any) in haystack from haystack. */ static void strcut (char *haystack, const char *needle) { haystack = strstr (haystack, needle); if (haystack) { char *src; for (src = haystack + strlen (needle); *src;) *haystack++ = *src++; *haystack = 0; } } /* Compare two target format name strings. Return a value indicating how "similar" they are. */ static int name_compare (const char *first, const char *second) { char *copy1; char *copy2; int result; copy1 = (char *) xmalloc (strlen (first) + 1); copy2 = (char *) xmalloc (strlen (second) + 1); /* Convert the names to lower case. */ stricpy (copy1, first); stricpy (copy2, second); /* Remove size and endian strings from the name. */ strcut (copy1, "big"); strcut (copy1, "little"); strcut (copy2, "big"); strcut (copy2, "little"); /* Return a value based on how many characters match, starting from the beginning. If both strings are the same then return 10 * their length. */ for (result = 0; copy1[result] == copy2[result]; result++) if (copy1[result] == 0) { result *= 10; break; } free (copy1); free (copy2); return result; } /* Set by closest_target_match() below. */ static const bfd_target *winner; /* Scan all the valid bfd targets looking for one that has the endianness requirement that was specified on the command line, and is the nearest match to the original output target. */ static int closest_target_match (const bfd_target *target, void *data) { const bfd_target *original = (const bfd_target *) data; if (command_line.endian == ENDIAN_BIG && target->byteorder != BFD_ENDIAN_BIG) return 0; if (command_line.endian == ENDIAN_LITTLE && target->byteorder != BFD_ENDIAN_LITTLE) return 0; /* Must be the same flavour. */ if (target->flavour != original->flavour) return 0; /* Ignore generic big and little endian elf vectors. */ if (strcmp (target->name, "elf32-big") == 0 || strcmp (target->name, "elf64-big") == 0 || strcmp (target->name, "elf32-little") == 0 || strcmp (target->name, "elf64-little") == 0) return 0; /* If we have not found a potential winner yet, then record this one. */ if (winner == NULL) { winner = target; return 0; } /* Oh dear, we now have two potential candidates for a successful match. Compare their names and choose the better one. */ if (name_compare (target->name, original->name) > name_compare (winner->name, original->name)) winner = target; /* Keep on searching until wqe have checked them all. */ return 0; } /* Return the BFD target format of the first input file. */ static const char * get_first_input_target (void) { const char *target = NULL; LANG_FOR_EACH_INPUT_STATEMENT (s) { if (s->header.type == lang_input_statement_enum && s->flags.real) { ldfile_open_file (s); if (s->the_bfd != NULL && bfd_check_format (s->the_bfd, bfd_object)) { target = bfd_get_target (s->the_bfd); if (target != NULL) break; } } } return target; } const char * lang_get_output_target (void) { const char *target; /* Has the user told us which output format to use? */ if (output_target != NULL) return output_target; /* No - has the current target been set to something other than the default? */ if (current_target != default_target && current_target != NULL) return current_target; /* No - can we determine the format of the first input file? */ target = get_first_input_target (); if (target != NULL) return target; /* Failed - use the default output target. */ return default_target; } /* Open the output file. */ static void open_output (const char *name) { lang_input_statement_type *f; char *out = lrealpath (name); for (f = (void *) input_file_chain.head; f != NULL; f = f->next_real_file) if (f->flags.real) { char *in = lrealpath (f->local_sym_name); if (filename_cmp (in, out) == 0) einfo (_("%F%P: input file '%s' is the same as output file\n"), f->filename); free (in); } free (out); output_target = lang_get_output_target (); /* Has the user requested a particular endianness on the command line? */ if (command_line.endian != ENDIAN_UNSET) { /* Get the chosen target. */ const bfd_target *target = bfd_iterate_over_targets (get_target, (void *) output_target); /* If the target is not supported, we cannot do anything. */ if (target != NULL) { enum bfd_endian desired_endian; if (command_line.endian == ENDIAN_BIG) desired_endian = BFD_ENDIAN_BIG; else desired_endian = BFD_ENDIAN_LITTLE; /* See if the target has the wrong endianness. This should not happen if the linker script has provided big and little endian alternatives, but some scrips don't do this. */ if (target->byteorder != desired_endian) { /* If it does, then see if the target provides an alternative with the correct endianness. */ if (target->alternative_target != NULL && (target->alternative_target->byteorder == desired_endian)) output_target = target->alternative_target->name; else { /* Try to find a target as similar as possible to the default target, but which has the desired endian characteristic. */ bfd_iterate_over_targets (closest_target_match, (void *) target); /* Oh dear - we could not find any targets that satisfy our requirements. */ if (winner == NULL) einfo (_("%P: warning: could not find any targets" " that match endianness requirement\n")); else output_target = winner->name; } } } } link_info.output_bfd = bfd_openw (name, output_target); if (link_info.output_bfd == NULL) { if (bfd_get_error () == bfd_error_invalid_target) einfo (_("%F%P: target %s not found\n"), output_target); einfo (_("%F%P: cannot open output file %s: %E\n"), name); } delete_output_file_on_failure = true; if (!bfd_set_format (link_info.output_bfd, bfd_object)) einfo (_("%F%P: %s: can not make object file: %E\n"), name); if (!bfd_set_arch_mach (link_info.output_bfd, ldfile_output_architecture, ldfile_output_machine)) einfo (_("%F%P: %s: can not set architecture: %E\n"), name); link_info.hash = bfd_link_hash_table_create (link_info.output_bfd); if (link_info.hash == NULL) einfo (_("%F%P: can not create hash table: %E\n")); bfd_set_gp_size (link_info.output_bfd, g_switch_value); } static void ldlang_open_output (lang_statement_union_type *statement) { switch (statement->header.type) { case lang_output_statement_enum: ASSERT (link_info.output_bfd == NULL); open_output (statement->output_statement.name); ldemul_set_output_arch (); if (config.magic_demand_paged && !bfd_link_relocatable (&link_info)) link_info.output_bfd->flags |= D_PAGED; else link_info.output_bfd->flags &= ~D_PAGED; if (config.text_read_only) link_info.output_bfd->flags |= WP_TEXT; else link_info.output_bfd->flags &= ~WP_TEXT; if (link_info.traditional_format) link_info.output_bfd->flags |= BFD_TRADITIONAL_FORMAT; else link_info.output_bfd->flags &= ~BFD_TRADITIONAL_FORMAT; break; case lang_target_statement_enum: current_target = statement->target_statement.target; break; default: break; } } static void init_opb (asection *s) { unsigned int x; opb_shift = 0; if (bfd_get_flavour (link_info.output_bfd) == bfd_target_elf_flavour && s != NULL && (s->flags & SEC_ELF_OCTETS) != 0) return; x = bfd_arch_mach_octets_per_byte (ldfile_output_architecture, ldfile_output_machine); if (x > 1) while ((x & 1) == 0) { x >>= 1; ++opb_shift; } ASSERT (x == 1); } /* Open all the input files. */ enum open_bfd_mode { OPEN_BFD_NORMAL = 0, OPEN_BFD_FORCE = 1, OPEN_BFD_RESCAN = 2 }; #if BFD_SUPPORTS_PLUGINS static lang_input_statement_type *plugin_insert = NULL; static struct bfd_link_hash_entry *plugin_undefs = NULL; #endif static void open_input_bfds (lang_statement_union_type *s, enum open_bfd_mode mode) { for (; s != NULL; s = s->header.next) { switch (s->header.type) { case lang_constructors_statement_enum: open_input_bfds (constructor_list.head, mode); break; case lang_output_section_statement_enum: open_input_bfds (s->output_section_statement.children.head, mode); break; case lang_wild_statement_enum: /* Maybe we should load the file's symbols. */ if ((mode & OPEN_BFD_RESCAN) == 0 && s->wild_statement.filename && !wildcardp (s->wild_statement.filename) && !archive_path (s->wild_statement.filename)) lookup_name (s->wild_statement.filename); open_input_bfds (s->wild_statement.children.head, mode); break; case lang_group_statement_enum: { struct bfd_link_hash_entry *undefs; #if BFD_SUPPORTS_PLUGINS lang_input_statement_type *plugin_insert_save; #endif /* We must continually search the entries in the group until no new symbols are added to the list of undefined symbols. */ do { #if BFD_SUPPORTS_PLUGINS plugin_insert_save = plugin_insert; #endif undefs = link_info.hash->undefs_tail; open_input_bfds (s->group_statement.children.head, mode | OPEN_BFD_FORCE); } while (undefs != link_info.hash->undefs_tail #if BFD_SUPPORTS_PLUGINS /* Objects inserted by a plugin, which are loaded before we hit this loop, may have added new undefs. */ || (plugin_insert != plugin_insert_save && plugin_undefs) #endif ); } break; case lang_target_statement_enum: current_target = s->target_statement.target; break; case lang_input_statement_enum: if (s->input_statement.flags.real) { lang_statement_union_type **os_tail; lang_statement_list_type add; bfd *abfd; s->input_statement.target = current_target; /* If we are being called from within a group, and this is an archive which has already been searched, then force it to be researched unless the whole archive has been loaded already. Do the same for a rescan. Likewise reload --as-needed shared libs. */ if (mode != OPEN_BFD_NORMAL #if BFD_SUPPORTS_PLUGINS && ((mode & OPEN_BFD_RESCAN) == 0 || plugin_insert == NULL) #endif && s->input_statement.flags.loaded && (abfd = s->input_statement.the_bfd) != NULL && ((bfd_get_format (abfd) == bfd_archive && !s->input_statement.flags.whole_archive) || (bfd_get_format (abfd) == bfd_object && ((abfd->flags) & DYNAMIC) != 0 && s->input_statement.flags.add_DT_NEEDED_for_regular && bfd_get_flavour (abfd) == bfd_target_elf_flavour && (elf_dyn_lib_class (abfd) & DYN_AS_NEEDED) != 0))) { s->input_statement.flags.loaded = false; s->input_statement.flags.reload = true; } os_tail = lang_os_list.tail; lang_list_init (&add); if (!load_symbols (&s->input_statement, &add)) config.make_executable = false; if (add.head != NULL) { /* If this was a script with output sections then tack any added statements on to the end of the list. This avoids having to reorder the output section statement list. Very likely the user forgot -T, and whatever we do here will not meet naive user expectations. */ if (os_tail != lang_os_list.tail) { einfo (_("%P: warning: %s contains output sections;" " did you forget -T?\n"), s->input_statement.filename); *stat_ptr->tail = add.head; stat_ptr->tail = add.tail; } else { *add.tail = s->header.next; s->header.next = add.head; } } } #if BFD_SUPPORTS_PLUGINS /* If we have found the point at which a plugin added new files, clear plugin_insert to enable archive rescan. */ if (&s->input_statement == plugin_insert) plugin_insert = NULL; #endif break; case lang_assignment_statement_enum: if (s->assignment_statement.exp->type.node_class != etree_assert) exp_fold_tree_no_dot (s->assignment_statement.exp); break; default: break; } } /* Exit if any of the files were missing. */ if (input_flags.missing_file) einfo ("%F"); } #ifdef ENABLE_LIBCTF /* Emit CTF errors and warnings. fp can be NULL to report errors/warnings that happened specifically at CTF open time. */ static void lang_ctf_errs_warnings (ctf_dict_t *fp) { ctf_next_t *i = NULL; char *text; int is_warning; int err; while ((text = ctf_errwarning_next (fp, &i, &is_warning, &err)) != NULL) { einfo (_("%s: %s\n"), is_warning ? _("CTF warning"): _("CTF error"), text); free (text); } if (err != ECTF_NEXT_END) { einfo (_("CTF error: cannot get CTF errors: `%s'\n"), ctf_errmsg (err)); } /* `err' returns errors from the error/warning iterator in particular. These never assert. But if we have an fp, that could have recorded an assertion failure: assert if it has done so. */ ASSERT (!fp || ctf_errno (fp) != ECTF_INTERNAL); } /* Open the CTF sections in the input files with libctf: if any were opened, create a fake input file that we'll write the merged CTF data to later on. */ static void ldlang_open_ctf (void) { int any_ctf = 0; int err; LANG_FOR_EACH_INPUT_STATEMENT (file) { asection *sect; /* Incoming files from the compiler have a single ctf_dict_t in them (which is presented to us by the libctf API in a ctf_archive_t wrapper): files derived from a previous relocatable link have a CTF archive containing possibly many CTF files. */ if ((file->the_ctf = ctf_bfdopen (file->the_bfd, &err)) == NULL) { if (err != ECTF_NOCTFDATA) { lang_ctf_errs_warnings (NULL); einfo (_("%P: warning: CTF section in %pB not loaded; " "its types will be discarded: %s\n"), file->the_bfd, ctf_errmsg (err)); } continue; } /* Prevent the contents of this section from being written, while requiring the section itself to be duplicated in the output, but only once. */ /* This section must exist if ctf_bfdopen() succeeded. */ sect = bfd_get_section_by_name (file->the_bfd, ".ctf"); sect->size = 0; sect->flags |= SEC_NEVER_LOAD | SEC_HAS_CONTENTS | SEC_LINKER_CREATED; if (any_ctf) sect->flags |= SEC_EXCLUDE; any_ctf = 1; } if (!any_ctf) { ctf_output = NULL; return; } if ((ctf_output = ctf_create (&err)) != NULL) return; einfo (_("%P: warning: CTF output not created: `%s'\n"), ctf_errmsg (err)); LANG_FOR_EACH_INPUT_STATEMENT (errfile) ctf_close (errfile->the_ctf); } /* Merge together CTF sections. After this, only the symtab-dependent function and data object sections need adjustment. */ static void lang_merge_ctf (void) { asection *output_sect; int flags = 0; if (!ctf_output) return; output_sect = bfd_get_section_by_name (link_info.output_bfd, ".ctf"); /* If the section was discarded, don't waste time merging. */ if (output_sect == NULL) { ctf_dict_close (ctf_output); ctf_output = NULL; LANG_FOR_EACH_INPUT_STATEMENT (file) { ctf_close (file->the_ctf); file->the_ctf = NULL; } return; } LANG_FOR_EACH_INPUT_STATEMENT (file) { if (!file->the_ctf) continue; /* Takes ownership of file->the_ctf. */ if (ctf_link_add_ctf (ctf_output, file->the_ctf, file->filename) < 0) { einfo (_("%P: warning: CTF section in %pB cannot be linked: `%s'\n"), file->the_bfd, ctf_errmsg (ctf_errno (ctf_output))); ctf_close (file->the_ctf); file->the_ctf = NULL; continue; } } if (!config.ctf_share_duplicated) flags = CTF_LINK_SHARE_UNCONFLICTED; else flags = CTF_LINK_SHARE_DUPLICATED; if (!config.ctf_variables) flags |= CTF_LINK_OMIT_VARIABLES_SECTION; if (bfd_link_relocatable (&link_info)) flags |= CTF_LINK_NO_FILTER_REPORTED_SYMS; if (ctf_link (ctf_output, flags) < 0) { lang_ctf_errs_warnings (ctf_output); einfo (_("%P: warning: CTF linking failed; " "output will have no CTF section: %s\n"), ctf_errmsg (ctf_errno (ctf_output))); if (output_sect) { output_sect->size = 0; output_sect->flags |= SEC_EXCLUDE; } } /* Output any lingering errors that didn't come from ctf_link. */ lang_ctf_errs_warnings (ctf_output); } /* Let the emulation acquire strings from the dynamic strtab to help it optimize the CTF, if supported. */ void ldlang_ctf_acquire_strings (struct elf_strtab_hash *dynstrtab) { ldemul_acquire_strings_for_ctf (ctf_output, dynstrtab); } /* Inform the emulation about the addition of a new dynamic symbol, in BFD internal format. */ void ldlang_ctf_new_dynsym (int symidx, struct elf_internal_sym *sym) { ldemul_new_dynsym_for_ctf (ctf_output, symidx, sym); } /* Write out the CTF section. Called early, if the emulation isn't going to need to dedup against the strtab and symtab, then possibly called from the target linker code if the dedup has happened. */ static void lang_write_ctf (int late) { size_t output_size; asection *output_sect; if (!ctf_output) return; if (late) { /* Emit CTF late if this emulation says it can do so. */ if (ldemul_emit_ctf_early ()) return; } else { if (!ldemul_emit_ctf_early ()) return; } /* Inform the emulation that all the symbols that will be received have been. */ ldemul_new_dynsym_for_ctf (ctf_output, 0, NULL); /* Emit CTF. */ output_sect = bfd_get_section_by_name (link_info.output_bfd, ".ctf"); if (output_sect) { output_sect->contents = ctf_link_write (ctf_output, &output_size, CTF_COMPRESSION_THRESHOLD); output_sect->size = output_size; output_sect->flags |= SEC_IN_MEMORY | SEC_KEEP; lang_ctf_errs_warnings (ctf_output); if (!output_sect->contents) { einfo (_("%P: warning: CTF section emission failed; " "output will have no CTF section: %s\n"), ctf_errmsg (ctf_errno (ctf_output))); output_sect->size = 0; output_sect->flags |= SEC_EXCLUDE; } } /* This also closes every CTF input file used in the link. */ ctf_dict_close (ctf_output); ctf_output = NULL; LANG_FOR_EACH_INPUT_STATEMENT (file) file->the_ctf = NULL; } /* Write out the CTF section late, if the emulation needs that. */ void ldlang_write_ctf_late (void) { /* Trigger a "late call", if the emulation needs one. */ lang_write_ctf (1); } #else static void ldlang_open_ctf (void) { LANG_FOR_EACH_INPUT_STATEMENT (file) { asection *sect; /* If built without CTF, warn and delete all CTF sections from the output. (The alternative would be to simply concatenate them, which does not yield a valid CTF section.) */ if ((sect = bfd_get_section_by_name (file->the_bfd, ".ctf")) != NULL) { einfo (_("%P: warning: CTF section in %pB not linkable: " "%P was built without support for CTF\n"), file->the_bfd); sect->size = 0; sect->flags |= SEC_EXCLUDE; } } } static void lang_merge_ctf (void) {} void ldlang_ctf_acquire_strings (struct elf_strtab_hash *dynstrtab ATTRIBUTE_UNUSED) {} void ldlang_ctf_new_dynsym (int symidx ATTRIBUTE_UNUSED, struct elf_internal_sym *sym ATTRIBUTE_UNUSED) {} static void lang_write_ctf (int late ATTRIBUTE_UNUSED) {} void ldlang_write_ctf_late (void) {} #endif /* Add the supplied name to the symbol table as an undefined reference. This is a two step process as the symbol table doesn't even exist at the time the ld command line is processed. First we put the name on a list, then, once the output file has been opened, transfer the name to the symbol table. */ typedef struct bfd_sym_chain ldlang_undef_chain_list_type; #define ldlang_undef_chain_list_head entry_symbol.next void ldlang_add_undef (const char *const name, bool cmdline ATTRIBUTE_UNUSED) { ldlang_undef_chain_list_type *new_undef; new_undef = stat_alloc (sizeof (*new_undef)); new_undef->next = ldlang_undef_chain_list_head; ldlang_undef_chain_list_head = new_undef; new_undef->name = xstrdup (name); if (link_info.output_bfd != NULL) insert_undefined (new_undef->name); } /* Insert NAME as undefined in the symbol table. */ static void insert_undefined (const char *name) { struct bfd_link_hash_entry *h; h = bfd_link_hash_lookup (link_info.hash, name, true, false, true); if (h == NULL) einfo (_("%F%P: bfd_link_hash_lookup failed: %E\n")); if (h->type == bfd_link_hash_new) { h->type = bfd_link_hash_undefined; h->u.undef.abfd = NULL; h->non_ir_ref_regular = true; bfd_link_add_undef (link_info.hash, h); } } /* Run through the list of undefineds created above and place them into the linker hash table as undefined symbols belonging to the script file. */ static void lang_place_undefineds (void) { ldlang_undef_chain_list_type *ptr; for (ptr = ldlang_undef_chain_list_head; ptr != NULL; ptr = ptr->next) insert_undefined (ptr->name); } /* Mark -u symbols against garbage collection. */ static void lang_mark_undefineds (void) { ldlang_undef_chain_list_type *ptr; if (is_elf_hash_table (link_info.hash)) for (ptr = ldlang_undef_chain_list_head; ptr != NULL; ptr = ptr->next) { struct elf_link_hash_entry *h = (struct elf_link_hash_entry *) bfd_link_hash_lookup (link_info.hash, ptr->name, false, false, true); if (h != NULL) h->mark = 1; } } /* Structure used to build the list of symbols that the user has required be defined. */ struct require_defined_symbol { const char *name; struct require_defined_symbol *next; }; /* The list of symbols that the user has required be defined. */ static struct require_defined_symbol *require_defined_symbol_list; /* Add a new symbol NAME to the list of symbols that are required to be defined. */ void ldlang_add_require_defined (const char *const name) { struct require_defined_symbol *ptr; ldlang_add_undef (name, true); ptr = stat_alloc (sizeof (*ptr)); ptr->next = require_defined_symbol_list; ptr->name = strdup (name); require_defined_symbol_list = ptr; } /* Check that all symbols the user required to be defined, are defined, raise an error if we find a symbol that is not defined. */ static void ldlang_check_require_defined_symbols (void) { struct require_defined_symbol *ptr; for (ptr = require_defined_symbol_list; ptr != NULL; ptr = ptr->next) { struct bfd_link_hash_entry *h; h = bfd_link_hash_lookup (link_info.hash, ptr->name, false, false, true); if (h == NULL || (h->type != bfd_link_hash_defined && h->type != bfd_link_hash_defweak)) einfo(_("%X%P: required symbol `%s' not defined\n"), ptr->name); } } /* Check for all readonly or some readwrite sections. */ static void check_input_sections (lang_statement_union_type *s, lang_output_section_statement_type *output_section_statement) { for (; s != NULL; s = s->header.next) { switch (s->header.type) { case lang_wild_statement_enum: walk_wild (&s->wild_statement, check_section_callback, output_section_statement); if (!output_section_statement->all_input_readonly) return; break; case lang_constructors_statement_enum: check_input_sections (constructor_list.head, output_section_statement); if (!output_section_statement->all_input_readonly) return; break; case lang_group_statement_enum: check_input_sections (s->group_statement.children.head, output_section_statement); if (!output_section_statement->all_input_readonly) return; break; default: break; } } } /* Update wildcard statements if needed. */ static void update_wild_statements (lang_statement_union_type *s) { struct wildcard_list *sec; switch (sort_section) { default: FAIL (); case none: break; case by_name: case by_alignment: for (; s != NULL; s = s->header.next) { switch (s->header.type) { default: break; case lang_wild_statement_enum: for (sec = s->wild_statement.section_list; sec != NULL; sec = sec->next) /* Don't sort .init/.fini sections. */ if (strcmp (sec->spec.name, ".init") != 0 && strcmp (sec->spec.name, ".fini") != 0) { switch (sec->spec.sorted) { case none: sec->spec.sorted = sort_section; break; case by_name: if (sort_section == by_alignment) sec->spec.sorted = by_name_alignment; break; case by_alignment: if (sort_section == by_name) sec->spec.sorted = by_alignment_name; break; default: break; } s->wild_statement.any_specs_sorted = true; } break; case lang_constructors_statement_enum: update_wild_statements (constructor_list.head); break; case lang_output_section_statement_enum: update_wild_statements (s->output_section_statement.children.head); break; case lang_group_statement_enum: update_wild_statements (s->group_statement.children.head); break; } } break; } } /* Open input files and attach to output sections. */ static void map_input_to_output_sections (lang_statement_union_type *s, const char *target, lang_output_section_statement_type *os) { for (; s != NULL; s = s->header.next) { lang_output_section_statement_type *tos; flagword flags; unsigned int type = 0; switch (s->header.type) { case lang_wild_statement_enum: wild (&s->wild_statement, target, os); break; case lang_constructors_statement_enum: map_input_to_output_sections (constructor_list.head, target, os); break; case lang_output_section_statement_enum: tos = &s->output_section_statement; if (tos->constraint == ONLY_IF_RW || tos->constraint == ONLY_IF_RO) { tos->all_input_readonly = true; check_input_sections (tos->children.head, tos); if (tos->all_input_readonly != (tos->constraint == ONLY_IF_RO)) tos->constraint = -1; } if (tos->constraint >= 0) map_input_to_output_sections (tos->children.head, target, tos); break; case lang_output_statement_enum: break; case lang_target_statement_enum: target = s->target_statement.target; break; case lang_group_statement_enum: map_input_to_output_sections (s->group_statement.children.head, target, os); break; case lang_data_statement_enum: /* Make sure that any sections mentioned in the expression are initialized. */ exp_init_os (s->data_statement.exp); /* The output section gets CONTENTS, ALLOC and LOAD, but these may be overridden by the script. */ flags = SEC_HAS_CONTENTS | SEC_ALLOC | SEC_LOAD; switch (os->sectype) { case normal_section: case overlay_section: case first_overlay_section: break; case noalloc_section: flags = SEC_HAS_CONTENTS; break; case readonly_section: flags |= SEC_READONLY; break; case typed_readonly_section: flags |= SEC_READONLY; /* Fall through. */ case type_section: if (os->sectype_value->type.node_class == etree_name && os->sectype_value->type.node_code == NAME) { const char *name = os->sectype_value->name.name; if (strcmp (name, "SHT_PROGBITS") == 0) type = SHT_PROGBITS; else if (strcmp (name, "SHT_STRTAB") == 0) type = SHT_STRTAB; else if (strcmp (name, "SHT_NOTE") == 0) type = SHT_NOTE; else if (strcmp (name, "SHT_NOBITS") == 0) type = SHT_NOBITS; else if (strcmp (name, "SHT_INIT_ARRAY") == 0) type = SHT_INIT_ARRAY; else if (strcmp (name, "SHT_FINI_ARRAY") == 0) type = SHT_FINI_ARRAY; else if (strcmp (name, "SHT_PREINIT_ARRAY") == 0) type = SHT_PREINIT_ARRAY; else einfo (_ ("%F%P: invalid type for output section `%s'\n"), os->name); } else { exp_fold_tree_no_dot (os->sectype_value); if (expld.result.valid_p) type = expld.result.value; else einfo (_ ("%F%P: invalid type for output section `%s'\n"), os->name); } break; case noload_section: if (bfd_get_flavour (link_info.output_bfd) == bfd_target_elf_flavour) flags = SEC_NEVER_LOAD | SEC_ALLOC; else flags = SEC_NEVER_LOAD | SEC_HAS_CONTENTS; break; } if (os->bfd_section == NULL) init_os (os, flags | SEC_READONLY); else os->bfd_section->flags |= flags; os->bfd_section->type = type; break; case lang_input_section_enum: break; case lang_fill_statement_enum: case lang_object_symbols_statement_enum: case lang_reloc_statement_enum: case lang_padding_statement_enum: case lang_input_statement_enum: if (os != NULL && os->bfd_section == NULL) init_os (os, 0); break; case lang_assignment_statement_enum: if (os != NULL && os->bfd_section == NULL) init_os (os, 0); /* Make sure that any sections mentioned in the assignment are initialized. */ exp_init_os (s->assignment_statement.exp); break; case lang_address_statement_enum: /* Mark the specified section with the supplied address. If this section was actually a segment marker, then the directive is ignored if the linker script explicitly processed the segment marker. Originally, the linker treated segment directives (like -Ttext on the command-line) as section directives. We honor the section directive semantics for backwards compatibility; linker scripts that do not specifically check for SEGMENT_START automatically get the old semantics. */ if (!s->address_statement.segment || !s->address_statement.segment->used) { const char *name = s->address_statement.section_name; /* Create the output section statement here so that orphans with a set address will be placed after other script sections. If we let the orphan placement code place them in amongst other sections then the address will affect following script sections, which is likely to surprise naive users. */ tos = lang_output_section_statement_lookup (name, 0, 1); tos->addr_tree = s->address_statement.address; if (tos->bfd_section == NULL) init_os (tos, 0); } break; case lang_insert_statement_enum: break; case lang_input_matcher_enum: FAIL (); } } } /* An insert statement snips out all the linker statements from the start of the list and places them after the output section statement specified by the insert. This operation is complicated by the fact that we keep a doubly linked list of output section statements as well as the singly linked list of all statements. FIXME someday: Twiddling with the list not only moves statements from the user's script but also input and group statements that are built from command line object files and --start-group. We only get away with this because the list pointers used by file_chain and input_file_chain are not reordered, and processing via statement_list after this point mostly ignores input statements. One exception is the map file, where LOAD and START GROUP/END GROUP can end up looking odd. */ static void process_insert_statements (lang_statement_union_type **start) { lang_statement_union_type **s; lang_output_section_statement_type *first_os = NULL; lang_output_section_statement_type *last_os = NULL; lang_output_section_statement_type *os; s = start; while (*s != NULL) { if ((*s)->header.type == lang_output_section_statement_enum) { /* Keep pointers to the first and last output section statement in the sequence we may be about to move. */ os = &(*s)->output_section_statement; ASSERT (last_os == NULL || last_os->next == os); last_os = os; /* Set constraint negative so that lang_output_section_find won't match this output section statement. At this stage in linking constraint has values in the range [-1, ONLY_IN_RW]. */ last_os->constraint = -2 - last_os->constraint; if (first_os == NULL) first_os = last_os; } else if ((*s)->header.type == lang_group_statement_enum) { /* A user might put -T between --start-group and --end-group. One way this odd construct might arise is from a wrapper around ld to change library search behaviour. For example: #! /bin/sh exec real_ld --start-group "$@" --end-group This isn't completely unreasonable so go looking inside a group statement for insert statements. */ process_insert_statements (&(*s)->group_statement.children.head); } else if ((*s)->header.type == lang_insert_statement_enum) { lang_insert_statement_type *i = &(*s)->insert_statement; lang_output_section_statement_type *where; lang_statement_union_type **ptr; lang_statement_union_type *first; if (link_info.non_contiguous_regions) { einfo (_("warning: INSERT statement in linker script is " "incompatible with --enable-non-contiguous-regions.\n")); } where = lang_output_section_find (i->where); if (where != NULL && i->is_before) { do where = where->prev; while (where != NULL && where->constraint < 0); } if (where == NULL) { einfo (_("%F%P: %s not found for insert\n"), i->where); return; } /* Deal with reordering the output section statement list. */ if (last_os != NULL) { asection *first_sec, *last_sec; struct lang_output_section_statement_struct **next; /* Snip out the output sections we are moving. */ first_os->prev->next = last_os->next; if (last_os->next == NULL) { next = &first_os->prev->next; lang_os_list.tail = (lang_statement_union_type **) next; } else last_os->next->prev = first_os->prev; /* Add them in at the new position. */ last_os->next = where->next; if (where->next == NULL) { next = &last_os->next; lang_os_list.tail = (lang_statement_union_type **) next; } else where->next->prev = last_os; first_os->prev = where; where->next = first_os; /* Move the bfd sections in the same way. */ first_sec = NULL; last_sec = NULL; for (os = first_os; os != NULL; os = os->next) { os->constraint = -2 - os->constraint; if (os->bfd_section != NULL && os->bfd_section->owner != NULL) { last_sec = os->bfd_section; if (first_sec == NULL) first_sec = last_sec; } if (os == last_os) break; } if (last_sec != NULL) { asection *sec = where->bfd_section; if (sec == NULL) sec = output_prev_sec_find (where); /* The place we want to insert must come after the sections we are moving. So if we find no section or if the section is the same as our last section, then no move is needed. */ if (sec != NULL && sec != last_sec) { /* Trim them off. */ if (first_sec->prev != NULL) first_sec->prev->next = last_sec->next; else link_info.output_bfd->sections = last_sec->next; if (last_sec->next != NULL) last_sec->next->prev = first_sec->prev; else link_info.output_bfd->section_last = first_sec->prev; /* Add back. */ last_sec->next = sec->next; if (sec->next != NULL) sec->next->prev = last_sec; else link_info.output_bfd->section_last = last_sec; first_sec->prev = sec; sec->next = first_sec; } } first_os = NULL; last_os = NULL; } ptr = insert_os_after (where); /* Snip everything from the start of the list, up to and including the insert statement we are currently processing. */ first = *start; *start = (*s)->header.next; /* Add them back where they belong, minus the insert. */ *s = *ptr; if (*s == NULL) statement_list.tail = s; *ptr = first; s = start; continue; } s = &(*s)->header.next; } /* Undo constraint twiddling. */ for (os = first_os; os != NULL; os = os->next) { os->constraint = -2 - os->constraint; if (os == last_os) break; } } /* An output section might have been removed after its statement was added. For example, ldemul_before_allocation can remove dynamic sections if they turn out to be not needed. Clean them up here. */ void strip_excluded_output_sections (void) { lang_output_section_statement_type *os; /* Run lang_size_sections (if not already done). */ if (expld.phase != lang_mark_phase_enum) { expld.phase = lang_mark_phase_enum; expld.dataseg.phase = exp_seg_none; one_lang_size_sections_pass (NULL, false); lang_reset_memory_regions (); } for (os = (void *) lang_os_list.head; os != NULL; os = os->next) { asection *output_section; bool exclude; if (os->constraint < 0) continue; output_section = os->bfd_section; if (output_section == NULL) continue; exclude = (output_section->rawsize == 0 && (output_section->flags & SEC_KEEP) == 0 && !bfd_section_removed_from_list (link_info.output_bfd, output_section)); /* Some sections have not yet been sized, notably .gnu.version, .dynsym, .dynstr and .hash. These all have SEC_LINKER_CREATED input sections, so don't drop output sections that have such input sections unless they are also marked SEC_EXCLUDE. */ if (exclude && output_section->map_head.s != NULL) { asection *s; for (s = output_section->map_head.s; s != NULL; s = s->map_head.s) if ((s->flags & SEC_EXCLUDE) == 0 && ((s->flags & SEC_LINKER_CREATED) != 0 || link_info.emitrelocations)) { exclude = false; break; } } if (exclude) { /* We don't set bfd_section to NULL since bfd_section of the removed output section statement may still be used. */ if (!os->update_dot) os->ignored = true; output_section->flags |= SEC_EXCLUDE; bfd_section_list_remove (link_info.output_bfd, output_section); link_info.output_bfd->section_count--; } } } /* Called from ldwrite to clear out asection.map_head and asection.map_tail for use as link_orders in ldwrite. */ void lang_clear_os_map (void) { lang_output_section_statement_type *os; if (map_head_is_link_order) return; for (os = (void *) lang_os_list.head; os != NULL; os = os->next) { asection *output_section; if (os->constraint < 0) continue; output_section = os->bfd_section; if (output_section == NULL) continue; /* TODO: Don't just junk map_head.s, turn them into link_orders. */ output_section->map_head.link_order = NULL; output_section->map_tail.link_order = NULL; } /* Stop future calls to lang_add_section from messing with map_head and map_tail link_order fields. */ map_head_is_link_order = true; } static void print_output_section_statement (lang_output_section_statement_type *output_section_statement) { asection *section = output_section_statement->bfd_section; int len; if (output_section_statement != abs_output_section) { minfo ("\n%s", output_section_statement->name); if (section != NULL) { print_dot = section->vma; len = strlen (output_section_statement->name); if (len >= SECTION_NAME_MAP_LENGTH - 1) { print_nl (); len = 0; } print_spaces (SECTION_NAME_MAP_LENGTH - len); minfo ("0x%V %W", section->vma, TO_ADDR (section->size)); if (section->vma != section->lma) minfo (_(" load address 0x%V"), section->lma); if (output_section_statement->update_dot_tree != NULL) exp_fold_tree (output_section_statement->update_dot_tree, bfd_abs_section_ptr, &print_dot); } print_nl (); } print_statement_list (output_section_statement->children.head, output_section_statement); } static void print_assignment (lang_assignment_statement_type *assignment, lang_output_section_statement_type *output_section) { bool is_dot; etree_type *tree; asection *osec; print_spaces (SECTION_NAME_MAP_LENGTH); if (assignment->exp->type.node_class == etree_assert) { is_dot = false; tree = assignment->exp->assert_s.child; } else { const char *dst = assignment->exp->assign.dst; is_dot = (dst[0] == '.' && dst[1] == 0); tree = assignment->exp; } osec = output_section->bfd_section; if (osec == NULL) osec = bfd_abs_section_ptr; if (assignment->exp->type.node_class != etree_provide) exp_fold_tree (tree, osec, &print_dot); else expld.result.valid_p = false; char buf[32]; const char *str = buf; if (expld.result.valid_p) { bfd_vma value; if (assignment->exp->type.node_class == etree_assert || is_dot || expld.assign_name != NULL) { value = expld.result.value; if (expld.result.section != NULL) value += expld.result.section->vma; buf[0] = '0'; buf[1] = 'x'; bfd_sprintf_vma (link_info.output_bfd, buf + 2, value); if (is_dot) print_dot = value; } else { struct bfd_link_hash_entry *h; h = bfd_link_hash_lookup (link_info.hash, assignment->exp->assign.dst, false, false, true); if (h != NULL && (h->type == bfd_link_hash_defined || h->type == bfd_link_hash_defweak)) { value = h->u.def.value; value += h->u.def.section->output_section->vma; value += h->u.def.section->output_offset; buf[0] = '['; buf[1] = '0'; buf[2] = 'x'; bfd_sprintf_vma (link_info.output_bfd, buf + 3, value); strcat (buf, "]"); } else str = "[unresolved]"; } } else { if (assignment->exp->type.node_class == etree_provide) str = "[!provide]"; else str = "*undef*"; } expld.assign_name = NULL; fprintf (config.map_file, "%-34s", str); exp_print_tree (assignment->exp); print_nl (); } static void print_input_statement (lang_input_statement_type *statm) { if (statm->filename != NULL) fprintf (config.map_file, "LOAD %s\n", statm->filename); } /* Print all symbols defined in a particular section. This is called via bfd_link_hash_traverse, or by print_all_symbols. */ bool print_one_symbol (struct bfd_link_hash_entry *hash_entry, void *ptr) { asection *sec = (asection *) ptr; if ((hash_entry->type == bfd_link_hash_defined || hash_entry->type == bfd_link_hash_defweak) && sec == hash_entry->u.def.section) { print_spaces (SECTION_NAME_MAP_LENGTH); minfo ("0x%V ", (hash_entry->u.def.value + hash_entry->u.def.section->output_offset + hash_entry->u.def.section->output_section->vma)); minfo (" %pT\n", hash_entry->root.string); } return true; } static int hash_entry_addr_cmp (const void *a, const void *b) { const struct bfd_link_hash_entry *l = *(const struct bfd_link_hash_entry **)a; const struct bfd_link_hash_entry *r = *(const struct bfd_link_hash_entry **)b; if (l->u.def.value < r->u.def.value) return -1; else if (l->u.def.value > r->u.def.value) return 1; else return 0; } static void print_all_symbols (asection *sec) { input_section_userdata_type *ud = bfd_section_userdata (sec); struct map_symbol_def *def; struct bfd_link_hash_entry **entries; unsigned int i; if (!ud) return; *ud->map_symbol_def_tail = 0; /* Sort the symbols by address. */ entries = (struct bfd_link_hash_entry **) obstack_alloc (&map_obstack, ud->map_symbol_def_count * sizeof (*entries)); for (i = 0, def = ud->map_symbol_def_head; def; def = def->next, i++) entries[i] = def->entry; qsort (entries, ud->map_symbol_def_count, sizeof (*entries), hash_entry_addr_cmp); /* Print the symbols. */ for (i = 0; i < ud->map_symbol_def_count; i++) ldemul_print_symbol (entries[i], sec); obstack_free (&map_obstack, entries); } /* Print information about an input section to the map file. */ static void print_input_section (asection *i, bool is_discarded) { bfd_size_type size = i->size; int len; bfd_vma addr; init_opb (i); minfo (" %s", i->name); len = 1 + strlen (i->name); if (len >= SECTION_NAME_MAP_LENGTH - 1) { print_nl (); len = 0; } print_spaces (SECTION_NAME_MAP_LENGTH - len); if (i->output_section != NULL && i->output_section->owner == link_info.output_bfd) addr = i->output_section->vma + i->output_offset; else { addr = print_dot; if (!is_discarded) size = 0; } char buf[32]; bfd_sprintf_vma (link_info.output_bfd, buf, addr); minfo ("0x%s %W %pB\n", buf, TO_ADDR (size), i->owner); if (size != i->rawsize && i->rawsize != 0) { len = SECTION_NAME_MAP_LENGTH + 3 + strlen (buf); print_spaces (len); minfo (_("%W (size before relaxing)\n"), TO_ADDR (i->rawsize)); } if (i->output_section != NULL && i->output_section->owner == link_info.output_bfd) { if (link_info.reduce_memory_overheads) bfd_link_hash_traverse (link_info.hash, ldemul_print_symbol, i); else print_all_symbols (i); /* Update print_dot, but make sure that we do not move it backwards - this could happen if we have overlays and a later overlay is shorter than an earier one. */ if (addr + TO_ADDR (size) > print_dot) print_dot = addr + TO_ADDR (size); } } static void print_fill_statement (lang_fill_statement_type *fill) { size_t size; unsigned char *p; fputs (" FILL mask 0x", config.map_file); for (p = fill->fill->data, size = fill->fill->size; size != 0; p++, size--) fprintf (config.map_file, "%02x", *p); fputs ("\n", config.map_file); } static void print_data_statement (lang_data_statement_type *data) { bfd_vma addr; bfd_size_type size; const char *name; init_opb (data->output_section); print_spaces (SECTION_NAME_MAP_LENGTH); addr = data->output_offset; if (data->output_section != NULL) addr += data->output_section->vma; switch (data->type) { default: abort (); case BYTE: size = BYTE_SIZE; name = "BYTE"; break; case SHORT: size = SHORT_SIZE; name = "SHORT"; break; case LONG: size = LONG_SIZE; name = "LONG"; break; case QUAD: size = QUAD_SIZE; name = "QUAD"; break; case SQUAD: size = QUAD_SIZE; name = "SQUAD"; break; } if (size < TO_SIZE ((unsigned) 1)) size = TO_SIZE ((unsigned) 1); minfo ("0x%V %W %s 0x%v", addr, TO_ADDR (size), name, data->value); if (data->exp->type.node_class != etree_value) { print_space (); exp_print_tree (data->exp); } print_nl (); print_dot = addr + TO_ADDR (size); } /* Print an address statement. These are generated by options like -Ttext. */ static void print_address_statement (lang_address_statement_type *address) { minfo (_("Address of section %s set to "), address->section_name); exp_print_tree (address->address); print_nl (); } /* Print a reloc statement. */ static void print_reloc_statement (lang_reloc_statement_type *reloc) { bfd_vma addr; bfd_size_type size; init_opb (reloc->output_section); print_spaces (SECTION_NAME_MAP_LENGTH); addr = reloc->output_offset; if (reloc->output_section != NULL) addr += reloc->output_section->vma; size = bfd_get_reloc_size (reloc->howto); minfo ("0x%V %W RELOC %s ", addr, TO_ADDR (size), reloc->howto->name); if (reloc->name != NULL) minfo ("%s+", reloc->name); else minfo ("%s+", reloc->section->name); exp_print_tree (reloc->addend_exp); print_nl (); print_dot = addr + TO_ADDR (size); } static void print_padding_statement (lang_padding_statement_type *s) { int len; bfd_vma addr; init_opb (s->output_section); minfo (" *fill*"); len = sizeof " *fill*" - 1; print_spaces (SECTION_NAME_MAP_LENGTH - len); addr = s->output_offset; if (s->output_section != NULL) addr += s->output_section->vma; minfo ("0x%V %W ", addr, TO_ADDR (s->size)); if (s->fill->size != 0) { size_t size; unsigned char *p; for (p = s->fill->data, size = s->fill->size; size != 0; p++, size--) fprintf (config.map_file, "%02x", *p); } print_nl (); print_dot = addr + TO_ADDR (s->size); } static void print_wild_statement (lang_wild_statement_type *w, lang_output_section_statement_type *os) { struct wildcard_list *sec; print_space (); if (w->exclude_name_list) { name_list *tmp; minfo ("EXCLUDE_FILE(%s", w->exclude_name_list->name); for (tmp = w->exclude_name_list->next; tmp; tmp = tmp->next) minfo (" %s", tmp->name); minfo (") "); } if (w->filenames_sorted) minfo ("SORT_BY_NAME("); if (w->filename != NULL) minfo ("%s", w->filename); else minfo ("*"); if (w->filenames_sorted) minfo (")"); minfo ("("); for (sec = w->section_list; sec; sec = sec->next) { int closing_paren = 0; switch (sec->spec.sorted) { case none: break; case by_name: minfo ("SORT_BY_NAME("); closing_paren = 1; break; case by_alignment: minfo ("SORT_BY_ALIGNMENT("); closing_paren = 1; break; case by_name_alignment: minfo ("SORT_BY_NAME(SORT_BY_ALIGNMENT("); closing_paren = 2; break; case by_alignment_name: minfo ("SORT_BY_ALIGNMENT(SORT_BY_NAME("); closing_paren = 2; break; case by_none: minfo ("SORT_NONE("); closing_paren = 1; break; case by_init_priority: minfo ("SORT_BY_INIT_PRIORITY("); closing_paren = 1; break; } if (sec->spec.exclude_name_list != NULL) { name_list *tmp; minfo ("EXCLUDE_FILE(%s", sec->spec.exclude_name_list->name); for (tmp = sec->spec.exclude_name_list->next; tmp; tmp = tmp->next) minfo (" %s", tmp->name); minfo (") "); } if (sec->spec.name != NULL) minfo ("%s", sec->spec.name); else minfo ("*"); for (;closing_paren > 0; closing_paren--) minfo (")"); if (sec->next) minfo (" "); } minfo (")"); print_nl (); print_statement_list (w->children.head, os); } /* Print a group statement. */ static void print_group (lang_group_statement_type *s, lang_output_section_statement_type *os) { fprintf (config.map_file, "START GROUP\n"); print_statement_list (s->children.head, os); fprintf (config.map_file, "END GROUP\n"); } /* Print the list of statements in S. This can be called for any statement type. */ static void print_statement_list (lang_statement_union_type *s, lang_output_section_statement_type *os) { while (s != NULL) { print_statement (s, os); s = s->header.next; } } /* Print the first statement in statement list S. This can be called for any statement type. */ static void print_statement (lang_statement_union_type *s, lang_output_section_statement_type *os) { switch (s->header.type) { default: fprintf (config.map_file, _("Fail with %d\n"), s->header.type); FAIL (); break; case lang_constructors_statement_enum: if (constructor_list.head != NULL) { if (constructors_sorted) minfo (" SORT (CONSTRUCTORS)\n"); else minfo (" CONSTRUCTORS\n"); print_statement_list (constructor_list.head, os); } break; case lang_wild_statement_enum: print_wild_statement (&s->wild_statement, os); break; case lang_address_statement_enum: print_address_statement (&s->address_statement); break; case lang_object_symbols_statement_enum: minfo (" CREATE_OBJECT_SYMBOLS\n"); break; case lang_fill_statement_enum: print_fill_statement (&s->fill_statement); break; case lang_data_statement_enum: print_data_statement (&s->data_statement); break; case lang_reloc_statement_enum: print_reloc_statement (&s->reloc_statement); break; case lang_input_section_enum: print_input_section (s->input_section.section, false); break; case lang_padding_statement_enum: print_padding_statement (&s->padding_statement); break; case lang_output_section_statement_enum: print_output_section_statement (&s->output_section_statement); break; case lang_assignment_statement_enum: print_assignment (&s->assignment_statement, os); break; case lang_target_statement_enum: fprintf (config.map_file, "TARGET(%s)\n", s->target_statement.target); break; case lang_output_statement_enum: minfo ("OUTPUT(%s", s->output_statement.name); if (output_target != NULL) minfo (" %s", output_target); minfo (")\n"); break; case lang_input_statement_enum: print_input_statement (&s->input_statement); break; case lang_group_statement_enum: print_group (&s->group_statement, os); break; case lang_insert_statement_enum: minfo ("INSERT %s %s\n", s->insert_statement.is_before ? "BEFORE" : "AFTER", s->insert_statement.where); break; } } static void print_statements (void) { print_statement_list (statement_list.head, abs_output_section); } /* Print the first N statements in statement list S to STDERR. If N == 0, nothing is printed. If N < 0, the entire list is printed. Intended to be called from GDB. */ void dprint_statement (lang_statement_union_type *s, int n) { FILE *map_save = config.map_file; config.map_file = stderr; if (n < 0) print_statement_list (s, abs_output_section); else { while (s && --n >= 0) { print_statement (s, abs_output_section); s = s->header.next; } } config.map_file = map_save; } static void insert_pad (lang_statement_union_type **ptr, fill_type *fill, bfd_size_type alignment_needed, asection *output_section, bfd_vma dot) { static fill_type zero_fill; lang_statement_union_type *pad = NULL; if (ptr != &statement_list.head) pad = ((lang_statement_union_type *) ((char *) ptr - offsetof (lang_statement_union_type, header.next))); if (pad != NULL && pad->header.type == lang_padding_statement_enum && pad->padding_statement.output_section == output_section) { /* Use the existing pad statement. */ } else if ((pad = *ptr) != NULL && pad->header.type == lang_padding_statement_enum && pad->padding_statement.output_section == output_section) { /* Use the existing pad statement. */ } else { /* Make a new padding statement, linked into existing chain. */ pad = stat_alloc (sizeof (lang_padding_statement_type)); pad->header.next = *ptr; *ptr = pad; pad->header.type = lang_padding_statement_enum; pad->padding_statement.output_section = output_section; if (fill == NULL) fill = &zero_fill; pad->padding_statement.fill = fill; } pad->padding_statement.output_offset = dot - output_section->vma; pad->padding_statement.size = alignment_needed; if (!(output_section->flags & SEC_FIXED_SIZE)) output_section->size = TO_SIZE (dot + TO_ADDR (alignment_needed) - output_section->vma); } /* Work out how much this section will move the dot point. */ static bfd_vma size_input_section (lang_statement_union_type **this_ptr, lang_output_section_statement_type *output_section_statement, fill_type *fill, bool *removed, bfd_vma dot) { lang_input_section_type *is = &((*this_ptr)->input_section); asection *i = is->section; asection *o = output_section_statement->bfd_section; *removed = 0; if (link_info.non_contiguous_regions) { /* If the input section I has already been successfully assigned to an output section other than O, don't bother with it and let the caller remove it from the list. Keep processing in case we have already handled O, because the repeated passes have reinitialized its size. */ if (i->already_assigned && i->already_assigned != o) { *removed = 1; return dot; } } if (i->sec_info_type == SEC_INFO_TYPE_JUST_SYMS) i->output_offset = i->vma - o->vma; else if (((i->flags & SEC_EXCLUDE) != 0) || output_section_statement->ignored) i->output_offset = dot - o->vma; else { bfd_size_type alignment_needed; /* Align this section first to the input sections requirement, then to the output section's requirement. If this alignment is greater than any seen before, then record it too. Perform the alignment by inserting a magic 'padding' statement. */ if (output_section_statement->subsection_alignment != NULL) i->alignment_power = exp_get_power (output_section_statement->subsection_alignment, "subsection alignment"); if (o->alignment_power < i->alignment_power) o->alignment_power = i->alignment_power; alignment_needed = align_power (dot, i->alignment_power) - dot; if (alignment_needed != 0) { insert_pad (this_ptr, fill, TO_SIZE (alignment_needed), o, dot); dot += alignment_needed; } if (link_info.non_contiguous_regions) { /* If I would overflow O, let the caller remove I from the list. */ if (output_section_statement->region) { bfd_vma end = output_section_statement->region->origin + output_section_statement->region->length; if (dot + TO_ADDR (i->size) > end) { if (i->flags & SEC_LINKER_CREATED) einfo (_("%F%P: Output section `%pA' not large enough for " "the linker-created stubs section `%pA'.\n"), i->output_section, i); if (i->rawsize && i->rawsize != i->size) einfo (_("%F%P: Relaxation not supported with " "--enable-non-contiguous-regions (section `%pA' " "would overflow `%pA' after it changed size).\n"), i, i->output_section); *removed = 1; dot = end; i->output_section = NULL; return dot; } } } /* Remember where in the output section this input section goes. */ i->output_offset = dot - o->vma; /* Mark how big the output section must be to contain this now. */ dot += TO_ADDR (i->size); if (!(o->flags & SEC_FIXED_SIZE)) o->size = TO_SIZE (dot - o->vma); if (link_info.non_contiguous_regions) { /* Record that I was successfully assigned to O, and update its actual output section too. */ i->already_assigned = o; i->output_section = o; } } return dot; } struct check_sec { asection *sec; bool warned; }; static int sort_sections_by_lma (const void *arg1, const void *arg2) { const asection *sec1 = ((const struct check_sec *) arg1)->sec; const asection *sec2 = ((const struct check_sec *) arg2)->sec; if (sec1->lma < sec2->lma) return -1; else if (sec1->lma > sec2->lma) return 1; else if (sec1->id < sec2->id) return -1; else if (sec1->id > sec2->id) return 1; return 0; } static int sort_sections_by_vma (const void *arg1, const void *arg2) { const asection *sec1 = ((const struct check_sec *) arg1)->sec; const asection *sec2 = ((const struct check_sec *) arg2)->sec; if (sec1->vma < sec2->vma) return -1; else if (sec1->vma > sec2->vma) return 1; else if (sec1->id < sec2->id) return -1; else if (sec1->id > sec2->id) return 1; return 0; } #define IS_TBSS(s) \ ((s->flags & (SEC_LOAD | SEC_THREAD_LOCAL)) == SEC_THREAD_LOCAL) #define IGNORE_SECTION(s) \ ((s->flags & SEC_ALLOC) == 0 || IS_TBSS (s)) /* Check to see if any allocated sections overlap with other allocated sections. This can happen if a linker script specifies the output section addresses of the two sections. Also check whether any memory region has overflowed. */ static void lang_check_section_addresses (void) { asection *s, *p; struct check_sec *sections; size_t i, count; bfd_vma addr_mask; bfd_vma s_start; bfd_vma s_end; bfd_vma p_start = 0; bfd_vma p_end = 0; lang_memory_region_type *m; bool overlays; /* Detect address space overflow on allocated sections. */ addr_mask = ((bfd_vma) 1 << (bfd_arch_bits_per_address (link_info.output_bfd) - 1)) - 1; addr_mask = (addr_mask << 1) + 1; for (s = link_info.output_bfd->sections; s != NULL; s = s->next) if ((s->flags & SEC_ALLOC) != 0) { s_end = (s->vma + s->size) & addr_mask; if (s_end != 0 && s_end < (s->vma & addr_mask)) einfo (_("%X%P: section %s VMA wraps around address space\n"), s->name); else { s_end = (s->lma + s->size) & addr_mask; if (s_end != 0 && s_end < (s->lma & addr_mask)) einfo (_("%X%P: section %s LMA wraps around address space\n"), s->name); } } if (bfd_count_sections (link_info.output_bfd) <= 1) return; count = bfd_count_sections (link_info.output_bfd); sections = XNEWVEC (struct check_sec, count); /* Scan all sections in the output list. */ count = 0; for (s = link_info.output_bfd->sections; s != NULL; s = s->next) { if (IGNORE_SECTION (s) || s->size == 0) continue; sections[count].sec = s; sections[count].warned = false; count++; } if (count <= 1) { free (sections); return; } qsort (sections, count, sizeof (*sections), sort_sections_by_lma); /* First check section LMAs. There should be no overlap of LMAs on loadable sections, even with overlays. */ for (p = NULL, i = 0; i < count; i++) { s = sections[i].sec; init_opb (s); if ((s->flags & SEC_LOAD) != 0) { s_start = s->lma; s_end = s_start + TO_ADDR (s->size) - 1; /* Look for an overlap. We have sorted sections by lma, so we know that s_start >= p_start. Besides the obvious case of overlap when the current section starts before the previous one ends, we also must have overlap if the previous section wraps around the address space. */ if (p != NULL && (s_start <= p_end || p_end < p_start)) { einfo (_("%X%P: section %s LMA [%V,%V]" " overlaps section %s LMA [%V,%V]\n"), s->name, s_start, s_end, p->name, p_start, p_end); sections[i].warned = true; } p = s; p_start = s_start; p_end = s_end; } } /* If any non-zero size allocated section (excluding tbss) starts at exactly the same VMA as another such section, then we have overlays. Overlays generated by the OVERLAY keyword will have this property. It is possible to intentionally generate overlays that fail this test, but it would be unusual. */ qsort (sections, count, sizeof (*sections), sort_sections_by_vma); overlays = false; p_start = sections[0].sec->vma; for (i = 1; i < count; i++) { s_start = sections[i].sec->vma; if (p_start == s_start) { overlays = true; break; } p_start = s_start; } /* Now check section VMAs if no overlays were detected. */ if (!overlays) { for (p = NULL, i = 0; i < count; i++) { s = sections[i].sec; init_opb (s); s_start = s->vma; s_end = s_start + TO_ADDR (s->size) - 1; if (p != NULL && !sections[i].warned && (s_start <= p_end || p_end < p_start)) einfo (_("%X%P: section %s VMA [%V,%V]" " overlaps section %s VMA [%V,%V]\n"), s->name, s_start, s_end, p->name, p_start, p_end); p = s; p_start = s_start; p_end = s_end; } } free (sections); /* If any memory region has overflowed, report by how much. We do not issue this diagnostic for regions that had sections explicitly placed outside their bounds; os_region_check's diagnostics are adequate for that case. FIXME: It is conceivable that m->current - (m->origin + m->length) might overflow a 32-bit integer. There is, alas, no way to print a bfd_vma quantity in decimal. */ for (m = lang_memory_region_list; m; m = m->next) if (m->had_full_message) { unsigned long over = m->current - (m->origin + m->length); einfo (ngettext ("%X%P: region `%s' overflowed by %lu byte\n", "%X%P: region `%s' overflowed by %lu bytes\n", over), m->name_list.name, over); } } /* Make sure the new address is within the region. We explicitly permit the current address to be at the exact end of the region when the address is non-zero, in case the region is at the end of addressable memory and the calculation wraps around. */ static void os_region_check (lang_output_section_statement_type *os, lang_memory_region_type *region, etree_type *tree, bfd_vma rbase) { if ((region->current < region->origin || (region->current - region->origin > region->length)) && ((region->current != region->origin + region->length) || rbase == 0)) { if (tree != NULL) { einfo (_("%X%P: address 0x%v of %pB section `%s'" " is not within region `%s'\n"), region->current, os->bfd_section->owner, os->bfd_section->name, region->name_list.name); } else if (!region->had_full_message) { region->had_full_message = true; einfo (_("%X%P: %pB section `%s' will not fit in region `%s'\n"), os->bfd_section->owner, os->bfd_section->name, region->name_list.name); } } } static void ldlang_check_relro_region (lang_statement_union_type *s) { seg_align_type *seg = &expld.dataseg; if (seg->relro == exp_seg_relro_start) { if (!seg->relro_start_stat) seg->relro_start_stat = s; else { ASSERT (seg->relro_start_stat == s); } } else if (seg->relro == exp_seg_relro_end) { if (!seg->relro_end_stat) seg->relro_end_stat = s; else { ASSERT (seg->relro_end_stat == s); } } } /* Set the sizes for all the output sections. */ static bfd_vma lang_size_sections_1 (lang_statement_union_type **prev, lang_output_section_statement_type *output_section_statement, fill_type *fill, bfd_vma dot, bool *relax, bool check_regions) { lang_statement_union_type *s; lang_statement_union_type *prev_s = NULL; bool removed_prev_s = false; /* Size up the sections from their constituent parts. */ for (s = *prev; s != NULL; prev_s = s, s = s->header.next) { bool removed = false; switch (s->header.type) { case lang_output_section_statement_enum: { bfd_vma newdot, after, dotdelta; lang_output_section_statement_type *os; lang_memory_region_type *r; int section_alignment = 0; os = &s->output_section_statement; init_opb (os->bfd_section); if (os->constraint == -1) break; /* FIXME: We shouldn't need to zero section vmas for ld -r here, in lang_insert_orphan, or in the default linker scripts. This is covering for coff backend linker bugs. See PR6945. */ if (os->addr_tree == NULL && bfd_link_relocatable (&link_info) && (bfd_get_flavour (link_info.output_bfd) == bfd_target_coff_flavour)) os->addr_tree = exp_intop (0); if (os->addr_tree != NULL) { exp_fold_tree (os->addr_tree, bfd_abs_section_ptr, &dot); if (expld.result.valid_p) { dot = expld.result.value; if (expld.result.section != NULL) dot += expld.result.section->vma; } else if (expld.phase != lang_mark_phase_enum) einfo (_("%F%P:%pS: non constant or forward reference" " address expression for section %s\n"), os->addr_tree, os->name); } if (os->bfd_section == NULL) /* This section was removed or never actually created. */ break; /* If this is a COFF shared library section, use the size and address from the input section. FIXME: This is COFF specific; it would be cleaner if there were some other way to do this, but nothing simple comes to mind. */ if (((bfd_get_flavour (link_info.output_bfd) == bfd_target_ecoff_flavour) || (bfd_get_flavour (link_info.output_bfd) == bfd_target_coff_flavour)) && (os->bfd_section->flags & SEC_COFF_SHARED_LIBRARY) != 0) { asection *input; if (os->children.head == NULL || os->children.head->header.next != NULL || (os->children.head->header.type != lang_input_section_enum)) einfo (_("%X%P: internal error on COFF shared library" " section %s\n"), os->name); input = os->children.head->input_section.section; bfd_set_section_vma (os->bfd_section, bfd_section_vma (input)); if (!(os->bfd_section->flags & SEC_FIXED_SIZE)) os->bfd_section->size = input->size; break; } newdot = dot; dotdelta = 0; if (bfd_is_abs_section (os->bfd_section)) { /* No matter what happens, an abs section starts at zero. */ ASSERT (os->bfd_section->vma == 0); } else { if (os->addr_tree == NULL) { /* No address specified for this section, get one from the region specification. */ if (os->region == NULL || ((os->bfd_section->flags & (SEC_ALLOC | SEC_LOAD)) && os->region->name_list.name[0] == '*' && strcmp (os->region->name_list.name, DEFAULT_MEMORY_REGION) == 0)) { os->region = lang_memory_default (os->bfd_section); } /* If a loadable section is using the default memory region, and some non default memory regions were defined, issue an error message. */ if (!os->ignored && !IGNORE_SECTION (os->bfd_section) && !bfd_link_relocatable (&link_info) && check_regions && strcmp (os->region->name_list.name, DEFAULT_MEMORY_REGION) == 0 && lang_memory_region_list != NULL && (strcmp (lang_memory_region_list->name_list.name, DEFAULT_MEMORY_REGION) != 0 || lang_memory_region_list->next != NULL) && lang_sizing_iteration == 1) { /* By default this is an error rather than just a warning because if we allocate the section to the default memory region we can end up creating an excessively large binary, or even seg faulting when attempting to perform a negative seek. See sources.redhat.com/ml/binutils/2003-04/msg00423.html for an example of this. This behaviour can be overridden by the using the --no-check-sections switch. */ if (command_line.check_section_addresses) einfo (_("%F%P: error: no memory region specified" " for loadable section `%s'\n"), bfd_section_name (os->bfd_section)); else einfo (_("%P: warning: no memory region specified" " for loadable section `%s'\n"), bfd_section_name (os->bfd_section)); } newdot = os->region->current; section_alignment = os->bfd_section->alignment_power; } else section_alignment = exp_get_power (os->section_alignment, "section alignment"); /* Align to what the section needs. */ if (section_alignment > 0) { bfd_vma savedot = newdot; bfd_vma diff = 0; newdot = align_power (newdot, section_alignment); dotdelta = newdot - savedot; if (lang_sizing_iteration == 1) diff = dotdelta; else if (lang_sizing_iteration > 1) { /* Only report adjustments that would change alignment from what we have already reported. */ diff = newdot - os->bfd_section->vma; if (!(diff & (((bfd_vma) 1 << section_alignment) - 1))) diff = 0; } if (diff != 0 && (config.warn_section_align || os->addr_tree != NULL)) einfo (_("%P: warning: " "start of section %s changed by %ld\n"), os->name, (long) diff); } bfd_set_section_vma (os->bfd_section, newdot); os->bfd_section->output_offset = 0; } lang_size_sections_1 (&os->children.head, os, os->fill, newdot, relax, check_regions); os->processed_vma = true; if (bfd_is_abs_section (os->bfd_section) || os->ignored) /* Except for some special linker created sections, no output section should change from zero size after strip_excluded_output_sections. A non-zero size on an ignored section indicates that some input section was not sized early enough. */ ASSERT (os->bfd_section->size == 0); else { dot = os->bfd_section->vma; /* Put the section within the requested block size, or align at the block boundary. */ after = ((dot + TO_ADDR (os->bfd_section->size) + os->block_value - 1) & - (bfd_vma) os->block_value); if (!(os->bfd_section->flags & SEC_FIXED_SIZE)) os->bfd_section->size = TO_SIZE (after - os->bfd_section->vma); } /* Set section lma. */ r = os->region; if (r == NULL) r = lang_memory_region_lookup (DEFAULT_MEMORY_REGION, false); if (os->load_base) { bfd_vma lma = exp_get_abs_int (os->load_base, 0, "load base"); os->bfd_section->lma = lma; } else if (os->lma_region != NULL) { bfd_vma lma = os->lma_region->current; if (os->align_lma_with_input) lma += dotdelta; else { /* When LMA_REGION is the same as REGION, align the LMA as we did for the VMA, possibly including alignment from the bfd section. If a different region, then only align according to the value in the output statement. */ if (os->lma_region != os->region) section_alignment = exp_get_power (os->section_alignment, "section alignment"); if (section_alignment > 0) lma = align_power (lma, section_alignment); } os->bfd_section->lma = lma; } else if (r->last_os != NULL && (os->bfd_section->flags & SEC_ALLOC) != 0) { bfd_vma lma; asection *last; last = r->last_os->output_section_statement.bfd_section; /* A backwards move of dot should be accompanied by an explicit assignment to the section LMA (ie. os->load_base set) because backwards moves can create overlapping LMAs. */ if (dot < last->vma && os->bfd_section->size != 0 && dot + TO_ADDR (os->bfd_section->size) <= last->vma) { /* If dot moved backwards then leave lma equal to vma. This is the old default lma, which might just happen to work when the backwards move is sufficiently large. Nag if this changes anything, so people can fix their linker scripts. */ if (last->vma != last->lma) einfo (_("%P: warning: dot moved backwards " "before `%s'\n"), os->name); } else { /* If this is an overlay, set the current lma to that at the end of the previous section. */ if (os->sectype == overlay_section) lma = last->lma + TO_ADDR (last->size); /* Otherwise, keep the same lma to vma relationship as the previous section. */ else lma = os->bfd_section->vma + last->lma - last->vma; if (section_alignment > 0) lma = align_power (lma, section_alignment); os->bfd_section->lma = lma; } } os->processed_lma = true; /* Keep track of normal sections using the default lma region. We use this to set the lma for following sections. Overlays or other linker script assignment to lma might mean that the default lma == vma is incorrect. To avoid warnings about dot moving backwards when using -Ttext, don't start tracking sections until we find one of non-zero size or with lma set differently to vma. Do this tracking before we short-cut the loop so that we track changes for the case where the section size is zero, but the lma is set differently to the vma. This is important, if an orphan section is placed after an otherwise empty output section that has an explicit lma set, we want that lma reflected in the orphans lma. */ if (((!IGNORE_SECTION (os->bfd_section) && (os->bfd_section->size != 0 || (r->last_os == NULL && os->bfd_section->vma != os->bfd_section->lma) || (r->last_os != NULL && dot >= (r->last_os->output_section_statement .bfd_section->vma)))) || os->sectype == first_overlay_section) && os->lma_region == NULL && !bfd_link_relocatable (&link_info)) r->last_os = s; if (bfd_is_abs_section (os->bfd_section) || os->ignored) break; /* .tbss sections effectively have zero size. */ if (!IS_TBSS (os->bfd_section) || bfd_link_relocatable (&link_info)) dotdelta = TO_ADDR (os->bfd_section->size); else dotdelta = 0; dot += dotdelta; if (os->update_dot_tree != 0) exp_fold_tree (os->update_dot_tree, bfd_abs_section_ptr, &dot); /* Update dot in the region ? We only do this if the section is going to be allocated, since unallocated sections do not contribute to the region's overall size in memory. */ if (os->region != NULL && (os->bfd_section->flags & (SEC_ALLOC | SEC_LOAD))) { os->region->current = dot; if (check_regions) /* Make sure the new address is within the region. */ os_region_check (os, os->region, os->addr_tree, os->bfd_section->vma); if (os->lma_region != NULL && os->lma_region != os->region && ((os->bfd_section->flags & SEC_LOAD) || os->align_lma_with_input)) { os->lma_region->current = os->bfd_section->lma + dotdelta; if (check_regions) os_region_check (os, os->lma_region, NULL, os->bfd_section->lma); } } } break; case lang_constructors_statement_enum: dot = lang_size_sections_1 (&constructor_list.head, output_section_statement, fill, dot, relax, check_regions); break; case lang_data_statement_enum: { unsigned int size = 0; s->data_statement.output_offset = dot - output_section_statement->bfd_section->vma; s->data_statement.output_section = output_section_statement->bfd_section; /* We might refer to provided symbols in the expression, and need to mark them as needed. */ exp_fold_tree (s->data_statement.exp, bfd_abs_section_ptr, &dot); switch (s->data_statement.type) { default: abort (); case QUAD: case SQUAD: size = QUAD_SIZE; break; case LONG: size = LONG_SIZE; break; case SHORT: size = SHORT_SIZE; break; case BYTE: size = BYTE_SIZE; break; } if (size < TO_SIZE ((unsigned) 1)) size = TO_SIZE ((unsigned) 1); dot += TO_ADDR (size); if (!(output_section_statement->bfd_section->flags & SEC_FIXED_SIZE)) output_section_statement->bfd_section->size = TO_SIZE (dot - output_section_statement->bfd_section->vma); } break; case lang_reloc_statement_enum: { int size; s->reloc_statement.output_offset = dot - output_section_statement->bfd_section->vma; s->reloc_statement.output_section = output_section_statement->bfd_section; size = bfd_get_reloc_size (s->reloc_statement.howto); dot += TO_ADDR (size); if (!(output_section_statement->bfd_section->flags & SEC_FIXED_SIZE)) output_section_statement->bfd_section->size = TO_SIZE (dot - output_section_statement->bfd_section->vma); } break; case lang_wild_statement_enum: dot = lang_size_sections_1 (&s->wild_statement.children.head, output_section_statement, fill, dot, relax, check_regions); break; case lang_object_symbols_statement_enum: link_info.create_object_symbols_section = output_section_statement->bfd_section; output_section_statement->bfd_section->flags |= SEC_KEEP; break; case lang_output_statement_enum: case lang_target_statement_enum: break; case lang_input_section_enum: { asection *i; i = s->input_section.section; if (relax) { bool again; if (!bfd_relax_section (i->owner, i, &link_info, &again)) einfo (_("%F%P: can't relax section: %E\n")); if (again) *relax = true; } dot = size_input_section (prev, output_section_statement, fill, &removed, dot); } break; case lang_input_statement_enum: break; case lang_fill_statement_enum: s->fill_statement.output_section = output_section_statement->bfd_section; fill = s->fill_statement.fill; break; case lang_assignment_statement_enum: { bfd_vma newdot = dot; etree_type *tree = s->assignment_statement.exp; expld.dataseg.relro = exp_seg_relro_none; exp_fold_tree (tree, output_section_statement->bfd_section, &newdot); ldlang_check_relro_region (s); expld.dataseg.relro = exp_seg_relro_none; /* This symbol may be relative to this section. */ if ((tree->type.node_class == etree_provided || tree->type.node_class == etree_assign) && (tree->assign.dst [0] != '.' || tree->assign.dst [1] != '\0')) output_section_statement->update_dot = 1; if (!output_section_statement->ignored) { if (output_section_statement == abs_output_section) { /* If we don't have an output section, then just adjust the default memory address. */ lang_memory_region_lookup (DEFAULT_MEMORY_REGION, false)->current = newdot; } else if (newdot != dot) { /* Insert a pad after this statement. We can't put the pad before when relaxing, in case the assignment references dot. */ insert_pad (&s->header.next, fill, TO_SIZE (newdot - dot), output_section_statement->bfd_section, dot); /* Don't neuter the pad below when relaxing. */ s = s->header.next; /* If dot is advanced, this implies that the section should have space allocated to it, unless the user has explicitly stated that the section should not be allocated. */ if (output_section_statement->sectype != noalloc_section && (output_section_statement->sectype != noload_section || (bfd_get_flavour (link_info.output_bfd) == bfd_target_elf_flavour))) output_section_statement->bfd_section->flags |= SEC_ALLOC; } dot = newdot; } } break; case lang_padding_statement_enum: /* If this is the first time lang_size_sections is called, we won't have any padding statements. If this is the second or later passes when relaxing, we should allow padding to shrink. If padding is needed on this pass, it will be added back in. */ s->padding_statement.size = 0; /* Make sure output_offset is valid. If relaxation shrinks the section and this pad isn't needed, it's possible to have output_offset larger than the final size of the section. bfd_set_section_contents will complain even for a pad size of zero. */ s->padding_statement.output_offset = dot - output_section_statement->bfd_section->vma; break; case lang_group_statement_enum: dot = lang_size_sections_1 (&s->group_statement.children.head, output_section_statement, fill, dot, relax, check_regions); break; case lang_insert_statement_enum: break; /* We can only get here when relaxing is turned on. */ case lang_address_statement_enum: break; default: FAIL (); break; } /* If an input section doesn't fit in the current output section, remove it from the list. Handle the case where we have to remove an input_section statement here: there is a special case to remove the first element of the list. */ if (link_info.non_contiguous_regions && removed) { /* If we removed the first element during the previous iteration, override the loop assignment of prev_s. */ if (removed_prev_s) prev_s = NULL; if (prev_s) { /* If there was a real previous input section, just skip the current one. */ prev_s->header.next=s->header.next; s = prev_s; removed_prev_s = false; } else { /* Remove the first input section of the list. */ *prev = s->header.next; removed_prev_s = true; } /* Move to next element, unless we removed the head of the list. */ if (!removed_prev_s) prev = &s->header.next; } else { prev = &s->header.next; removed_prev_s = false; } } return dot; } /* Callback routine that is used in _bfd_elf_map_sections_to_segments. The BFD library has set NEW_SEGMENT to TRUE iff it thinks that CURRENT_SECTION and PREVIOUS_SECTION ought to be placed into different segments. We are allowed an opportunity to override this decision. */ bool ldlang_override_segment_assignment (struct bfd_link_info *info ATTRIBUTE_UNUSED, bfd *abfd ATTRIBUTE_UNUSED, asection *current_section, asection *previous_section, bool new_segment) { lang_output_section_statement_type *cur; lang_output_section_statement_type *prev; /* The checks below are only necessary when the BFD library has decided that the two sections ought to be placed into the same segment. */ if (new_segment) return true; /* Paranoia checks. */ if (current_section == NULL || previous_section == NULL) return new_segment; /* If this flag is set, the target never wants code and non-code sections comingled in the same segment. */ if (config.separate_code && ((current_section->flags ^ previous_section->flags) & SEC_CODE)) return true; /* Find the memory regions associated with the two sections. We call lang_output_section_find() here rather than scanning the list of output sections looking for a matching section pointer because if we have a large number of sections then a hash lookup is faster. */ cur = lang_output_section_find (current_section->name); prev = lang_output_section_find (previous_section->name); /* More paranoia. */ if (cur == NULL || prev == NULL) return new_segment; /* If the regions are different then force the sections to live in different segments. See the email thread starting at the following URL for the reasons why this is necessary: http://sourceware.org/ml/binutils/2007-02/msg00216.html */ return cur->region != prev->region; } void one_lang_size_sections_pass (bool *relax, bool check_regions) { lang_statement_iteration++; if (expld.phase != lang_mark_phase_enum) lang_sizing_iteration++; lang_size_sections_1 (&statement_list.head, abs_output_section, 0, 0, relax, check_regions); } static bool lang_size_segment (void) { /* If XXX_SEGMENT_ALIGN XXX_SEGMENT_END pair was seen, check whether a page could be saved in the data segment. */ seg_align_type *seg = &expld.dataseg; bfd_vma first, last; first = -seg->base & (seg->commonpagesize - 1); last = seg->end & (seg->commonpagesize - 1); if (first && last && ((seg->base & ~(seg->commonpagesize - 1)) != (seg->end & ~(seg->commonpagesize - 1))) && first + last <= seg->commonpagesize) { seg->phase = exp_seg_adjust; return true; } seg->phase = exp_seg_done; return false; } static bfd_vma lang_size_relro_segment_1 (void) { seg_align_type *seg = &expld.dataseg; bfd_vma relro_end, desired_end; asection *sec; /* Compute the expected PT_GNU_RELRO/PT_LOAD segment end. */ relro_end = (seg->relro_end + seg->relropagesize - 1) & -seg->relropagesize; /* Adjust by the offset arg of XXX_SEGMENT_RELRO_END. */ desired_end = relro_end - seg->relro_offset; /* For sections in the relro segment.. */ for (sec = link_info.output_bfd->section_last; sec; sec = sec->prev) if ((sec->flags & SEC_ALLOC) != 0 && sec->vma >= seg->base && sec->vma < seg->relro_end - seg->relro_offset) { /* Where do we want to put this section so that it ends as desired? */ bfd_vma start, end, bump; end = start = sec->vma; if (!IS_TBSS (sec)) end += TO_ADDR (sec->size); bump = desired_end - end; /* We'd like to increase START by BUMP, but we must heed alignment so the increase might be less than optimum. */ start += bump; start &= ~(((bfd_vma) 1 << sec->alignment_power) - 1); /* This is now the desired end for the previous section. */ desired_end = start; } seg->phase = exp_seg_relro_adjust; ASSERT (desired_end >= seg->base); seg->base = desired_end; return relro_end; } static bool lang_size_relro_segment (bool *relax, bool check_regions) { bool do_reset = false; if (link_info.relro && expld.dataseg.relro_end) { bfd_vma data_initial_base = expld.dataseg.base; bfd_vma data_relro_end = lang_size_relro_segment_1 (); lang_reset_memory_regions (); one_lang_size_sections_pass (relax, check_regions); /* Assignments to dot, or to output section address in a user script have increased padding over the original. Revert. */ if (expld.dataseg.relro_end > data_relro_end) { expld.dataseg.base = data_initial_base; do_reset = true; } } else if (lang_size_segment ()) do_reset = true; return do_reset; } void lang_size_sections (bool *relax, bool check_regions) { expld.phase = lang_allocating_phase_enum; expld.dataseg.phase = exp_seg_none; one_lang_size_sections_pass (relax, check_regions); if (expld.dataseg.phase != exp_seg_end_seen) expld.dataseg.phase = exp_seg_done; if (expld.dataseg.phase == exp_seg_end_seen) { bool do_reset = lang_size_relro_segment (relax, check_regions); if (do_reset) { lang_reset_memory_regions (); one_lang_size_sections_pass (relax, check_regions); } if (link_info.relro && expld.dataseg.relro_end) { link_info.relro_start = expld.dataseg.base; link_info.relro_end = expld.dataseg.relro_end; } } } static lang_output_section_statement_type *current_section; static lang_assignment_statement_type *current_assign; static bool prefer_next_section; /* Worker function for lang_do_assignments. Recursiveness goes here. */ static bfd_vma lang_do_assignments_1 (lang_statement_union_type *s, lang_output_section_statement_type *current_os, fill_type *fill, bfd_vma dot, bool *found_end) { for (; s != NULL; s = s->header.next) { switch (s->header.type) { case lang_constructors_statement_enum: dot = lang_do_assignments_1 (constructor_list.head, current_os, fill, dot, found_end); break; case lang_output_section_statement_enum: { lang_output_section_statement_type *os; bfd_vma newdot; os = &(s->output_section_statement); os->after_end = *found_end; init_opb (os->bfd_section); newdot = dot; if (os->bfd_section != NULL) { if (!os->ignored && (os->bfd_section->flags & SEC_ALLOC) != 0) { current_section = os; prefer_next_section = false; } newdot = os->bfd_section->vma; } newdot = lang_do_assignments_1 (os->children.head, os, os->fill, newdot, found_end); if (!os->ignored) { if (os->bfd_section != NULL) { newdot = os->bfd_section->vma; /* .tbss sections effectively have zero size. */ if (!IS_TBSS (os->bfd_section) || bfd_link_relocatable (&link_info)) newdot += TO_ADDR (os->bfd_section->size); if (os->update_dot_tree != NULL) exp_fold_tree (os->update_dot_tree, bfd_abs_section_ptr, &newdot); } dot = newdot; } } break; case lang_wild_statement_enum: dot = lang_do_assignments_1 (s->wild_statement.children.head, current_os, fill, dot, found_end); break; case lang_object_symbols_statement_enum: case lang_output_statement_enum: case lang_target_statement_enum: break; case lang_data_statement_enum: exp_fold_tree (s->data_statement.exp, bfd_abs_section_ptr, &dot); if (expld.result.valid_p) { s->data_statement.value = expld.result.value; if (expld.result.section != NULL) s->data_statement.value += expld.result.section->vma; } else if (expld.phase == lang_final_phase_enum) einfo (_("%F%P: invalid data statement\n")); { unsigned int size; switch (s->data_statement.type) { default: abort (); case QUAD: case SQUAD: size = QUAD_SIZE; break; case LONG: size = LONG_SIZE; break; case SHORT: size = SHORT_SIZE; break; case BYTE: size = BYTE_SIZE; break; } if (size < TO_SIZE ((unsigned) 1)) size = TO_SIZE ((unsigned) 1); dot += TO_ADDR (size); } break; case lang_reloc_statement_enum: exp_fold_tree (s->reloc_statement.addend_exp, bfd_abs_section_ptr, &dot); if (expld.result.valid_p) s->reloc_statement.addend_value = expld.result.value; else if (expld.phase == lang_final_phase_enum) einfo (_("%F%P: invalid reloc statement\n")); dot += TO_ADDR (bfd_get_reloc_size (s->reloc_statement.howto)); break; case lang_input_section_enum: { asection *in = s->input_section.section; if ((in->flags & SEC_EXCLUDE) == 0) dot += TO_ADDR (in->size); } break; case lang_input_statement_enum: break; case lang_fill_statement_enum: fill = s->fill_statement.fill; break; case lang_assignment_statement_enum: current_assign = &s->assignment_statement; if (current_assign->exp->type.node_class != etree_assert) { const char *p = current_assign->exp->assign.dst; if (current_os == abs_output_section && p[0] == '.' && p[1] == 0) prefer_next_section = true; while (*p == '_') ++p; if (strcmp (p, "end") == 0) *found_end = true; } exp_fold_tree (s->assignment_statement.exp, (current_os->bfd_section != NULL ? current_os->bfd_section : bfd_und_section_ptr), &dot); break; case lang_padding_statement_enum: dot += TO_ADDR (s->padding_statement.size); break; case lang_group_statement_enum: dot = lang_do_assignments_1 (s->group_statement.children.head, current_os, fill, dot, found_end); break; case lang_insert_statement_enum: break; case lang_address_statement_enum: break; default: FAIL (); break; } } return dot; } void lang_do_assignments (lang_phase_type phase) { bool found_end = false; current_section = NULL; prefer_next_section = false; expld.phase = phase; lang_statement_iteration++; lang_do_assignments_1 (statement_list.head, abs_output_section, NULL, 0, &found_end); } /* For an assignment statement outside of an output section statement, choose the best of neighbouring output sections to use for values of "dot". */ asection * section_for_dot (void) { asection *s; /* Assignments belong to the previous output section, unless there has been an assignment to "dot", in which case following assignments belong to the next output section. (The assumption is that an assignment to "dot" is setting up the address for the next output section.) Except that past the assignment to "_end" we always associate with the previous section. This exception is for targets like SH that define an alloc .stack or other weirdness after non-alloc sections. */ if (current_section == NULL || prefer_next_section) { lang_statement_union_type *stmt; lang_output_section_statement_type *os; for (stmt = (lang_statement_union_type *) current_assign; stmt != NULL; stmt = stmt->header.next) if (stmt->header.type == lang_output_section_statement_enum) break; os = stmt ? &stmt->output_section_statement : NULL; while (os != NULL && !os->after_end && (os->bfd_section == NULL || (os->bfd_section->flags & SEC_EXCLUDE) != 0 || bfd_section_removed_from_list (link_info.output_bfd, os->bfd_section))) os = os->next; if (current_section == NULL || os == NULL || !os->after_end) { if (os != NULL) s = os->bfd_section; else s = link_info.output_bfd->section_last; while (s != NULL && ((s->flags & SEC_ALLOC) == 0 || (s->flags & SEC_THREAD_LOCAL) != 0)) s = s->prev; if (s != NULL) return s; return bfd_abs_section_ptr; } } s = current_section->bfd_section; /* The section may have been stripped. */ while (s != NULL && ((s->flags & SEC_EXCLUDE) != 0 || (s->flags & SEC_ALLOC) == 0 || (s->flags & SEC_THREAD_LOCAL) != 0 || bfd_section_removed_from_list (link_info.output_bfd, s))) s = s->prev; if (s == NULL) s = link_info.output_bfd->sections; while (s != NULL && ((s->flags & SEC_ALLOC) == 0 || (s->flags & SEC_THREAD_LOCAL) != 0)) s = s->next; if (s != NULL) return s; return bfd_abs_section_ptr; } /* Array of __start/__stop/.startof./.sizeof/ symbols. */ static struct bfd_link_hash_entry **start_stop_syms; static size_t start_stop_count = 0; static size_t start_stop_alloc = 0; /* Give start/stop SYMBOL for SEC a preliminary definition, and add it to start_stop_syms. */ static void lang_define_start_stop (const char *symbol, asection *sec) { struct bfd_link_hash_entry *h; h = bfd_define_start_stop (link_info.output_bfd, &link_info, symbol, sec); if (h != NULL) { if (start_stop_count == start_stop_alloc) { start_stop_alloc = 2 * start_stop_alloc + 10; start_stop_syms = xrealloc (start_stop_syms, start_stop_alloc * sizeof (*start_stop_syms)); } start_stop_syms[start_stop_count++] = h; } } /* Check for input sections whose names match references to __start_SECNAME or __stop_SECNAME symbols. Give the symbols preliminary definitions. */ static void lang_init_start_stop (void) { bfd *abfd; asection *s; char leading_char = bfd_get_symbol_leading_char (link_info.output_bfd); for (abfd = link_info.input_bfds; abfd != NULL; abfd = abfd->link.next) for (s = abfd->sections; s != NULL; s = s->next) { const char *ps; const char *secname = s->name; for (ps = secname; *ps != '\0'; ps++) if (!ISALNUM ((unsigned char) *ps) && *ps != '_') break; if (*ps == '\0') { char *symbol = (char *) xmalloc (10 + strlen (secname)); symbol[0] = leading_char; sprintf (symbol + (leading_char != 0), "__start_%s", secname); lang_define_start_stop (symbol, s); symbol[1] = leading_char; memcpy (symbol + 1 + (leading_char != 0), "__stop", 6); lang_define_start_stop (symbol + 1, s); free (symbol); } } } /* Iterate over start_stop_syms. */ static void foreach_start_stop (void (*func) (struct bfd_link_hash_entry *)) { size_t i; for (i = 0; i < start_stop_count; ++i) func (start_stop_syms[i]); } /* __start and __stop symbols are only supposed to be defined by the linker for orphan sections, but we now extend that to sections that map to an output section of the same name. The symbols were defined early for --gc-sections, before we mapped input to output sections, so undo those that don't satisfy this rule. */ static void undef_start_stop (struct bfd_link_hash_entry *h) { if (h->ldscript_def) return; if (h->u.def.section->output_section == NULL || h->u.def.section->output_section->owner != link_info.output_bfd || strcmp (h->u.def.section->name, h->u.def.section->output_section->name) != 0) { asection *sec = bfd_get_section_by_name (link_info.output_bfd, h->u.def.section->name); if (sec != NULL) { /* When there are more than one input sections with the same section name, SECNAME, linker picks the first one to define __start_SECNAME and __stop_SECNAME symbols. When the first input section is removed by comdat group, we need to check if there is still an output section with section name SECNAME. */ asection *i; for (i = sec->map_head.s; i != NULL; i = i->map_head.s) if (strcmp (h->u.def.section->name, i->name) == 0) { h->u.def.section = i; return; } } h->type = bfd_link_hash_undefined; h->u.undef.abfd = NULL; if (is_elf_hash_table (link_info.hash)) { const struct elf_backend_data *bed; struct elf_link_hash_entry *eh = (struct elf_link_hash_entry *) h; unsigned int was_forced = eh->forced_local; bed = get_elf_backend_data (link_info.output_bfd); (*bed->elf_backend_hide_symbol) (&link_info, eh, true); if (!eh->ref_regular_nonweak) h->type = bfd_link_hash_undefweak; eh->def_regular = 0; eh->forced_local = was_forced; } } } static void lang_undef_start_stop (void) { foreach_start_stop (undef_start_stop); } /* Check for output sections whose names match references to .startof.SECNAME or .sizeof.SECNAME symbols. Give the symbols preliminary definitions. */ static void lang_init_startof_sizeof (void) { asection *s; for (s = link_info.output_bfd->sections; s != NULL; s = s->next) { const char *secname = s->name; char *symbol = (char *) xmalloc (10 + strlen (secname)); sprintf (symbol, ".startof.%s", secname); lang_define_start_stop (symbol, s); memcpy (symbol + 1, ".size", 5); lang_define_start_stop (symbol + 1, s); free (symbol); } } /* Set .startof., .sizeof., __start and __stop symbols final values. */ static void set_start_stop (struct bfd_link_hash_entry *h) { if (h->ldscript_def || h->type != bfd_link_hash_defined) return; if (h->root.string[0] == '.') { /* .startof. or .sizeof. symbol. .startof. already has final value. */ if (h->root.string[2] == 'i') { /* .sizeof. */ h->u.def.value = TO_ADDR (h->u.def.section->size); h->u.def.section = bfd_abs_section_ptr; } } else { /* __start or __stop symbol. */ int has_lead = bfd_get_symbol_leading_char (link_info.output_bfd) != 0; h->u.def.section = h->u.def.section->output_section; if (h->root.string[4 + has_lead] == 'o') { /* __stop_ */ h->u.def.value = TO_ADDR (h->u.def.section->size); } } } static void lang_finalize_start_stop (void) { foreach_start_stop (set_start_stop); } static void lang_symbol_tweaks (void) { /* Give initial values for __start and __stop symbols, so that ELF gc_sections will keep sections referenced by these symbols. Must be done before lang_do_assignments. */ if (config.build_constructors) lang_init_start_stop (); /* Make __ehdr_start hidden, and set def_regular even though it is likely undefined at this stage. For lang_check_relocs. */ if (is_elf_hash_table (link_info.hash) && !bfd_link_relocatable (&link_info)) { struct elf_link_hash_entry *h = (struct elf_link_hash_entry *) bfd_link_hash_lookup (link_info.hash, "__ehdr_start", false, false, true); /* Only adjust the export class if the symbol was referenced and not defined, otherwise leave it alone. */ if (h != NULL && (h->root.type == bfd_link_hash_new || h->root.type == bfd_link_hash_undefined || h->root.type == bfd_link_hash_undefweak || h->root.type == bfd_link_hash_common)) { const struct elf_backend_data *bed; bed = get_elf_backend_data (link_info.output_bfd); (*bed->elf_backend_hide_symbol) (&link_info, h, true); if (ELF_ST_VISIBILITY (h->other) != STV_INTERNAL) h->other = (h->other & ~ELF_ST_VISIBILITY (-1)) | STV_HIDDEN; h->def_regular = 1; h->root.linker_def = 1; h->root.rel_from_abs = 1; } } } static void lang_end (void) { struct bfd_link_hash_entry *h; bool warn; if ((bfd_link_relocatable (&link_info) && !link_info.gc_sections) || bfd_link_dll (&link_info)) warn = entry_from_cmdline; else warn = true; /* Force the user to specify a root when generating a relocatable with --gc-sections, unless --gc-keep-exported was also given. */ if (bfd_link_relocatable (&link_info) && link_info.gc_sections && !link_info.gc_keep_exported) { struct bfd_sym_chain *sym; for (sym = link_info.gc_sym_list; sym != NULL; sym = sym->next) { h = bfd_link_hash_lookup (link_info.hash, sym->name, false, false, false); if (h != NULL && (h->type == bfd_link_hash_defined || h->type == bfd_link_hash_defweak) && !bfd_is_const_section (h->u.def.section)) break; } if (!sym) einfo (_("%F%P: --gc-sections requires a defined symbol root " "specified by -e or -u\n")); } if (entry_symbol.name == NULL) { /* No entry has been specified. Look for the default entry, but don't warn if we don't find it. */ entry_symbol.name = entry_symbol_default; warn = false; } h = bfd_link_hash_lookup (link_info.hash, entry_symbol.name, false, false, true); if (h != NULL && (h->type == bfd_link_hash_defined || h->type == bfd_link_hash_defweak) && h->u.def.section->output_section != NULL) { bfd_vma val; val = (h->u.def.value + bfd_section_vma (h->u.def.section->output_section) + h->u.def.section->output_offset); if (!bfd_set_start_address (link_info.output_bfd, val)) einfo (_("%F%P: %s: can't set start address\n"), entry_symbol.name); } else { bfd_vma val; const char *send; /* We couldn't find the entry symbol. Try parsing it as a number. */ val = bfd_scan_vma (entry_symbol.name, &send, 0); if (*send == '\0') { if (!bfd_set_start_address (link_info.output_bfd, val)) einfo (_("%F%P: can't set start address\n")); } /* BZ 2004952: Only use the start of the entry section for executables. */ else if bfd_link_executable (&link_info) { asection *ts; /* Can't find the entry symbol, and it's not a number. Use the first address in the text section. */ ts = bfd_get_section_by_name (link_info.output_bfd, entry_section); if (ts != NULL) { if (warn) einfo (_("%P: warning: cannot find entry symbol %s;" " defaulting to %V\n"), entry_symbol.name, bfd_section_vma (ts)); if (!bfd_set_start_address (link_info.output_bfd, bfd_section_vma (ts))) einfo (_("%F%P: can't set start address\n")); } else { if (warn) einfo (_("%P: warning: cannot find entry symbol %s;" " not setting start address\n"), entry_symbol.name); } } else { if (warn) einfo (_("%P: warning: cannot find entry symbol %s;" " not setting start address\n"), entry_symbol.name); } } } /* This is a small function used when we want to ignore errors from BFD. */ static void ignore_bfd_errors (const char *fmt ATTRIBUTE_UNUSED, va_list ap ATTRIBUTE_UNUSED) { /* Don't do anything. */ } /* Check that the architecture of all the input files is compatible with the output file. Also call the backend to let it do any other checking that is needed. */ static void lang_check (void) { lang_input_statement_type *file; bfd *input_bfd; const bfd_arch_info_type *compatible; for (file = (void *) file_chain.head; file != NULL; file = file->next) { #if BFD_SUPPORTS_PLUGINS /* Don't check format of files claimed by plugin. */ if (file->flags.claimed) continue; #endif /* BFD_SUPPORTS_PLUGINS */ input_bfd = file->the_bfd; compatible = bfd_arch_get_compatible (input_bfd, link_info.output_bfd, command_line.accept_unknown_input_arch); /* In general it is not possible to perform a relocatable link between differing object formats when the input file has relocations, because the relocations in the input format may not have equivalent representations in the output format (and besides BFD does not translate relocs for other link purposes than a final link). */ if (!file->flags.just_syms && (bfd_link_relocatable (&link_info) || link_info.emitrelocations) && (compatible == NULL || (bfd_get_flavour (input_bfd) != bfd_get_flavour (link_info.output_bfd))) && (bfd_get_file_flags (input_bfd) & HAS_RELOC) != 0) { einfo (_("%F%P: relocatable linking with relocations from" " format %s (%pB) to format %s (%pB) is not supported\n"), bfd_get_target (input_bfd), input_bfd, bfd_get_target (link_info.output_bfd), link_info.output_bfd); /* einfo with %F exits. */ } if (compatible == NULL) { if (command_line.warn_mismatch) einfo (_("%X%P: %s architecture of input file `%pB'" " is incompatible with %s output\n"), bfd_printable_name (input_bfd), input_bfd, bfd_printable_name (link_info.output_bfd)); } /* If the input bfd has no contents, it shouldn't set the private data of the output bfd. */ else if (!file->flags.just_syms && ((input_bfd->flags & DYNAMIC) != 0 || bfd_count_sections (input_bfd) != 0)) { bfd_error_handler_type pfn = NULL; /* If we aren't supposed to warn about mismatched input files, temporarily set the BFD error handler to a function which will do nothing. We still want to call bfd_merge_private_bfd_data, since it may set up information which is needed in the output file. */ if (!command_line.warn_mismatch) pfn = bfd_set_error_handler (ignore_bfd_errors); if (!bfd_merge_private_bfd_data (input_bfd, &link_info)) { if (command_line.warn_mismatch) einfo (_("%X%P: failed to merge target specific data" " of file %pB\n"), input_bfd); } if (!command_line.warn_mismatch) bfd_set_error_handler (pfn); } } } /* Look through all the global common symbols and attach them to the correct section. The -sort-common command line switch may be used to roughly sort the entries by alignment. */ static void lang_common (void) { if (link_info.inhibit_common_definition) return; if (bfd_link_relocatable (&link_info) && !command_line.force_common_definition) return; if (!config.sort_common) bfd_link_hash_traverse (link_info.hash, lang_one_common, NULL); else { unsigned int power; if (config.sort_common == sort_descending) { for (power = 4; power > 0; power--) bfd_link_hash_traverse (link_info.hash, lang_one_common, &power); power = 0; bfd_link_hash_traverse (link_info.hash, lang_one_common, &power); } else { for (power = 0; power <= 4; power++) bfd_link_hash_traverse (link_info.hash, lang_one_common, &power); power = (unsigned int) -1; bfd_link_hash_traverse (link_info.hash, lang_one_common, &power); } } } /* Place one common symbol in the correct section. */ static bool lang_one_common (struct bfd_link_hash_entry *h, void *info) { unsigned int power_of_two; bfd_vma size; asection *section; if (h->type != bfd_link_hash_common) return true; size = h->u.c.size; power_of_two = h->u.c.p->alignment_power; if (config.sort_common == sort_descending && power_of_two < *(unsigned int *) info) return true; else if (config.sort_common == sort_ascending && power_of_two > *(unsigned int *) info) return true; section = h->u.c.p->section; if (!bfd_define_common_symbol (link_info.output_bfd, &link_info, h)) einfo (_("%F%P: could not define common symbol `%pT': %E\n"), h->root.string); if (config.map_file != NULL) { static bool header_printed; int len; char *name; char buf[32]; if (!header_printed) { minfo (_("\nAllocating common symbols\n")); minfo (_("Common symbol size file\n\n")); header_printed = true; } name = bfd_demangle (link_info.output_bfd, h->root.string, DMGL_ANSI | DMGL_PARAMS); if (name == NULL) { minfo ("%s", h->root.string); len = strlen (h->root.string); } else { minfo ("%s", name); len = strlen (name); free (name); } if (len >= 19) { print_nl (); len = 0; } sprintf (buf, "%" PRIx64, (uint64_t) size); fprintf (config.map_file, "%*s0x%-16s", 20 - len, "", buf); minfo ("%pB\n", section->owner); } return true; } /* Handle a single orphan section S, placing the orphan into an appropriate output section. The effects of the --orphan-handling command line option are handled here. */ static void ldlang_place_orphan (asection *s) { if (config.orphan_handling == orphan_handling_discard) { lang_output_section_statement_type *os; os = lang_output_section_statement_lookup (DISCARD_SECTION_NAME, 0, 1); if (os->addr_tree == NULL && (bfd_link_relocatable (&link_info) || (s->flags & (SEC_LOAD | SEC_ALLOC)) == 0)) os->addr_tree = exp_intop (0); lang_add_section (&os->children, s, NULL, NULL, os); } else { lang_output_section_statement_type *os; const char *name = s->name; int constraint = 0; if (config.orphan_handling == orphan_handling_error) einfo (_("%X%P: error: unplaced orphan section `%pA' from `%pB'\n"), s, s->owner); if (config.unique_orphan_sections || unique_section_p (s, NULL)) constraint = SPECIAL; os = ldemul_place_orphan (s, name, constraint); if (os == NULL) { os = lang_output_section_statement_lookup (name, constraint, 1); if (os->addr_tree == NULL && (bfd_link_relocatable (&link_info) || (s->flags & (SEC_LOAD | SEC_ALLOC)) == 0)) os->addr_tree = exp_intop (0); lang_add_section (&os->children, s, NULL, NULL, os); } if (config.orphan_handling == orphan_handling_warn) einfo (_("%P: warning: orphan section `%pA' from `%pB' being " "placed in section `%s'\n"), s, s->owner, os->name); } } /* Run through the input files and ensure that every input section has somewhere to go. If one is found without a destination then create an input request and place it into the statement tree. */ static void lang_place_orphans (void) { LANG_FOR_EACH_INPUT_STATEMENT (file) { asection *s; for (s = file->the_bfd->sections; s != NULL; s = s->next) { if (s->output_section == NULL) { /* This section of the file is not attached, root around for a sensible place for it to go. */ if (file->flags.just_syms) bfd_link_just_syms (file->the_bfd, s, &link_info); else if (lang_discard_section_p (s)) s->output_section = bfd_abs_section_ptr; else if (strcmp (s->name, "COMMON") == 0) { /* This is a lonely common section which must have come from an archive. We attach to the section with the wildcard. */ if (!bfd_link_relocatable (&link_info) || command_line.force_common_definition) { if (default_common_section == NULL) default_common_section = lang_output_section_statement_lookup (".bss", 0, 1); lang_add_section (&default_common_section->children, s, NULL, NULL, default_common_section); } } else ldlang_place_orphan (s); } } } } void lang_set_flags (lang_memory_region_type *ptr, const char *flags, int invert) { flagword *ptr_flags; ptr_flags = invert ? &ptr->not_flags : &ptr->flags; while (*flags) { switch (*flags) { /* PR 17900: An exclamation mark in the attributes reverses the sense of any of the attributes that follow. */ case '!': invert = !invert; ptr_flags = invert ? &ptr->not_flags : &ptr->flags; break; case 'A': case 'a': *ptr_flags |= SEC_ALLOC; break; case 'R': case 'r': *ptr_flags |= SEC_READONLY; break; case 'W': case 'w': *ptr_flags |= SEC_DATA; break; case 'X': case 'x': *ptr_flags |= SEC_CODE; break; case 'L': case 'l': case 'I': case 'i': *ptr_flags |= SEC_LOAD; break; default: einfo (_("%F%P: invalid character %c (%d) in flags\n"), *flags, *flags); break; } flags++; } } /* Call a function on each real input file. This function will be called on an archive, but not on the elements. */ void lang_for_each_input_file (void (*func) (lang_input_statement_type *)) { lang_input_statement_type *f; for (f = (void *) input_file_chain.head; f != NULL; f = f->next_real_file) if (f->flags.real) func (f); } /* Call a function on each real file. The function will be called on all the elements of an archive which are included in the link, but will not be called on the archive file itself. */ void lang_for_each_file (void (*func) (lang_input_statement_type *)) { LANG_FOR_EACH_INPUT_STATEMENT (f) { if (f->flags.real) func (f); } } void ldlang_add_file (lang_input_statement_type *entry) { lang_statement_append (&file_chain, entry, &entry->next); /* The BFD linker needs to have a list of all input BFDs involved in a link. */ ASSERT (link_info.input_bfds_tail != &entry->the_bfd->link.next && entry->the_bfd->link.next == NULL); ASSERT (entry->the_bfd != link_info.output_bfd); *link_info.input_bfds_tail = entry->the_bfd; link_info.input_bfds_tail = &entry->the_bfd->link.next; bfd_set_usrdata (entry->the_bfd, entry); bfd_set_gp_size (entry->the_bfd, g_switch_value); /* Look through the sections and check for any which should not be included in the link. We need to do this now, so that we can notice when the backend linker tries to report multiple definition errors for symbols which are in sections we aren't going to link. FIXME: It might be better to entirely ignore symbols which are defined in sections which are going to be discarded. This would require modifying the backend linker for each backend which might set the SEC_LINK_ONCE flag. If we do this, we should probably handle SEC_EXCLUDE in the same way. */ bfd_map_over_sections (entry->the_bfd, section_already_linked, entry); } void lang_add_output (const char *name, int from_script) { /* Make -o on command line override OUTPUT in script. */ if (!had_output_filename || !from_script) { output_filename = name; had_output_filename = true; } } lang_output_section_statement_type * lang_enter_output_section_statement (const char *output_section_statement_name, etree_type *address_exp, enum section_type sectype, etree_type *sectype_value, etree_type *align, etree_type *subalign, etree_type *ebase, int constraint, int align_with_input) { lang_output_section_statement_type *os; os = lang_output_section_statement_lookup (output_section_statement_name, constraint, 2); current_section = os; if (os->addr_tree == NULL) { os->addr_tree = address_exp; } os->sectype = sectype; if (sectype == type_section || sectype == typed_readonly_section) os->sectype_value = sectype_value; else if (sectype == noload_section) os->flags = SEC_NEVER_LOAD; else os->flags = SEC_NO_FLAGS; os->block_value = 1; /* Make next things chain into subchain of this. */ push_stat_ptr (&os->children); os->align_lma_with_input = align_with_input == ALIGN_WITH_INPUT; if (os->align_lma_with_input && align != NULL) einfo (_("%F%P:%pS: error: align with input and explicit align specified\n"), NULL); os->subsection_alignment = subalign; os->section_alignment = align; os->load_base = ebase; return os; } void lang_final (void) { lang_output_statement_type *new_stmt; new_stmt = new_stat (lang_output_statement, stat_ptr); new_stmt->name = output_filename; } /* Reset the current counters in the regions. */ void lang_reset_memory_regions (void) { lang_memory_region_type *p = lang_memory_region_list; asection *o; lang_output_section_statement_type *os; for (p = lang_memory_region_list; p != NULL; p = p->next) { p->current = p->origin; p->last_os = NULL; } for (os = (void *) lang_os_list.head; os != NULL; os = os->next) { os->processed_vma = false; os->processed_lma = false; } for (o = link_info.output_bfd->sections; o != NULL; o = o->next) { /* Save the last size for possible use by bfd_relax_section. */ o->rawsize = o->size; if (!(o->flags & SEC_FIXED_SIZE)) o->size = 0; } } /* Worker for lang_gc_sections_1. */ static void gc_section_callback (lang_wild_statement_type *ptr, struct wildcard_list *sec ATTRIBUTE_UNUSED, asection *section, lang_input_statement_type *file ATTRIBUTE_UNUSED, void *data ATTRIBUTE_UNUSED) { /* If the wild pattern was marked KEEP, the member sections should be as well. */ if (ptr->keep_sections) section->flags |= SEC_KEEP; } /* Iterate over sections marking them against GC. */ static void lang_gc_sections_1 (lang_statement_union_type *s) { for (; s != NULL; s = s->header.next) { switch (s->header.type) { case lang_wild_statement_enum: walk_wild (&s->wild_statement, gc_section_callback, NULL); break; case lang_constructors_statement_enum: lang_gc_sections_1 (constructor_list.head); break; case lang_output_section_statement_enum: lang_gc_sections_1 (s->output_section_statement.children.head); break; case lang_group_statement_enum: lang_gc_sections_1 (s->group_statement.children.head); break; default: break; } } } static void lang_gc_sections (void) { /* Keep all sections so marked in the link script. */ lang_gc_sections_1 (statement_list.head); /* SEC_EXCLUDE is ignored when doing a relocatable link, except in the special case of .stabstr debug info. (See bfd/stabs.c) Twiddle the flag here, to simplify later linker code. */ if (bfd_link_relocatable (&link_info)) { LANG_FOR_EACH_INPUT_STATEMENT (f) { asection *sec; #if BFD_SUPPORTS_PLUGINS if (f->flags.claimed) continue; #endif for (sec = f->the_bfd->sections; sec != NULL; sec = sec->next) if ((sec->flags & SEC_DEBUGGING) == 0 || strcmp (sec->name, ".stabstr") != 0) sec->flags &= ~SEC_EXCLUDE; } } if (link_info.gc_sections) bfd_gc_sections (link_info.output_bfd, &link_info); } /* Worker for lang_find_relro_sections_1. */ static void find_relro_section_callback (lang_wild_statement_type *ptr ATTRIBUTE_UNUSED, struct wildcard_list *sec ATTRIBUTE_UNUSED, asection *section, lang_input_statement_type *file ATTRIBUTE_UNUSED, void *data) { /* Discarded, excluded and ignored sections effectively have zero size. */ if (section->output_section != NULL && section->output_section->owner == link_info.output_bfd && (section->output_section->flags & SEC_EXCLUDE) == 0 && !IGNORE_SECTION (section) && section->size != 0) { bool *has_relro_section = (bool *) data; *has_relro_section = true; } } /* Iterate over sections for relro sections. */ static void lang_find_relro_sections_1 (lang_statement_union_type *s, bool *has_relro_section) { if (*has_relro_section) return; for (; s != NULL; s = s->header.next) { if (s == expld.dataseg.relro_end_stat) break; switch (s->header.type) { case lang_wild_statement_enum: walk_wild (&s->wild_statement, find_relro_section_callback, has_relro_section); break; case lang_constructors_statement_enum: lang_find_relro_sections_1 (constructor_list.head, has_relro_section); break; case lang_output_section_statement_enum: lang_find_relro_sections_1 (s->output_section_statement.children.head, has_relro_section); break; case lang_group_statement_enum: lang_find_relro_sections_1 (s->group_statement.children.head, has_relro_section); break; default: break; } } } static void lang_find_relro_sections (void) { bool has_relro_section = false; /* Check all sections in the link script. */ lang_find_relro_sections_1 (expld.dataseg.relro_start_stat, &has_relro_section); if (!has_relro_section) link_info.relro = false; } /* Relax all sections until bfd_relax_section gives up. */ void lang_relax_sections (bool need_layout) { /* NB: Also enable relaxation to layout sections for DT_RELR. */ if (RELAXATION_ENABLED || link_info.enable_dt_relr) { /* We may need more than one relaxation pass. */ int i = link_info.relax_pass; /* The backend can use it to determine the current pass. */ link_info.relax_pass = 0; while (i--) { /* Keep relaxing until bfd_relax_section gives up. */ bool relax_again; link_info.relax_trip = -1; do { link_info.relax_trip++; /* Note: pe-dll.c does something like this also. If you find you need to change this code, you probably need to change pe-dll.c also. DJ */ /* Do all the assignments with our current guesses as to section sizes. */ lang_do_assignments (lang_assigning_phase_enum); /* We must do this after lang_do_assignments, because it uses size. */ lang_reset_memory_regions (); /* Perform another relax pass - this time we know where the globals are, so can make a better guess. */ relax_again = false; lang_size_sections (&relax_again, false); } while (relax_again); link_info.relax_pass++; } need_layout = true; } if (need_layout) { /* Final extra sizing to report errors. */ lang_do_assignments (lang_assigning_phase_enum); lang_reset_memory_regions (); lang_size_sections (NULL, true); } } #if BFD_SUPPORTS_PLUGINS /* Find the insert point for the plugin's replacement files. We place them after the first claimed real object file, or if the first claimed object is an archive member, after the last real object file immediately preceding the archive. In the event no objects have been claimed at all, we return the first dummy object file on the list as the insert point; that works, but the callee must be careful when relinking the file_chain as it is not actually on that chain, only the statement_list and the input_file list; in that case, the replacement files must be inserted at the head of the file_chain. */ static lang_input_statement_type * find_replacements_insert_point (bool *before) { lang_input_statement_type *claim1, *lastobject; lastobject = (void *) input_file_chain.head; for (claim1 = (void *) file_chain.head; claim1 != NULL; claim1 = claim1->next) { if (claim1->flags.claimed) { *before = claim1->flags.claim_archive; return claim1->flags.claim_archive ? lastobject : claim1; } /* Update lastobject if this is a real object file. */ if (claim1->the_bfd != NULL && claim1->the_bfd->my_archive == NULL) lastobject = claim1; } /* No files were claimed by the plugin. Choose the last object file found on the list (maybe the first, dummy entry) as the insert point. */ *before = false; return lastobject; } /* Find where to insert ADD, an archive element or shared library added during a rescan. */ static lang_input_statement_type ** find_rescan_insertion (lang_input_statement_type *add) { bfd *add_bfd = add->the_bfd; lang_input_statement_type *f; lang_input_statement_type *last_loaded = NULL; lang_input_statement_type *before = NULL; lang_input_statement_type **iter = NULL; if (add_bfd->my_archive != NULL) add_bfd = add_bfd->my_archive; /* First look through the input file chain, to find an object file before the one we've rescanned. Normal object files always appear on both the input file chain and the file chain, so this lets us get quickly to somewhere near the correct place on the file chain if it is full of archive elements. Archives don't appear on the file chain, but if an element has been extracted then their input_statement->next points at it. */ for (f = (void *) input_file_chain.head; f != NULL; f = f->next_real_file) { if (f->the_bfd == add_bfd) { before = last_loaded; if (f->next != NULL) return &f->next->next; } if (f->the_bfd != NULL && f->next != NULL) last_loaded = f; } for (iter = before ? &before->next : &file_chain.head->input_statement.next; *iter != NULL; iter = &(*iter)->next) if (!(*iter)->flags.claim_archive && (*iter)->the_bfd->my_archive == NULL) break; return iter; } /* Insert SRCLIST into DESTLIST after given element by chaining on FIELD as the next-pointer. (Counterintuitively does not need a pointer to the actual after-node itself, just its chain field.) */ static void lang_list_insert_after (lang_statement_list_type *destlist, lang_statement_list_type *srclist, lang_statement_union_type **field) { *(srclist->tail) = *field; *field = srclist->head; if (destlist->tail == field) destlist->tail = srclist->tail; } /* Detach new nodes added to DESTLIST since the time ORIGLIST was taken as a copy of it and leave them in ORIGLIST. */ static void lang_list_remove_tail (lang_statement_list_type *destlist, lang_statement_list_type *origlist) { union lang_statement_union **savetail; /* Check that ORIGLIST really is an earlier state of DESTLIST. */ ASSERT (origlist->head == destlist->head); savetail = origlist->tail; origlist->head = *(savetail); origlist->tail = destlist->tail; destlist->tail = savetail; *savetail = NULL; } static lang_statement_union_type ** find_next_input_statement (lang_statement_union_type **s) { for ( ; *s; s = &(*s)->header.next) { lang_statement_union_type **t; switch ((*s)->header.type) { case lang_input_statement_enum: return s; case lang_wild_statement_enum: t = &(*s)->wild_statement.children.head; break; case lang_group_statement_enum: t = &(*s)->group_statement.children.head; break; case lang_output_section_statement_enum: t = &(*s)->output_section_statement.children.head; break; default: continue; } t = find_next_input_statement (t); if (*t) return t; } return s; } #endif /* BFD_SUPPORTS_PLUGINS */ /* Add NAME to the list of garbage collection entry points. */ void lang_add_gc_name (const char *name) { struct bfd_sym_chain *sym; if (name == NULL) return; sym = stat_alloc (sizeof (*sym)); sym->next = link_info.gc_sym_list; sym->name = name; link_info.gc_sym_list = sym; } /* Check relocations. */ static void lang_check_relocs (void) { if (link_info.check_relocs_after_open_input) { bfd *abfd; for (abfd = link_info.input_bfds; abfd != (bfd *) NULL; abfd = abfd->link.next) if (!bfd_link_check_relocs (abfd, &link_info)) { /* No object output, fail return. */ config.make_executable = false; /* Note: we do not abort the loop, but rather continue the scan in case there are other bad relocations to report. */ } } } /* Look through all output sections looking for places where we can propagate forward the lma region. */ static void lang_propagate_lma_regions (void) { lang_output_section_statement_type *os; for (os = (void *) lang_os_list.head; os != NULL; os = os->next) { if (os->prev != NULL && os->lma_region == NULL && os->load_base == NULL && os->addr_tree == NULL && os->region == os->prev->region) os->lma_region = os->prev->lma_region; } } static void warn_non_contiguous_discards (void) { LANG_FOR_EACH_INPUT_STATEMENT (file) { if ((file->the_bfd->flags & (BFD_LINKER_CREATED | DYNAMIC)) != 0 || file->flags.just_syms) continue; for (asection *s = file->the_bfd->sections; s != NULL; s = s->next) if (s->output_section == NULL && (s->flags & SEC_LINKER_CREATED) == 0) einfo (_("%P: warning: --enable-non-contiguous-regions " "discards section `%pA' from `%pB'\n"), s, file->the_bfd); } } static void reset_one_wild (lang_statement_union_type *statement) { if (statement->header.type == lang_wild_statement_enum) { lang_wild_statement_type *stmt = &statement->wild_statement; lang_list_init (&stmt->matching_sections); } } static void reset_resolved_wilds (void) { lang_for_each_statement (reset_one_wild); } void lang_process (void) { /* Finalize dynamic list. */ if (link_info.dynamic_list) lang_finalize_version_expr_head (&link_info.dynamic_list->head); current_target = default_target; /* Open the output file. */ lang_for_each_statement (ldlang_open_output); init_opb (NULL); ldemul_create_output_section_statements (); /* Add to the hash table all undefineds on the command line. */ lang_place_undefineds (); if (!bfd_section_already_linked_table_init ()) einfo (_("%F%P: can not create hash table: %E\n")); /* A first pass through the memory regions ensures that if any region references a symbol for its origin or length then this symbol will be added to the symbol table. Having these symbols in the symbol table means that when we call open_input_bfds PROVIDE statements will trigger to provide any needed symbols. The regions origins and lengths are not assigned as a result of this call. */ lang_do_memory_regions (false); /* Create a bfd for each input file. */ current_target = default_target; lang_statement_iteration++; open_input_bfds (statement_list.head, OPEN_BFD_NORMAL); /* Now that open_input_bfds has processed assignments and provide statements we can give values to symbolic origin/length now. */ lang_do_memory_regions (true); #if BFD_SUPPORTS_PLUGINS if (link_info.lto_plugin_active) { lang_statement_list_type added; lang_statement_list_type files, inputfiles; ldemul_before_plugin_all_symbols_read (); /* Now all files are read, let the plugin(s) decide if there are any more to be added to the link before we call the emulation's after_open hook. We create a private list of input statements for this purpose, which we will eventually insert into the global statement list after the first claimed file. */ added = *stat_ptr; /* We need to manipulate all three chains in synchrony. */ files = file_chain; inputfiles = input_file_chain; if (plugin_call_all_symbols_read ()) einfo (_("%F%P: %s: plugin reported error after all symbols read\n"), plugin_error_plugin ()); link_info.lto_all_symbols_read = true; /* Open any newly added files, updating the file chains. */ plugin_undefs = link_info.hash->undefs_tail; open_input_bfds (*added.tail, OPEN_BFD_NORMAL); if (plugin_undefs == link_info.hash->undefs_tail) plugin_undefs = NULL; /* Restore the global list pointer now they have all been added. */ lang_list_remove_tail (stat_ptr, &added); /* And detach the fresh ends of the file lists. */ lang_list_remove_tail (&file_chain, &files); lang_list_remove_tail (&input_file_chain, &inputfiles); /* Were any new files added? */ if (added.head != NULL) { /* If so, we will insert them into the statement list immediately after the first input file that was claimed by the plugin, unless that file was an archive in which case it is inserted immediately before. */ bool before; lang_statement_union_type **prev; plugin_insert = find_replacements_insert_point (&before); /* If a plugin adds input files without having claimed any, we don't really have a good idea where to place them. Just putting them at the start or end of the list is liable to leave them outside the crtbegin...crtend range. */ ASSERT (plugin_insert != NULL); /* Splice the new statement list into the old one. */ prev = &plugin_insert->header.next; if (before) { prev = find_next_input_statement (prev); if (*prev != (void *) plugin_insert->next_real_file) { /* We didn't find the expected input statement. Fall back to adding after plugin_insert. */ prev = &plugin_insert->header.next; } } lang_list_insert_after (stat_ptr, &added, prev); /* Likewise for the file chains. */ lang_list_insert_after (&input_file_chain, &inputfiles, (void *) &plugin_insert->next_real_file); /* We must be careful when relinking file_chain; we may need to insert the new files at the head of the list if the insert point chosen is the dummy first input file. */ if (plugin_insert->filename) lang_list_insert_after (&file_chain, &files, (void *) &plugin_insert->next); else lang_list_insert_after (&file_chain, &files, &file_chain.head); /* Rescan archives in case new undefined symbols have appeared. */ files = file_chain; lang_statement_iteration++; open_input_bfds (statement_list.head, OPEN_BFD_RESCAN); lang_list_remove_tail (&file_chain, &files); while (files.head != NULL) { lang_input_statement_type **insert; lang_input_statement_type **iter, *temp; bfd *my_arch; insert = find_rescan_insertion (&files.head->input_statement); /* All elements from an archive can be added at once. */ iter = &files.head->input_statement.next; my_arch = files.head->input_statement.the_bfd->my_archive; if (my_arch != NULL) for (; *iter != NULL; iter = &(*iter)->next) if ((*iter)->the_bfd->my_archive != my_arch) break; temp = *insert; *insert = &files.head->input_statement; files.head = (lang_statement_union_type *) *iter; *iter = temp; if (file_chain.tail == (lang_statement_union_type **) insert) file_chain.tail = (lang_statement_union_type **) iter; if (my_arch != NULL) { lang_input_statement_type *parent = bfd_usrdata (my_arch); if (parent != NULL) parent->next = (lang_input_statement_type *) ((char *) iter - offsetof (lang_input_statement_type, next)); } } } } #endif /* BFD_SUPPORTS_PLUGINS */ /* Make sure that nobody has tried to add a symbol to this list before now. */ ASSERT (link_info.gc_sym_list == NULL); link_info.gc_sym_list = &entry_symbol; if (entry_symbol.name == NULL) { link_info.gc_sym_list = ldlang_undef_chain_list_head; /* entry_symbol is normally initialied by a ENTRY definition in the linker script or the -e command line option. But if neither of these have been used, the target specific backend may still have provided an entry symbol via a call to lang_default_entry(). Unfortunately this value will not be processed until lang_end() is called, long after this function has finished. So detect this case here and add the target's entry symbol to the list of starting points for garbage collection resolution. */ lang_add_gc_name (entry_symbol_default); } lang_add_gc_name (link_info.init_function); lang_add_gc_name (link_info.fini_function); ldemul_after_open (); if (config.map_file != NULL) lang_print_asneeded (); ldlang_open_ctf (); bfd_section_already_linked_table_free (); /* Make sure that we're not mixing architectures. We call this after all the input files have been opened, but before we do any other processing, so that any operations merge_private_bfd_data does on the output file will be known during the rest of the link. */ lang_check (); /* Handle .exports instead of a version script if we're told to do so. */ if (command_line.version_exports_section) lang_do_version_exports_section (); /* Build all sets based on the information gathered from the input files. */ ldctor_build_sets (); lang_symbol_tweaks (); /* PR 13683: We must rerun the assignments prior to running garbage collection in order to make sure that all symbol aliases are resolved. */ lang_do_assignments (lang_mark_phase_enum); expld.phase = lang_first_phase_enum; /* Size up the common data. */ lang_common (); if (0) debug_prefix_tree (); resolve_wilds (); /* Remove unreferenced sections if asked to. */ lang_gc_sections (); lang_mark_undefineds (); /* Check relocations. */ lang_check_relocs (); ldemul_after_check_relocs (); /* There might have been new sections created (e.g. as result of checking relocs to need a .got, or suchlike), so to properly order them into our lists of matching sections reset them here. */ reset_resolved_wilds (); resolve_wilds (); /* Update wild statements in case the user gave --sort-section. Note how the option might have come after the linker script and so couldn't have been set when the wild statements were created. */ update_wild_statements (statement_list.head); /* Run through the contours of the script and attach input sections to the correct output sections. */ lang_statement_iteration++; map_input_to_output_sections (statement_list.head, NULL, NULL); /* Start at the statement immediately after the special abs_section output statement, so that it isn't reordered. */ process_insert_statements (&lang_os_list.head->header.next); ldemul_before_place_orphans (); /* Find any sections not attached explicitly and handle them. */ lang_place_orphans (); if (!bfd_link_relocatable (&link_info)) { asection *found; /* Merge SEC_MERGE sections. This has to be done after GC of sections, so that GCed sections are not merged, but before assigning dynamic symbols, since removing whole input sections is hard then. */ bfd_merge_sections (link_info.output_bfd, &link_info); /* Look for a text section and set the readonly attribute in it. */ found = bfd_get_section_by_name (link_info.output_bfd, ".text"); if (found != NULL) { if (config.text_read_only) found->flags |= SEC_READONLY; else found->flags &= ~SEC_READONLY; } } /* Merge together CTF sections. After this, only the symtab-dependent function and data object sections need adjustment. */ lang_merge_ctf (); /* Emit the CTF, iff the emulation doesn't need to do late emission after examining things laid out late, like the strtab. */ lang_write_ctf (0); /* Copy forward lma regions for output sections in same lma region. */ lang_propagate_lma_regions (); /* Defining __start/__stop symbols early for --gc-sections to work around a glibc build problem can result in these symbols being defined when they should not be. Fix them now. */ if (config.build_constructors) lang_undef_start_stop (); /* Define .startof./.sizeof. symbols with preliminary values before dynamic symbols are created. */ if (!bfd_link_relocatable (&link_info)) lang_init_startof_sizeof (); /* Do anything special before sizing sections. This is where ELF and other back-ends size dynamic sections. */ ldemul_before_allocation (); /* We must record the program headers before we try to fix the section positions, since they will affect SIZEOF_HEADERS. */ lang_record_phdrs (); /* Check relro sections. */ if (link_info.relro && !bfd_link_relocatable (&link_info)) lang_find_relro_sections (); /* Size up the sections. */ lang_size_sections (NULL, !RELAXATION_ENABLED); /* See if anything special should be done now we know how big everything is. This is where relaxation is done. */ ldemul_after_allocation (); /* Fix any __start, __stop, .startof. or .sizeof. symbols. */ lang_finalize_start_stop (); /* Do all the assignments again, to report errors. Assignment statements are processed multiple times, updating symbols; In open_input_bfds, lang_do_assignments, and lang_size_sections. Since lang_relax_sections calls lang_do_assignments, symbols are also updated in ldemul_after_allocation. */ lang_do_assignments (lang_final_phase_enum); ldemul_finish (); /* Convert absolute symbols to section relative. */ ldexp_finalize_syms (); /* Make sure that the section addresses make sense. */ if (command_line.check_section_addresses) lang_check_section_addresses (); if (link_info.non_contiguous_regions && link_info.non_contiguous_regions_warnings) warn_non_contiguous_discards (); /* Check any required symbols are known. */ ldlang_check_require_defined_symbols (); lang_end (); } /* EXPORTED TO YACC */ void lang_add_wild (struct wildcard_spec *filespec, struct wildcard_list *section_list, bool keep_sections) { struct wildcard_list *curr, *next; lang_wild_statement_type *new_stmt; bool any_specs_sorted = false; /* Reverse the list as the parser puts it back to front. */ for (curr = section_list, section_list = NULL; curr != NULL; section_list = curr, curr = next) { if (curr->spec.sorted != none && curr->spec.sorted != by_none) any_specs_sorted = true; next = curr->next; curr->next = section_list; } if (filespec != NULL && filespec->name != NULL) { if (strcmp (filespec->name, "*") == 0) filespec->name = NULL; else if (!wildcardp (filespec->name)) lang_has_input_file = true; } new_stmt = new_stat (lang_wild_statement, stat_ptr); new_stmt->filename = NULL; new_stmt->filenames_sorted = false; new_stmt->any_specs_sorted = any_specs_sorted; new_stmt->section_flag_list = NULL; new_stmt->exclude_name_list = NULL; if (filespec != NULL) { new_stmt->filename = filespec->name; new_stmt->filenames_sorted = filespec->sorted == by_name; new_stmt->section_flag_list = filespec->section_flag_list; new_stmt->exclude_name_list = filespec->exclude_name_list; } new_stmt->section_list = section_list; new_stmt->keep_sections = keep_sections; lang_list_init (&new_stmt->children); lang_list_init (&new_stmt->matching_sections); analyze_walk_wild_section_handler (new_stmt); if (0) { printf ("wild %s(", new_stmt->filename ? new_stmt->filename : "*"); for (curr = new_stmt->section_list; curr; curr = curr->next) printf ("%s ", curr->spec.name ? curr->spec.name : "*"); printf (")\n"); } } void lang_section_start (const char *name, etree_type *address, const segment_type *segment) { lang_address_statement_type *ad; ad = new_stat (lang_address_statement, stat_ptr); ad->section_name = name; ad->address = address; ad->segment = segment; } /* Set the start symbol to NAME. CMDLINE is nonzero if this is called because of a -e argument on the command line, or zero if this is called by ENTRY in a linker script. Command line arguments take precedence. */ void lang_add_entry (const char *name, bool cmdline) { if (entry_symbol.name == NULL || cmdline || !entry_from_cmdline) { entry_symbol.name = name; entry_from_cmdline = cmdline; } } /* Set the default start symbol to NAME. .em files should use this, not lang_add_entry, to override the use of "start" if neither the linker script nor the command line specifies an entry point. NAME must be permanently allocated. */ void lang_default_entry (const char *name) { entry_symbol_default = name; } void lang_add_target (const char *name) { lang_target_statement_type *new_stmt; new_stmt = new_stat (lang_target_statement, stat_ptr); new_stmt->target = name; } void lang_add_map (const char *name) { while (*name) { switch (*name) { case 'F': map_option_f = true; break; } name++; } } void lang_add_fill (fill_type *fill) { lang_fill_statement_type *new_stmt; new_stmt = new_stat (lang_fill_statement, stat_ptr); new_stmt->fill = fill; } void lang_add_data (int type, union etree_union *exp) { lang_data_statement_type *new_stmt; new_stmt = new_stat (lang_data_statement, stat_ptr); new_stmt->exp = exp; new_stmt->type = type; } void lang_add_string (const char *s) { bfd_vma len = strlen (s); bfd_vma i; bool escape = false; /* Add byte expressions until end of string. */ for (i = 0 ; i < len; i++) { char c = *s++; if (escape) { switch (c) { default: /* Ignore the escape. */ break; case 'n': c = '\n'; break; case 'r': c = '\r'; break; case 't': c = '\t'; break; case '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7': /* We have an octal number. */ { unsigned int value = c - '0'; c = *s; if ((c >= '0') && (c <= '7')) { value <<= 3; value += (c - '0'); i++; s++; c = *s; if ((c >= '0') && (c <= '7')) { value <<= 3; value += (c - '0'); i++; s++; } } if (value > 0xff) { /* octal: \777 is treated as '\077' + '7' */ value >>= 3; i--; s--; } c = value; } break; } lang_add_data (BYTE, exp_intop (c)); escape = false; } else { if (c == '\\') escape = true; else lang_add_data (BYTE, exp_intop (c)); } } /* Remeber to terminate the string. */ lang_add_data (BYTE, exp_intop (0)); } /* Create a new reloc statement. RELOC is the BFD relocation type to generate. HOWTO is the corresponding howto structure (we could look this up, but the caller has already done so). SECTION is the section to generate a reloc against, or NAME is the name of the symbol to generate a reloc against. Exactly one of SECTION and NAME must be NULL. ADDEND is an expression for the addend. */ void lang_add_reloc (bfd_reloc_code_real_type reloc, reloc_howto_type *howto, asection *section, const char *name, union etree_union *addend) { lang_reloc_statement_type *p = new_stat (lang_reloc_statement, stat_ptr); p->reloc = reloc; p->howto = howto; p->section = section; p->name = name; p->addend_exp = addend; p->addend_value = 0; p->output_section = NULL; p->output_offset = 0; } lang_assignment_statement_type * lang_add_assignment (etree_type *exp) { lang_assignment_statement_type *new_stmt; new_stmt = new_stat (lang_assignment_statement, stat_ptr); new_stmt->exp = exp; return new_stmt; } void lang_add_attribute (enum statement_enum attribute) { new_statement (attribute, sizeof (lang_statement_header_type), stat_ptr); } void lang_startup (const char *name) { if (first_file->filename != NULL) { einfo (_("%F%P: multiple STARTUP files\n")); } first_file->filename = name; first_file->local_sym_name = name; first_file->flags.real = true; } void lang_float (bool maybe) { lang_float_flag = maybe; } /* Work out the load- and run-time regions from a script statement, and store them in *LMA_REGION and *REGION respectively. MEMSPEC is the name of the run-time region, or the value of DEFAULT_MEMORY_REGION if the statement didn't specify one. LMA_MEMSPEC is the name of the load-time region, or null if the statement didn't specify one.HAVE_LMA_P is TRUE if the statement had an explicit load address. It is an error to specify both a load region and a load address. */ static void lang_get_regions (lang_memory_region_type **region, lang_memory_region_type **lma_region, const char *memspec, const char *lma_memspec, bool have_lma, bool have_vma) { *lma_region = lang_memory_region_lookup (lma_memspec, false); /* If no runtime region or VMA has been specified, but the load region has been specified, then use the load region for the runtime region as well. */ if (lma_memspec != NULL && !have_vma && strcmp (memspec, DEFAULT_MEMORY_REGION) == 0) *region = *lma_region; else *region = lang_memory_region_lookup (memspec, false); if (have_lma && lma_memspec != 0) einfo (_("%X%P:%pS: section has both a load address and a load region\n"), NULL); } void lang_leave_output_section_statement (fill_type *fill, const char *memspec, lang_output_section_phdr_list *phdrs, const char *lma_memspec) { lang_get_regions (¤t_section->region, ¤t_section->lma_region, memspec, lma_memspec, current_section->load_base != NULL, current_section->addr_tree != NULL); current_section->fill = fill; current_section->phdrs = phdrs; pop_stat_ptr (); } /* Set the output format type. -oformat overrides scripts. */ void lang_add_output_format (const char *format, const char *big, const char *little, int from_script) { if (output_target == NULL || !from_script) { if (command_line.endian == ENDIAN_BIG && big != NULL) format = big; else if (command_line.endian == ENDIAN_LITTLE && little != NULL) format = little; output_target = format; } } void lang_add_insert (const char *where, int is_before) { lang_insert_statement_type *new_stmt; new_stmt = new_stat (lang_insert_statement, stat_ptr); new_stmt->where = where; new_stmt->is_before = is_before; saved_script_handle = previous_script_handle; } /* Enter a group. This creates a new lang_group_statement, and sets stat_ptr to build new statements within the group. */ void lang_enter_group (void) { lang_group_statement_type *g; g = new_stat (lang_group_statement, stat_ptr); lang_list_init (&g->children); push_stat_ptr (&g->children); } /* Leave a group. This just resets stat_ptr to start writing to the regular list of statements again. Note that this will not work if groups can occur inside anything else which can adjust stat_ptr, but currently they can't. */ void lang_leave_group (void) { pop_stat_ptr (); } /* Add a new program header. This is called for each entry in a PHDRS command in a linker script. */ void lang_new_phdr (const char *name, etree_type *type, bool filehdr, bool phdrs, etree_type *at, etree_type *flags) { struct lang_phdr *n, **pp; bool hdrs; n = stat_alloc (sizeof (struct lang_phdr)); n->next = NULL; n->name = name; n->type = exp_get_vma (type, 0, "program header type"); n->filehdr = filehdr; n->phdrs = phdrs; n->at = at; n->flags = flags; hdrs = n->type == 1 && (phdrs || filehdr); for (pp = &lang_phdr_list; *pp != NULL; pp = &(*pp)->next) if (hdrs && (*pp)->type == 1 && !((*pp)->filehdr || (*pp)->phdrs)) { einfo (_("%X%P:%pS: PHDRS and FILEHDR are not supported" " when prior PT_LOAD headers lack them\n"), NULL); hdrs = false; } *pp = n; } /* Record the program header information in the output BFD. FIXME: We should not be calling an ELF specific function here. */ static void lang_record_phdrs (void) { unsigned int alc; asection **secs; lang_output_section_phdr_list *last; struct lang_phdr *l; lang_output_section_statement_type *os; alc = 10; secs = (asection **) xmalloc (alc * sizeof (asection *)); last = NULL; for (l = lang_phdr_list; l != NULL; l = l->next) { unsigned int c; flagword flags; bfd_vma at; c = 0; for (os = (void *) lang_os_list.head; os != NULL; os = os->next) { lang_output_section_phdr_list *pl; if (os->constraint < 0) continue; pl = os->phdrs; if (pl != NULL) last = pl; else { if (os->sectype == noload_section || os->bfd_section == NULL || (os->bfd_section->flags & SEC_ALLOC) == 0) continue; /* Don't add orphans to PT_INTERP header. */ if (l->type == 3) continue; if (last == NULL) { lang_output_section_statement_type *tmp_os; /* If we have not run across a section with a program header assigned to it yet, then scan forwards to find one. This prevents inconsistencies in the linker's behaviour when a script has specified just a single header and there are sections in that script which are not assigned to it, and which occur before the first use of that header. See here for more details: http://sourceware.org/ml/binutils/2007-02/msg00291.html */ for (tmp_os = os; tmp_os; tmp_os = tmp_os->next) if (tmp_os->phdrs) { last = tmp_os->phdrs; break; } if (last == NULL) einfo (_("%F%P: no sections assigned to phdrs\n")); } pl = last; } if (os->bfd_section == NULL) continue; for (; pl != NULL; pl = pl->next) { if (strcmp (pl->name, l->name) == 0) { if (c >= alc) { alc *= 2; secs = (asection **) xrealloc (secs, alc * sizeof (asection *)); } secs[c] = os->bfd_section; ++c; pl->used = true; } } } if (l->flags == NULL) flags = 0; else flags = exp_get_vma (l->flags, 0, "phdr flags"); if (l->at == NULL) at = 0; else at = exp_get_vma (l->at, 0, "phdr load address"); if (!bfd_record_phdr (link_info.output_bfd, l->type, l->flags != NULL, flags, l->at != NULL, at, l->filehdr, l->phdrs, c, secs)) einfo (_("%F%P: bfd_record_phdr failed: %E\n")); } free (secs); /* Make sure all the phdr assignments succeeded. */ for (os = (void *) lang_os_list.head; os != NULL; os = os->next) { lang_output_section_phdr_list *pl; if (os->constraint < 0 || os->bfd_section == NULL) continue; for (pl = os->phdrs; pl != NULL; pl = pl->next) if (!pl->used && strcmp (pl->name, "NONE") != 0) einfo (_("%X%P: section `%s' assigned to non-existent phdr `%s'\n"), os->name, pl->name); } } /* Record a list of sections which may not be cross referenced. */ void lang_add_nocrossref (lang_nocrossref_type *l) { struct lang_nocrossrefs *n; n = (struct lang_nocrossrefs *) xmalloc (sizeof *n); n->next = nocrossref_list; n->list = l; n->onlyfirst = false; nocrossref_list = n; /* Set notice_all so that we get informed about all symbols. */ link_info.notice_all = true; } /* Record a section that cannot be referenced from a list of sections. */ void lang_add_nocrossref_to (lang_nocrossref_type *l) { lang_add_nocrossref (l); nocrossref_list->onlyfirst = true; } /* Overlay handling. We handle overlays with some static variables. */ /* The overlay virtual address. */ static etree_type *overlay_vma; /* And subsection alignment. */ static etree_type *overlay_subalign; /* An expression for the maximum section size seen so far. */ static etree_type *overlay_max; /* A list of all the sections in this overlay. */ struct overlay_list { struct overlay_list *next; lang_output_section_statement_type *os; }; static struct overlay_list *overlay_list; /* Start handling an overlay. */ void lang_enter_overlay (etree_type *vma_expr, etree_type *subalign) { /* The grammar should prevent nested overlays from occurring. */ ASSERT (overlay_vma == NULL && overlay_subalign == NULL && overlay_max == NULL); overlay_vma = vma_expr; overlay_subalign = subalign; } /* Start a section in an overlay. We handle this by calling lang_enter_output_section_statement with the correct VMA. lang_leave_overlay sets up the LMA and memory regions. */ void lang_enter_overlay_section (const char *name) { struct overlay_list *n; etree_type *size; lang_enter_output_section_statement (name, overlay_vma, overlay_section, 0, 0, overlay_subalign, 0, 0, 0); /* If this is the first section, then base the VMA of future sections on this one. This will work correctly even if `.' is used in the addresses. */ if (overlay_list == NULL) overlay_vma = exp_nameop (ADDR, name); /* Remember the section. */ n = (struct overlay_list *) xmalloc (sizeof *n); n->os = current_section; n->next = overlay_list; overlay_list = n; size = exp_nameop (SIZEOF, name); /* Arrange to work out the maximum section end address. */ if (overlay_max == NULL) overlay_max = size; else overlay_max = exp_binop (MAX_K, overlay_max, size); } /* Finish a section in an overlay. There isn't any special to do here. */ void lang_leave_overlay_section (fill_type *fill, lang_output_section_phdr_list *phdrs) { const char *name; char *clean, *s2; const char *s1; char *buf; name = current_section->name; /* For now, assume that DEFAULT_MEMORY_REGION is the run-time memory region and that no load-time region has been specified. It doesn't really matter what we say here, since lang_leave_overlay will override it. */ lang_leave_output_section_statement (fill, DEFAULT_MEMORY_REGION, phdrs, 0); /* Define the magic symbols. */ clean = (char *) xmalloc (strlen (name) + 1); s2 = clean; for (s1 = name; *s1 != '\0'; s1++) if (ISALNUM (*s1) || *s1 == '_') *s2++ = *s1; *s2 = '\0'; buf = (char *) xmalloc (strlen (clean) + sizeof "__load_start_"); sprintf (buf, "__load_start_%s", clean); lang_add_assignment (exp_provide (buf, exp_nameop (LOADADDR, name), false)); buf = (char *) xmalloc (strlen (clean) + sizeof "__load_stop_"); sprintf (buf, "__load_stop_%s", clean); lang_add_assignment (exp_provide (buf, exp_binop ('+', exp_nameop (LOADADDR, name), exp_nameop (SIZEOF, name)), false)); free (clean); } /* Finish an overlay. If there are any overlay wide settings, this looks through all the sections in the overlay and sets them. */ void lang_leave_overlay (etree_type *lma_expr, int nocrossrefs, fill_type *fill, const char *memspec, lang_output_section_phdr_list *phdrs, const char *lma_memspec) { lang_memory_region_type *region; lang_memory_region_type *lma_region; struct overlay_list *l; lang_nocrossref_type *nocrossref; lang_get_regions (®ion, &lma_region, memspec, lma_memspec, lma_expr != NULL, false); nocrossref = NULL; /* After setting the size of the last section, set '.' to end of the overlay region. */ if (overlay_list != NULL) { overlay_list->os->update_dot = 1; overlay_list->os->update_dot_tree = exp_assign (".", exp_binop ('+', overlay_vma, overlay_max), false); } l = overlay_list; while (l != NULL) { struct overlay_list *next; if (fill != NULL && l->os->fill == NULL) l->os->fill = fill; l->os->region = region; l->os->lma_region = lma_region; /* The first section has the load address specified in the OVERLAY statement. The rest are worked out from that. The base address is not needed (and should be null) if an LMA region was specified. */ if (l->next == 0) { l->os->load_base = lma_expr; l->os->sectype = first_overlay_section; } if (phdrs != NULL && l->os->phdrs == NULL) l->os->phdrs = phdrs; if (nocrossrefs) { lang_nocrossref_type *nc; nc = (lang_nocrossref_type *) xmalloc (sizeof *nc); nc->name = l->os->name; nc->next = nocrossref; nocrossref = nc; } next = l->next; free (l); l = next; } if (nocrossref != NULL) lang_add_nocrossref (nocrossref); overlay_vma = NULL; overlay_list = NULL; overlay_max = NULL; overlay_subalign = NULL; } /* Version handling. This is only useful for ELF. */ /* If PREV is NULL, return first version pattern matching particular symbol. If PREV is non-NULL, return first version pattern matching particular symbol after PREV (previously returned by lang_vers_match). */ static struct bfd_elf_version_expr * lang_vers_match (struct bfd_elf_version_expr_head *head, struct bfd_elf_version_expr *prev, const char *sym) { const char *c_sym; const char *cxx_sym = sym; const char *java_sym = sym; struct bfd_elf_version_expr *expr = NULL; enum demangling_styles curr_style; curr_style = CURRENT_DEMANGLING_STYLE; cplus_demangle_set_style (no_demangling); c_sym = bfd_demangle (link_info.output_bfd, sym, DMGL_NO_OPTS); if (!c_sym) c_sym = sym; cplus_demangle_set_style (curr_style); if (head->mask & BFD_ELF_VERSION_CXX_TYPE) { cxx_sym = bfd_demangle (link_info.output_bfd, sym, DMGL_PARAMS | DMGL_ANSI); if (!cxx_sym) cxx_sym = sym; } if (head->mask & BFD_ELF_VERSION_JAVA_TYPE) { java_sym = bfd_demangle (link_info.output_bfd, sym, DMGL_JAVA); if (!java_sym) java_sym = sym; } if (head->htab && (prev == NULL || prev->literal)) { struct bfd_elf_version_expr e; switch (prev ? prev->mask : 0) { case 0: if (head->mask & BFD_ELF_VERSION_C_TYPE) { e.pattern = c_sym; expr = (struct bfd_elf_version_expr *) htab_find ((htab_t) head->htab, &e); while (expr && strcmp (expr->pattern, c_sym) == 0) if (expr->mask == BFD_ELF_VERSION_C_TYPE) goto out_ret; else expr = expr->next; } /* Fallthrough */ case BFD_ELF_VERSION_C_TYPE: if (head->mask & BFD_ELF_VERSION_CXX_TYPE) { e.pattern = cxx_sym; expr = (struct bfd_elf_version_expr *) htab_find ((htab_t) head->htab, &e); while (expr && strcmp (expr->pattern, cxx_sym) == 0) if (expr->mask == BFD_ELF_VERSION_CXX_TYPE) goto out_ret; else expr = expr->next; } /* Fallthrough */ case BFD_ELF_VERSION_CXX_TYPE: if (head->mask & BFD_ELF_VERSION_JAVA_TYPE) { e.pattern = java_sym; expr = (struct bfd_elf_version_expr *) htab_find ((htab_t) head->htab, &e); while (expr && strcmp (expr->pattern, java_sym) == 0) if (expr->mask == BFD_ELF_VERSION_JAVA_TYPE) goto out_ret; else expr = expr->next; } /* Fallthrough */ default: break; } } /* Finally, try the wildcards. */ if (prev == NULL || prev->literal) expr = head->remaining; else expr = prev->next; for (; expr; expr = expr->next) { const char *s; if (!expr->pattern) continue; if (expr->pattern[0] == '*' && expr->pattern[1] == '\0') break; if (expr->mask == BFD_ELF_VERSION_JAVA_TYPE) s = java_sym; else if (expr->mask == BFD_ELF_VERSION_CXX_TYPE) s = cxx_sym; else s = c_sym; if (fnmatch (expr->pattern, s, 0) == 0) break; } out_ret: if (c_sym != sym) free ((char *) c_sym); if (cxx_sym != sym) free ((char *) cxx_sym); if (java_sym != sym) free ((char *) java_sym); return expr; } /* Return NULL if the PATTERN argument is a glob pattern, otherwise, return a pointer to the symbol name with any backslash quotes removed. */ static const char * realsymbol (const char *pattern) { const char *p; bool changed = false, backslash = false; char *s, *symbol = (char *) xmalloc (strlen (pattern) + 1); for (p = pattern, s = symbol; *p != '\0'; ++p) { /* It is a glob pattern only if there is no preceding backslash. */ if (backslash) { /* Remove the preceding backslash. */ *(s - 1) = *p; backslash = false; changed = true; } else { if (*p == '?' || *p == '*' || *p == '[') { free (symbol); return NULL; } *s++ = *p; backslash = *p == '\\'; } } if (changed) { *s = '\0'; return symbol; } else { free (symbol); return pattern; } } /* This is called for each variable name or match expression. NEW_NAME is the name of the symbol to match, or, if LITERAL_P is FALSE, a glob pattern to be matched against symbol names. */ struct bfd_elf_version_expr * lang_new_vers_pattern (struct bfd_elf_version_expr *orig, const char *new_name, const char *lang, bool literal_p) { struct bfd_elf_version_expr *ret; ret = (struct bfd_elf_version_expr *) xmalloc (sizeof *ret); ret->next = orig; ret->symver = 0; ret->script = 0; ret->literal = true; ret->pattern = literal_p ? new_name : realsymbol (new_name); if (ret->pattern == NULL) { ret->pattern = new_name; ret->literal = false; } if (lang == NULL || strcasecmp (lang, "C") == 0) ret->mask = BFD_ELF_VERSION_C_TYPE; else if (strcasecmp (lang, "C++") == 0) ret->mask = BFD_ELF_VERSION_CXX_TYPE; else if (strcasecmp (lang, "Java") == 0) ret->mask = BFD_ELF_VERSION_JAVA_TYPE; else { einfo (_("%X%P: unknown language `%s' in version information\n"), lang); ret->mask = BFD_ELF_VERSION_C_TYPE; } return ldemul_new_vers_pattern (ret); } /* This is called for each set of variable names and match expressions. */ struct bfd_elf_version_tree * lang_new_vers_node (struct bfd_elf_version_expr *globals, struct bfd_elf_version_expr *locals) { struct bfd_elf_version_tree *ret; ret = (struct bfd_elf_version_tree *) xcalloc (1, sizeof *ret); ret->globals.list = globals; ret->locals.list = locals; ret->match = lang_vers_match; ret->name_indx = (unsigned int) -1; return ret; } /* This static variable keeps track of version indices. */ static int version_index; static hashval_t version_expr_head_hash (const void *p) { const struct bfd_elf_version_expr *e = (const struct bfd_elf_version_expr *) p; return htab_hash_string (e->pattern); } static int version_expr_head_eq (const void *p1, const void *p2) { const struct bfd_elf_version_expr *e1 = (const struct bfd_elf_version_expr *) p1; const struct bfd_elf_version_expr *e2 = (const struct bfd_elf_version_expr *) p2; return strcmp (e1->pattern, e2->pattern) == 0; } static void lang_finalize_version_expr_head (struct bfd_elf_version_expr_head *head) { size_t count = 0; struct bfd_elf_version_expr *e, *next; struct bfd_elf_version_expr **list_loc, **remaining_loc; for (e = head->list; e; e = e->next) { if (e->literal) count++; head->mask |= e->mask; } if (count) { head->htab = htab_create (count * 2, version_expr_head_hash, version_expr_head_eq, NULL); list_loc = &head->list; remaining_loc = &head->remaining; for (e = head->list; e; e = next) { next = e->next; if (!e->literal) { *remaining_loc = e; remaining_loc = &e->next; } else { void **loc = htab_find_slot ((htab_t) head->htab, e, INSERT); if (*loc) { struct bfd_elf_version_expr *e1, *last; e1 = (struct bfd_elf_version_expr *) *loc; last = NULL; do { if (e1->mask == e->mask) { last = NULL; break; } last = e1; e1 = e1->next; } while (e1 && strcmp (e1->pattern, e->pattern) == 0); if (last == NULL) { /* This is a duplicate. */ /* FIXME: Memory leak. Sometimes pattern is not xmalloced alone, but in larger chunk of memory. */ /* free (e->pattern); */ free (e); } else { e->next = last->next; last->next = e; } } else { *loc = e; *list_loc = e; list_loc = &e->next; } } } *remaining_loc = NULL; *list_loc = head->remaining; } else head->remaining = head->list; } /* This is called when we know the name and dependencies of the version. */ void lang_register_vers_node (const char *name, struct bfd_elf_version_tree *version, struct bfd_elf_version_deps *deps) { struct bfd_elf_version_tree *t, **pp; struct bfd_elf_version_expr *e1; if (name == NULL) name = ""; if (link_info.version_info != NULL && (name[0] == '\0' || link_info.version_info->name[0] == '\0')) { einfo (_("%X%P: anonymous version tag cannot be combined" " with other version tags\n")); free (version); return; } /* Make sure this node has a unique name. */ for (t = link_info.version_info; t != NULL; t = t->next) if (strcmp (t->name, name) == 0) einfo (_("%X%P: duplicate version tag `%s'\n"), name); lang_finalize_version_expr_head (&version->globals); lang_finalize_version_expr_head (&version->locals); /* Check the global and local match names, and make sure there aren't any duplicates. */ for (e1 = version->globals.list; e1 != NULL; e1 = e1->next) { for (t = link_info.version_info; t != NULL; t = t->next) { struct bfd_elf_version_expr *e2; if (t->locals.htab && e1->literal) { e2 = (struct bfd_elf_version_expr *) htab_find ((htab_t) t->locals.htab, e1); while (e2 && strcmp (e1->pattern, e2->pattern) == 0) { if (e1->mask == e2->mask) einfo (_("%X%P: duplicate expression `%s'" " in version information\n"), e1->pattern); e2 = e2->next; } } else if (!e1->literal) for (e2 = t->locals.remaining; e2 != NULL; e2 = e2->next) if (strcmp (e1->pattern, e2->pattern) == 0 && e1->mask == e2->mask) einfo (_("%X%P: duplicate expression `%s'" " in version information\n"), e1->pattern); } } for (e1 = version->locals.list; e1 != NULL; e1 = e1->next) { for (t = link_info.version_info; t != NULL; t = t->next) { struct bfd_elf_version_expr *e2; if (t->globals.htab && e1->literal) { e2 = (struct bfd_elf_version_expr *) htab_find ((htab_t) t->globals.htab, e1); while (e2 && strcmp (e1->pattern, e2->pattern) == 0) { if (e1->mask == e2->mask) einfo (_("%X%P: duplicate expression `%s'" " in version information\n"), e1->pattern); e2 = e2->next; } } else if (!e1->literal) for (e2 = t->globals.remaining; e2 != NULL; e2 = e2->next) if (strcmp (e1->pattern, e2->pattern) == 0 && e1->mask == e2->mask) einfo (_("%X%P: duplicate expression `%s'" " in version information\n"), e1->pattern); } } version->deps = deps; version->name = name; if (name[0] != '\0') { ++version_index; version->vernum = version_index; } else version->vernum = 0; for (pp = &link_info.version_info; *pp != NULL; pp = &(*pp)->next) ; *pp = version; } /* This is called when we see a version dependency. */ struct bfd_elf_version_deps * lang_add_vers_depend (struct bfd_elf_version_deps *list, const char *name) { struct bfd_elf_version_deps *ret; struct bfd_elf_version_tree *t; ret = (struct bfd_elf_version_deps *) xmalloc (sizeof *ret); ret->next = list; for (t = link_info.version_info; t != NULL; t = t->next) { if (strcmp (t->name, name) == 0) { ret->version_needed = t; return ret; } } einfo (_("%X%P: unable to find version dependency `%s'\n"), name); ret->version_needed = NULL; return ret; } static void lang_do_version_exports_section (void) { struct bfd_elf_version_expr *greg = NULL, *lreg; LANG_FOR_EACH_INPUT_STATEMENT (is) { asection *sec = bfd_get_section_by_name (is->the_bfd, ".exports"); char *contents, *p; bfd_size_type len; if (sec == NULL) continue; len = sec->size; contents = (char *) xmalloc (len); if (!bfd_get_section_contents (is->the_bfd, sec, contents, 0, len)) einfo (_("%X%P: unable to read .exports section contents\n"), sec); p = contents; while (p < contents + len) { greg = lang_new_vers_pattern (greg, p, NULL, false); p = strchr (p, '\0') + 1; } /* Do not free the contents, as we used them creating the regex. */ /* Do not include this section in the link. */ sec->flags |= SEC_EXCLUDE | SEC_KEEP; } lreg = lang_new_vers_pattern (NULL, "*", NULL, false); lang_register_vers_node (command_line.version_exports_section, lang_new_vers_node (greg, lreg), NULL); } /* Evaluate LENGTH and ORIGIN parts of MEMORY spec. This is initially called with UPDATE_REGIONS_P set to FALSE, in this case no errors are thrown, however, references to symbols in the origin and length fields will be pushed into the symbol table, this allows PROVIDE statements to then provide these symbols. This function is called a second time with UPDATE_REGIONS_P set to TRUE, this time the we update the actual region data structures, and throw errors if missing symbols are encountered. */ static void lang_do_memory_regions (bool update_regions_p) { lang_memory_region_type *r = lang_memory_region_list; for (; r != NULL; r = r->next) { if (r->origin_exp) { exp_fold_tree_no_dot (r->origin_exp); if (update_regions_p) { if (expld.result.valid_p) { r->origin = expld.result.value; r->current = r->origin; } else einfo (_("%P: invalid origin for memory region %s\n"), r->name_list.name); } } if (r->length_exp) { exp_fold_tree_no_dot (r->length_exp); if (update_regions_p) { if (expld.result.valid_p) r->length = expld.result.value; else einfo (_("%P: invalid length for memory region %s\n"), r->name_list.name); } } } } void lang_add_unique (const char *name) { struct unique_sections *ent; for (ent = unique_section_list; ent; ent = ent->next) if (strcmp (ent->name, name) == 0) return; ent = (struct unique_sections *) xmalloc (sizeof *ent); ent->name = xstrdup (name); ent->next = unique_section_list; unique_section_list = ent; } /* Append the list of dynamic symbols to the existing one. */ void lang_append_dynamic_list (struct bfd_elf_dynamic_list **list_p, struct bfd_elf_version_expr *dynamic) { if (*list_p) { struct bfd_elf_version_expr *tail; for (tail = dynamic; tail->next != NULL; tail = tail->next) ; tail->next = (*list_p)->head.list; (*list_p)->head.list = dynamic; } else { struct bfd_elf_dynamic_list *d; d = (struct bfd_elf_dynamic_list *) xcalloc (1, sizeof *d); d->head.list = dynamic; d->match = lang_vers_match; *list_p = d; } } /* Append the list of C++ typeinfo dynamic symbols to the existing one. */ void lang_append_dynamic_list_cpp_typeinfo (void) { const char *symbols[] = { "typeinfo name for*", "typeinfo for*" }; struct bfd_elf_version_expr *dynamic = NULL; unsigned int i; for (i = 0; i < ARRAY_SIZE (symbols); i++) dynamic = lang_new_vers_pattern (dynamic, symbols [i], "C++", false); lang_append_dynamic_list (&link_info.dynamic_list, dynamic); } /* Append the list of C++ operator new and delete dynamic symbols to the existing one. */ void lang_append_dynamic_list_cpp_new (void) { const char *symbols[] = { "operator new*", "operator delete*" }; struct bfd_elf_version_expr *dynamic = NULL; unsigned int i; for (i = 0; i < ARRAY_SIZE (symbols); i++) dynamic = lang_new_vers_pattern (dynamic, symbols [i], "C++", false); lang_append_dynamic_list (&link_info.dynamic_list, dynamic); } /* Scan a space and/or comma separated string of features. */ void lang_ld_feature (char *str) { char *p, *q; p = str; while (*p) { char sep; while (*p == ',' || ISSPACE (*p)) ++p; if (!*p) break; q = p + 1; while (*q && *q != ',' && !ISSPACE (*q)) ++q; sep = *q; *q = 0; if (strcasecmp (p, "SANE_EXPR") == 0) config.sane_expr = true; else einfo (_("%X%P: unknown feature `%s'\n"), p); *q = sep; p = q; } } /* Pretty print memory amount. */ static void lang_print_memory_size (uint64_t sz) { if ((sz & 0x3fffffff) == 0) printf ("%10" PRIu64 " GB", sz >> 30); else if ((sz & 0xfffff) == 0) printf ("%10" PRIu64 " MB", sz >> 20); else if ((sz & 0x3ff) == 0) printf ("%10" PRIu64 " KB", sz >> 10); else printf (" %10" PRIu64 " B", sz); } /* Implement --print-memory-usage: disply per region memory usage. */ void lang_print_memory_usage (void) { lang_memory_region_type *r; printf ("Memory region Used Size Region Size %%age Used\n"); for (r = lang_memory_region_list; r->next != NULL; r = r->next) { bfd_vma used_length = r->current - r->origin; printf ("%16s: ",r->name_list.name); lang_print_memory_size (used_length); lang_print_memory_size (r->length); if (r->length != 0) { double percent = used_length * 100.0 / r->length; printf (" %6.2f%%", percent); } printf ("\n"); } }