# This shell script emits a C file. -*- C -*- # It does some substitutions. if [ -z "$MACHINE" ]; then OUTPUT_ARCH=${ARCH} else OUTPUT_ARCH=${ARCH}:${MACHINE} fi cat >e${EMULATION_NAME}.c < #include "ldlex.h" #include "ldmisc.h" #include "ldctor.h" #include "coff/internal.h" #include "../bfd/libcoff.h" #define TARGET_IS_${EMULATION_NAME} static struct internal_extra_pe_aouthdr pe; static int dll; extern const char *output_filename; static void gld_${EMULATION_NAME}_before_parse (void) { const bfd_arch_info_type *arch = bfd_scan_arch ("${OUTPUT_ARCH}"); if (arch) { ldfile_output_architecture = arch->arch; ldfile_output_machine = arch->mach; ldfile_output_machine_name = arch->printable_name; } else ldfile_output_architecture = bfd_arch_${ARCH}; output_filename = "a.exe"; } /* PE format extra command line options. */ /* Used for setting flags in the PE header. */ #define OPTION_BASE_FILE (300 + 1) #define OPTION_DLL (OPTION_BASE_FILE + 1) #define OPTION_FILE_ALIGNMENT (OPTION_DLL + 1) #define OPTION_IMAGE_BASE (OPTION_FILE_ALIGNMENT + 1) #define OPTION_MAJOR_IMAGE_VERSION (OPTION_IMAGE_BASE + 1) #define OPTION_MAJOR_OS_VERSION (OPTION_MAJOR_IMAGE_VERSION + 1) #define OPTION_MAJOR_SUBSYSTEM_VERSION (OPTION_MAJOR_OS_VERSION + 1) #define OPTION_MINOR_IMAGE_VERSION (OPTION_MAJOR_SUBSYSTEM_VERSION + 1) #define OPTION_MINOR_OS_VERSION (OPTION_MINOR_IMAGE_VERSION + 1) #define OPTION_MINOR_SUBSYSTEM_VERSION (OPTION_MINOR_OS_VERSION + 1) #define OPTION_SECTION_ALIGNMENT (OPTION_MINOR_SUBSYSTEM_VERSION + 1) #define OPTION_STACK (OPTION_SECTION_ALIGNMENT + 1) #define OPTION_SUBSYSTEM (OPTION_STACK + 1) #define OPTION_HEAP (OPTION_SUBSYSTEM + 1) static void gld${EMULATION_NAME}_add_options (int ns ATTRIBUTE_UNUSED, char **shortopts ATTRIBUTE_UNUSED, int nl, struct option **longopts, int nrl ATTRIBUTE_UNUSED, struct option **really_longopts ATTRIBUTE_UNUSED) { static const struct option xtra_long[] = { /* PE options */ {"base-file", required_argument, NULL, OPTION_BASE_FILE}, {"dll", no_argument, NULL, OPTION_DLL}, {"file-alignment", required_argument, NULL, OPTION_FILE_ALIGNMENT}, {"heap", required_argument, NULL, OPTION_HEAP}, {"image-base", required_argument, NULL, OPTION_IMAGE_BASE}, {"major-image-version", required_argument, NULL, OPTION_MAJOR_IMAGE_VERSION}, {"major-os-version", required_argument, NULL, OPTION_MAJOR_OS_VERSION}, {"major-subsystem-version", required_argument, NULL, OPTION_MAJOR_SUBSYSTEM_VERSION}, {"minor-image-version", required_argument, NULL, OPTION_MINOR_IMAGE_VERSION}, {"minor-os-version", required_argument, NULL, OPTION_MINOR_OS_VERSION}, {"minor-subsystem-version", required_argument, NULL, OPTION_MINOR_SUBSYSTEM_VERSION}, {"section-alignment", required_argument, NULL, OPTION_SECTION_ALIGNMENT}, {"stack", required_argument, NULL, OPTION_STACK}, {"subsystem", required_argument, NULL, OPTION_SUBSYSTEM}, {NULL, no_argument, NULL, 0} }; *longopts = (struct option *) xrealloc (*longopts, nl * sizeof (struct option) + sizeof (xtra_long)); memcpy (*longopts + nl, &xtra_long, sizeof (xtra_long)); } /* PE/WIN32; added routines to get the subsystem type, heap and/or stack parameters which may be input from the command line */ typedef struct { void *ptr; int size; int value; char *symbol; int inited; } definfo; #define D(field,symbol,def) {&pe.field,sizeof(pe.field), def, symbol,0} static definfo init[] = { /* imagebase must be first */ #define IMAGEBASEOFF 0 D(ImageBase,"__image_base__", BEOS_EXE_IMAGE_BASE), #define DLLOFF 1 {&dll, sizeof(dll), 0, "__dll__", 0}, D(SectionAlignment,"__section_alignment__", PE_DEF_SECTION_ALIGNMENT), D(FileAlignment,"__file_alignment__", PE_DEF_FILE_ALIGNMENT), D(MajorOperatingSystemVersion,"__major_os_version__", 4), D(MinorOperatingSystemVersion,"__minor_os_version__", 0), D(MajorImageVersion,"__major_image_version__", 1), D(MinorImageVersion,"__minor_image_version__", 0), D(MajorSubsystemVersion,"__major_subsystem_version__", 4), D(MinorSubsystemVersion,"__minor_subsystem_version__", 0), D(Subsystem,"__subsystem__", 3), D(SizeOfStackReserve,"__size_of_stack_reserve__", 0x2000000), D(SizeOfStackCommit,"__size_of_stack_commit__", 0x1000), D(SizeOfHeapReserve,"__size_of_heap_reserve__", 0x100000), D(SizeOfHeapCommit,"__size_of_heap_commit__", 0x1000), D(LoaderFlags,"__loader_flags__", 0x0), { NULL, 0, 0, NULL, 0 } }; static void set_pe_name (char *name, long val) { int i; /* Find the name and set it. */ for (i = 0; init[i].ptr; i++) { if (strcmp (name, init[i].symbol) == 0) { init[i].value = val; init[i].inited = 1; return; } } abort(); } static void set_pe_subsystem (void) { const char *sver; int len; int i; static const struct { const char *name; const int value; const char *entry; } v[] = { { "native", 1, "_NtProcessStartup" }, { "windows", 2, "_WinMainCRTStartup" }, { "wwindows", 2, "_wWinMainCRTStartup" }, { "console", 3, "_mainCRTStartup" }, { "wconsole", 3, "_wmainCRTStartup" }, #if 0 /* The Microsoft linker does not recognize this. */ { "os2", 5, "" }, #endif { "posix", 7, "___PosixProcessStartup"}, { 0, 0, 0 } }; sver = strchr (optarg, ':'); if (sver == NULL) len = strlen (optarg); else { char *end; len = sver - optarg; set_pe_name ("__major_subsystem_version__", strtoul (sver + 1, &end, 0)); if (*end == '.') set_pe_name ("__minor_subsystem_version__", strtoul (end + 1, &end, 0)); if (*end != '\0') einfo ("%P: warning: bad version number in -subsystem option\n"); } for (i = 0; v[i].name; i++) { if (strncmp (optarg, v[i].name, len) == 0 && v[i].name[len] == '\0') { set_pe_name ("__subsystem__", v[i].value); /* If the subsystem is windows, we use a different entry point. We also register the entry point as an undefined symbol. from lang_add_entry() The reason we do this is so that the user doesn't have to because they would have to use the -u switch if they were specifying an entry point other than _mainCRTStartup. Specifically, if creating a windows application, entry point _WinMainCRTStartup must be specified. What I have found for non console applications (entry not _mainCRTStartup) is that the .obj that contains mainCRTStartup is brought in since it is the first encountered in libc.lib and it has other symbols in it which will be pulled in by the link process. To avoid this, adding -u with the entry point name specified forces the correct .obj to be used. We can avoid making the user do this by always adding the entry point name as an undefined symbol. */ lang_add_entry (v[i].entry, 1); return; } } einfo ("%P%F: invalid subsystem type %s\n", optarg); } static void set_pe_value (char *name) { char *end; set_pe_name (name, strtoul (optarg, &end, 0)); if (end == optarg) { einfo ("%P%F: invalid hex number for PE parameter '%s'\n", optarg); } optarg = end; } static void set_pe_stack_heap (char *resname, char *comname) { set_pe_value (resname); if (*optarg == ',') { optarg++; set_pe_value (comname); } else if (*optarg) { einfo ("%P%F: strange hex info for PE parameter '%s'\n", optarg); } } static bfd_boolean gld${EMULATION_NAME}_handle_option (int optc) { switch (optc) { default: return FALSE; case OPTION_BASE_FILE: link_info.base_file = fopen (optarg, FOPEN_WB); if (link_info.base_file == NULL) { fprintf (stderr, "%s: Can't open base file %s\n", program_name, optarg); xexit (1); } break; /* PE options */ case OPTION_HEAP: set_pe_stack_heap ("__size_of_heap_reserve__", "__size_of_heap_commit__"); break; case OPTION_STACK: set_pe_stack_heap ("__size_of_stack_reserve__", "__size_of_stack_commit__"); break; case OPTION_SUBSYSTEM: set_pe_subsystem (); break; case OPTION_MAJOR_OS_VERSION: set_pe_value ("__major_os_version__"); break; case OPTION_MINOR_OS_VERSION: set_pe_value ("__minor_os_version__"); break; case OPTION_MAJOR_SUBSYSTEM_VERSION: set_pe_value ("__major_subsystem_version__"); break; case OPTION_MINOR_SUBSYSTEM_VERSION: set_pe_value ("__minor_subsystem_version__"); break; case OPTION_MAJOR_IMAGE_VERSION: set_pe_value ("__major_image_version__"); break; case OPTION_MINOR_IMAGE_VERSION: set_pe_value ("__minor_image_version__"); break; case OPTION_FILE_ALIGNMENT: set_pe_value ("__file_alignment__"); break; case OPTION_SECTION_ALIGNMENT: set_pe_value ("__section_alignment__"); break; case OPTION_DLL: set_pe_name ("__dll__", 1); break; case OPTION_IMAGE_BASE: set_pe_value ("__image_base__"); break; } return TRUE; } /* Assign values to the special symbols before the linker script is read. */ static void gld_${EMULATION_NAME}_set_symbols (void) { /* Run through and invent symbols for all the names and insert the defaults. */ int j; lang_statement_list_type *save; if (!init[IMAGEBASEOFF].inited) { if (link_info.relocatable) init[IMAGEBASEOFF].value = 0; else if (init[DLLOFF].value) init[IMAGEBASEOFF].value = BEOS_DLL_IMAGE_BASE; else init[IMAGEBASEOFF].value = BEOS_EXE_IMAGE_BASE; } /* Don't do any symbol assignments if this is a relocatable link. */ if (link_info.relocatable) return; /* Glue the assignments into the abs section */ save = stat_ptr; stat_ptr = &(abs_output_section->children); for (j = 0; init[j].ptr; j++) { long val = init[j].value; lang_add_assignment (exp_assop ('=', init[j].symbol, exp_intop (val))); if (init[j].size == sizeof(short)) *(short *)init[j].ptr = val; else if (init[j].size == sizeof(int)) *(int *)init[j].ptr = val; else if (init[j].size == sizeof(long)) *(long *)init[j].ptr = val; /* This might be a long long or other special type. */ else if (init[j].size == sizeof(bfd_vma)) *(bfd_vma *)init[j].ptr = val; else abort(); } /* Restore the pointer. */ stat_ptr = save; if (pe.FileAlignment > pe.SectionAlignment) { einfo ("%P: warning, file alignment > section alignment.\n"); } } static void gld_${EMULATION_NAME}_after_open (void) { /* Pass the wacky PE command line options into the output bfd. FIXME: This should be done via a function, rather than by including an internal BFD header. */ if (!coff_data(output_bfd)->pe) { einfo ("%F%P: PE operations on non PE file.\n"); } pe_data(output_bfd)->pe_opthdr = pe; pe_data(output_bfd)->dll = init[DLLOFF].value; } /* Callback functions for qsort in sort_sections. */ static int sort_by_file_name (const void *a, const void *b) { const lang_statement_union_type *const *ra = a; const lang_statement_union_type *const *rb = b; int i, a_sec, b_sec; i = strcmp ((*ra)->input_section.ifile->the_bfd->my_archive->filename, (*rb)->input_section.ifile->the_bfd->my_archive->filename); if (i != 0) return i; i = strcmp ((*ra)->input_section.ifile->filename, (*rb)->input_section.ifile->filename); if (i != 0) return i; /* the tail idata4/5 are the only ones without relocs to an idata$6 section unless we are importing by ordinal, so sort them to last to terminate the IAT and HNT properly. if no reloc this one is import by ordinal so we have to sort by section contents */ if ( ((*ra)->input_section.section->reloc_count + (*rb)->input_section.section->reloc_count) ) { i = (((*ra)->input_section.section->reloc_count > (*rb)->input_section.section->reloc_count) ? -1 : 0); if ( i != 0) return i; return (((*ra)->input_section.section->reloc_count > (*rb)->input_section.section->reloc_count) ? 0 : 1); } else { if ( (strcmp( (*ra)->input_section.section->name, ".idata$6") == 0) ) return 0; /* don't sort .idata$6 or .idata$7 FIXME dlltool eliminate .idata$7 */ if (! bfd_get_section_contents ((*ra)->input_section.ifile->the_bfd, (*ra)->input_section.section, &a_sec, (file_ptr) 0, (bfd_size_type)sizeof(a_sec))) einfo ("%F%B: Can't read contents of section .idata: %E\n", (*ra)->input_section.ifile->the_bfd); if (! bfd_get_section_contents ((*rb)->input_section.ifile->the_bfd, (*rb)->input_section.section, &b_sec, (file_ptr) 0, (bfd_size_type)sizeof(b_sec) )) einfo ("%F%B: Can't read contents of section .idata: %E\n", (*rb)->input_section.ifile->the_bfd); i = ((a_sec < b_sec) ? -1 : 0); if ( i != 0) return i; return ((a_sec < b_sec) ? 0 : 1); } return 0; } static int sort_by_section_name (const void *a, const void *b) { const lang_statement_union_type *const *ra = a; const lang_statement_union_type *const *rb = b; int i; i = strcmp ((*ra)->input_section.section->name, (*rb)->input_section.section->name); /* this is a hack to make .stab and .stabstr last, so we don't have to fix strip/objcopy for .reloc sections. FIXME stripping images with a .rsrc section still needs to be fixed */ if ( i != 0) { if ((strncmp ((*ra)->input_section.section->name, ".stab", 5) == 0) && (strncmp ((*rb)->input_section.section->name, ".stab", 5) != 0)) return 1; return i; } return i; } /* Subroutine of sort_sections to a contiguous subset of a list of sections. NEXT_AFTER is the element after the last one to sort. The result is a pointer to the last element's "next" pointer. */ static lang_statement_union_type ** sort_sections_1 (lang_statement_union_type **startptr, lang_statement_union_type *next_after, int count, int (*sort_func) (const void *, const void *)) { lang_statement_union_type **vec; lang_statement_union_type *p; int i; lang_statement_union_type **ret; if (count == 0) return startptr; vec = ((lang_statement_union_type **) xmalloc (count * sizeof (lang_statement_union_type *))); for (p = *startptr, i = 0; i < count; i++, p = p->header.next) vec[i] = p; qsort (vec, count, sizeof (vec[0]), sort_func); /* Fill in the next pointers again. */ *startptr = vec[0]; for (i = 0; i < count - 1; i++) vec[i]->header.next = vec[i + 1]; vec[i]->header.next = next_after; ret = &vec[i]->header.next; free (vec); return ret; } /* Sort the .idata\$foo input sections of archives into filename order. The reason is so dlltool can arrange to have the pe dll import information generated correctly - the head of the list goes into dh.o, the tail into dt.o, and the guts into ds[nnnn].o. Note that this is only needed for the .idata section. FIXME: This may no longer be necessary with grouped sections. Instead of sorting on dh.o, ds[nnnn].o, dt.o, one could, for example, have dh.o use .idata\$4h, have ds[nnnn].o use .idata\$4s[nnnn], and have dt.o use .idata\$4t. This would have to be elaborated upon to handle multiple dll's [assuming such an eloboration is possible of course]. We also sort sections in '\$' wild statements. These are created by the place_orphans routine to implement grouped sections. */ static void sort_sections (lang_statement_union_type *s) { for (; s ; s = s->header.next) switch (s->header.type) { case lang_output_section_statement_enum: sort_sections (s->output_section_statement.children.head); break; case lang_wild_statement_enum: { lang_statement_union_type **p = &s->wild_statement.children.head; struct wildcard_list *sec; for (sec = s->wild_statement.section_list; sec; sec = sec->next) { /* Is this the .idata section? */ if (sec->spec.name != NULL && strncmp (sec->spec.name, ".idata", 6) == 0) { /* Sort the children. We want to sort any objects in the same archive. In order to handle the case of including a single archive multiple times, we sort all the children by archive name and then by object name. After sorting them, we re-thread the pointer chain. */ while (*p) { lang_statement_union_type *start = *p; if (start->header.type != lang_input_section_enum || !start->input_section.ifile->the_bfd->my_archive) p = &(start->header.next); else { lang_statement_union_type *end; int count; for (end = start, count = 0; end && (end->header.type == lang_input_section_enum); end = end->header.next) count++; p = sort_sections_1 (p, end, count, sort_by_file_name); } } break; } /* If this is a collection of grouped sections, sort them. The linker script must explicitly mention "*(.foo\$)" or "*(.foo\$*)". Don't sort them if \$ is not the last character (not sure if this is really useful, but it allows explicitly mentioning some \$ sections and letting the linker handle the rest). */ if (sec->spec.name != NULL) { char *q = strchr (sec->spec.name, '\$'); if (q != NULL && (q[1] == '\0' || (q[1] == '*' && q[2] == '\0'))) { lang_statement_union_type *end; int count; for (end = *p, count = 0; end; end = end->header.next) { if (end->header.type != lang_input_section_enum) abort (); count++; } (void) sort_sections_1 (p, end, count, sort_by_section_name); } break; } } } break; default: break; } } static void gld_${EMULATION_NAME}_before_allocation (void) { extern lang_statement_list_type *stat_ptr; #ifdef TARGET_IS_ppcpe /* Here we rummage through the found bfds to collect toc information */ { LANG_FOR_EACH_INPUT_STATEMENT (is) { if (!ppc_process_before_allocation(is->the_bfd, &link_info)) { einfo("Errors encountered processing file %s\n", is->filename); } } } /* We have seen it all. Allocate it, and carry on */ ppc_allocate_toc_section (&link_info); #else #ifdef TARGET_IS_armpe /* FIXME: we should be able to set the size of the interworking stub section. Here we rummage through the found bfds to collect glue information. FIXME: should this be based on a command line option? krk@cygnus.com */ { LANG_FOR_EACH_INPUT_STATEMENT (is) { if (!arm_process_before_allocation (is->the_bfd, & link_info)) { einfo ("Errors encountered processing file %s", is->filename); } } } /* We have seen it all. Allocate it, and carry on */ arm_allocate_interworking_sections (& link_info); #endif /* TARGET_IS_armpe */ #endif /* TARGET_IS_ppcpe */ sort_sections (stat_ptr->head); } /* Place an orphan section. We use this to put sections with a '\$' in them into the right place. Any section with a '\$' in them (e.g. .text\$foo) gets mapped to the output section with everything from the '\$' on stripped (e.g. .text). See the Microsoft Portable Executable and Common Object File Format Specification 4.1, section 4.2, Grouped Sections. FIXME: This is now handled by the linker script using wildcards, but I'm leaving this here in case we want to enable it for sections which are not mentioned in the linker script. */ static bfd_boolean gld${EMULATION_NAME}_place_orphan (lang_input_statement_type *file, asection *s) { const char *secname; char *output_secname, *ps; lang_output_section_statement_type *os; lang_statement_union_type *l; if ((s->flags & SEC_ALLOC) == 0) return FALSE; /* Don't process grouped sections unless doing a final link. If they're marked as COMDAT sections, we don't want .text\$foo to end up in .text and then have .text disappear because it's marked link-once-discard. */ if (link_info.relocatable) return FALSE; secname = bfd_get_section_name (s->owner, s); /* Everything from the '\$' on gets deleted so don't allow '\$' as the first character. */ if (*secname == '\$') einfo ("%P%F: section %s has '\$' as first character\n", secname); if (strchr (secname + 1, '\$') == NULL) return FALSE; /* Look up the output section. The Microsoft specs say sections names in image files never contain a '\$'. Fortunately, lang_..._lookup creates the section if it doesn't exist. */ output_secname = xstrdup (secname); ps = strchr (output_secname + 1, '\$'); *ps = 0; os = lang_output_section_statement_lookup (output_secname); /* Find the '\$' wild statement for this section. We currently require the linker script to explicitly mention "*(.foo\$)". FIXME: ppcpe.sc has .CRT\$foo in the .rdata section. According to the Microsoft docs this isn't correct so it's not (currently) handled. */ ps[0] = '\$'; ps[1] = 0; for (l = os->children.head; l; l = l->header.next) if (l->header.type == lang_wild_statement_enum) { struct wildcard_list *sec; for (sec = l->wild_statement.section_list; sec; sec = sec->next) if (sec->spec.name && strcmp (sec->spec.name, output_secname) == 0) break; if (sec) break; } ps[0] = 0; if (l == NULL) #if 1 einfo ("%P%F: *(%s\$) missing from linker script\n", output_secname); #else /* FIXME: This block is untried. It exists to convey the intent, should one decide to not require *(.foo\$) to appear in the linker script. */ { lang_wild_statement_type *new; struct wildcard_list *tmp; tmp = (struct wildcard_list *) xmalloc (sizeof *tmp); tmp->next = NULL; tmp->spec.name = xmalloc (strlen (output_secname) + 2); sprintf (tmp->spec.name, "%s\$", output_secname); tmp->spec.exclude_name_list = NULL; tmp->sorted = FALSE; new = new_stat (lang_wild_statement, &os->children); new->filename = NULL; new->filenames_sorted = FALSE; new->section_list = tmp; new->keep_sections = FALSE; lang_list_init (&new->children); l = new; } #endif /* Link the input section in and we're done for now. The sections still have to be sorted, but that has to wait until all such sections have been processed by us. The sorting is done by sort_sections. */ lang_add_section (&l->wild_statement.children, s, os, file); return TRUE; } static char * gld_${EMULATION_NAME}_get_script (int *isfile) EOF # Scripts compiled in. # sed commands to quote an ld script as a C string. sc="-f stringify.sed" cat >>e${EMULATION_NAME}.c <> e${EMULATION_NAME}.c echo ' ; else if (link_info.relocatable) return' >> e${EMULATION_NAME}.c sed $sc ldscripts/${EMULATION_NAME}.xr >> e${EMULATION_NAME}.c echo ' ; else if (!config.text_read_only) return' >> e${EMULATION_NAME}.c sed $sc ldscripts/${EMULATION_NAME}.xbn >> e${EMULATION_NAME}.c echo ' ; else if (!config.magic_demand_paged) return' >> e${EMULATION_NAME}.c sed $sc ldscripts/${EMULATION_NAME}.xn >> e${EMULATION_NAME}.c echo ' ; else return' >> e${EMULATION_NAME}.c sed $sc ldscripts/${EMULATION_NAME}.x >> e${EMULATION_NAME}.c echo '; }' >> e${EMULATION_NAME}.c cat >>e${EMULATION_NAME}.c <