diff options
author | Alan Modra <amodra@gmail.com> | 2022-07-20 18:28:50 +0930 |
---|---|---|
committer | Alan Modra <amodra@gmail.com> | 2022-07-21 13:35:51 +0930 |
commit | 175b91507b83ad42607d2f6dadaf55b7b511bdbe (patch) | |
tree | 48fe5761c403740b0fa7d7137a6a454295bee5d6 | |
parent | 590207aa6ece509eabbc7a51aab87e5d9bcfc869 (diff) | |
download | gdb-175b91507b83ad42607d2f6dadaf55b7b511bdbe.zip gdb-175b91507b83ad42607d2f6dadaf55b7b511bdbe.tar.gz gdb-175b91507b83ad42607d2f6dadaf55b7b511bdbe.tar.bz2 |
miscellaneous dwarf.c tidies
* dwarf.c: Leading and trailing whitespace fixes.
(free_abbrev_list): New function.
(free_all_abbrevs): Use the above. Free cu_abbrev_map here too.
(process_abbrev_set): Print actual section name on error.
(get_type_abbrev_from_form): Add overflow check.
(free_debug_memory): Don't free cu_abbrev_map here..
(process_debug_info): ..or here. Warn on another case of not
finding a neeeded abbrev.
-rw-r--r-- | binutils/dwarf.c | 216 |
1 files changed, 106 insertions, 110 deletions
diff --git a/binutils/dwarf.c b/binutils/dwarf.c index 2b1eec4..267ed3b 100644 --- a/binutils/dwarf.c +++ b/binutils/dwarf.c @@ -806,7 +806,7 @@ fetch_indexed_value (dwarf_vma idx, pointer_size = 4; bias = 12; } - + dwarf_vma offset = idx * pointer_size; /* Offsets are biased by the size of the section header @@ -901,38 +901,41 @@ record_abbrev_list_for_cu (dwarf_vma start, dwarf_vma end, abbrev_list * list) next_free_abbrev_map_entry ++; } -static void -free_all_abbrevs (void) +static abbrev_list * +free_abbrev_list (abbrev_list *list) { - abbrev_list * list; + abbrev_entry *abbrv = list->first_abbrev; - for (list = abbrev_lists; list != NULL;) + while (abbrv) { - abbrev_list * next = list->next; - abbrev_entry * abbrv; + abbrev_attr *attr = abbrv->first_attr; - for (abbrv = list->first_abbrev; abbrv != NULL;) + while (attr) { - abbrev_entry * next_abbrev = abbrv->next; - abbrev_attr * attr; - - for (attr = abbrv->first_attr; attr;) - { - abbrev_attr *next_attr = attr->next; - - free (attr); - attr = next_attr; - } - - free (abbrv); - abbrv = next_abbrev; + abbrev_attr *next_attr = attr->next; + free (attr); + attr = next_attr; } - free (list); - list = next; + abbrev_entry *next_abbrev = abbrv->next; + free (abbrv); + abbrv = next_abbrev; } - abbrev_lists = NULL; + abbrev_list *next = list->next; + free (list); + return next; +} + +static void +free_all_abbrevs (void) +{ + while (abbrev_lists) + abbrev_lists = free_abbrev_list (abbrev_lists); + + free (cu_abbrev_map); + cu_abbrev_map = NULL; + next_free_abbrev_map_entry = 0; } static abbrev_list * @@ -978,7 +981,7 @@ find_abbrev_map_by_offset (dwarf_vma offset) && cu_abbrev_map[i].end > offset) return cu_abbrev_map + i; - return NULL; + return NULL; } static void @@ -1119,7 +1122,7 @@ process_abbrev_set (struct dwarf_section *section, } /* Report the missing single zero which ends the section. */ - error (_(".debug_abbrev section not zero terminated\n")); + error (_("%s section not zero terminated\n"), section->name); return NULL; } @@ -1842,7 +1845,7 @@ fetch_alt_indirect_string (dwarf_vma offset) dwarf_vmatoa ("x", offset)); return _("<offset is too big>"); } - + static const char * get_AT_name (unsigned long attribute) { @@ -2124,7 +2127,8 @@ get_type_abbrev_from_form (unsigned long form, case DW_FORM_ref4: case DW_FORM_ref8: case DW_FORM_ref_udata: - if (uvalue + cu_offset > (size_t) (cu_end - section->start)) + if (uvalue + cu_offset < uvalue + || uvalue + cu_offset > (size_t) (cu_end - section->start)) { warn (_("Unable to resolve ref form: uvalue %lx + cu_offset %lx > CU size %lx\n"), uvalue, (long) cu_offset, (long) (cu_end - section->start)); @@ -2161,7 +2165,7 @@ get_type_abbrev_from_form (unsigned long form, else *map_return = NULL; } - + READ_ULEB (abbrev_number, data, section->start + section->size); for (entry = map->list->first_abbrev; entry != NULL; entry = entry->next) @@ -2750,10 +2754,10 @@ read_and_display_attr_value (unsigned long attribute, if (form == DW_FORM_loclistx) { if (dwo) - { - index = fetch_indexed_value (uvalue, loclists_dwo, 0); - index += (offset_size == 8) ? 20 : 12; - } + { + index = fetch_indexed_value (uvalue, loclists_dwo, 0); + index += (offset_size == 8) ? 20 : 12; + } else if (debug_info_p == NULL) { index = fetch_indexed_value (uvalue, loclists, 0); @@ -2771,21 +2775,21 @@ read_and_display_attr_value (unsigned long attribute, else if (form == DW_FORM_rnglistx) { if (dwo) - { - index = fetch_indexed_value (uvalue, rnglists_dwo, 0); - index += (offset_size == 8) ? 20 : 12; - } + { + index = fetch_indexed_value (uvalue, rnglists_dwo, 0); + index += (offset_size == 8) ? 20 : 12; + } else - { - if (debug_info_p == NULL) - base = 0; - else - base = debug_info_p->rnglists_base; - /* We do not have a cached value this time, so we perform the - computation manually. */ - index = fetch_indexed_value (uvalue, rnglists, base); - index += base; - } + { + if (debug_info_p == NULL) + base = 0; + else + base = debug_info_p->rnglists_base; + /* We do not have a cached value this time, so we perform the + computation manually. */ + index = fetch_indexed_value (uvalue, rnglists, base); + index += base; + } } else { @@ -2811,7 +2815,7 @@ read_and_display_attr_value (unsigned long attribute, if (!do_loc) printf ("%c<0x%s>", delimiter, dwarf_vmatoa ("x", uvalue + cu_offset)); break; - + default: warn (_("Unrecognized form: 0x%lx\n"), form); /* What to do? Consume a byte maybe? */ @@ -2836,9 +2840,9 @@ read_and_display_attr_value (unsigned long attribute, case DW_AT_rnglists_base: if (debug_info_p->rnglists_base) warn (_("CU @ 0x%s has multiple rnglists_base values (0x%s and 0x%s)"), - dwarf_vmatoa ("x", debug_info_p->cu_offset), - dwarf_vmatoa ("x", debug_info_p->rnglists_base), - dwarf_vmatoa ("x", uvalue)); + dwarf_vmatoa ("x", debug_info_p->cu_offset), + dwarf_vmatoa ("x", debug_info_p->rnglists_base), + dwarf_vmatoa ("x", uvalue)); debug_info_p->rnglists_base = uvalue; break; case DW_AT_str_offsets_base: @@ -2988,7 +2992,7 @@ read_and_display_attr_value (unsigned long attribute, case DW_FORM_strx3: case DW_FORM_strx4: add_dwo_name (fetch_indexed_string (uvalue, this_set, offset_size, false, - debug_info_p->str_offsets_base), + debug_info_p->str_offsets_base), cu_offset); break; case DW_FORM_string: @@ -3022,7 +3026,7 @@ read_and_display_attr_value (unsigned long attribute, case DW_FORM_strx3: case DW_FORM_strx4: add_dwo_dir (fetch_indexed_string (uvalue, this_set, offset_size, false, - debug_info_p->str_offsets_base), + debug_info_p->str_offsets_base), cu_offset); break; case DW_FORM_string: @@ -3653,11 +3657,8 @@ process_debug_info (struct dwarf_section * section, introduce (section, false); free_all_abbrevs (); - free (cu_abbrev_map); - cu_abbrev_map = NULL; - next_free_abbrev_map_entry = 0; - /* In order to be able to resolve DW_FORM_ref_attr forms we need + /* In order to be able to resolve DW_FORM_ref_addr forms we need to load *all* of the abbrevs for all CUs in this .debug_info section. This does effectively mean that we (partially) read every CU header twice. */ @@ -4028,12 +4029,11 @@ process_debug_info (struct dwarf_section * section, /* Scan through the abbreviation list until we reach the correct entry. */ - if (list == NULL) - continue; - - for (entry = list->first_abbrev; entry != NULL; entry = entry->next) - if (entry->number == abbrev_number) - break; + entry = NULL; + if (list != NULL) + for (entry = list->first_abbrev; entry != NULL; entry = entry->next) + if (entry->number == abbrev_number) + break; if (entry == NULL) { @@ -4057,7 +4057,7 @@ process_debug_info (struct dwarf_section * section, break; case DW_TAG_compile_unit: case DW_TAG_skeleton_unit: - need_base_address = 1; + need_base_address = 1; need_dwo_info = do_loc; break; case DW_TAG_entry_point: @@ -4440,7 +4440,7 @@ display_debug_sup (struct dwarf_section * section, SAFE_BYTE_GET_AND_INC (is_supplementary, start, 1, end); if (is_supplementary != 0 && is_supplementary != 1) - warn (_("corrupt .debug_sup section: is_supplementary not 0 or 1\n")); + warn (_("corrupt .debug_sup section: is_supplementary not 0 or 1\n")); sup_filename = start; if (is_supplementary && sup_filename[0] != 0) @@ -5619,7 +5619,7 @@ display_debug_lines_decoded (struct dwarf_section * section, printf ("%s %11d %#18" DWARF_VMA_FMT "x", newFileName, state_machine_regs.line, state_machine_regs.address); - } + } else { if (xop == -DW_LNE_end_sequence) @@ -6073,7 +6073,7 @@ display_debug_macro (struct dwarf_section *section, load_debug_section_with_follow (str, file); load_debug_section_with_follow (line, file); load_debug_section_with_follow (str_index, file); - + introduce (section, false); while (curr < end) @@ -6525,7 +6525,7 @@ display_loc_list (struct dwarf_section *section, /* Check base address specifiers. */ if (is_max_address (begin, pointer_size) - && !is_max_address (end, pointer_size)) + && !is_max_address (end, pointer_size)) { base_address = end; print_dwarf_vma (begin, pointer_size); @@ -6703,7 +6703,7 @@ display_loclists_list (struct dwarf_section * section, case DW_LLE_default_location: begin = end = 0; break; - + case DW_LLE_offset_pair: READ_ULEB (begin, start, section_end); begin += base_address; @@ -6999,7 +6999,7 @@ display_offset_entry_loclists (struct dwarf_section *section) unsigned char * start = section->start; unsigned char * const end = start + section->size; - introduce (section, false); + introduce (section, false); do { @@ -7048,14 +7048,14 @@ display_offset_entry_loclists (struct dwarf_section *section) section->name, segment_selector_size); return 0; } - + if (offset_entry_count == 0) { warn (_("The %s section contains a table without offset\n"), section->name); return 0; } - + printf (_("\n Offset Entries starting at 0x%lx:\n"), (long)(start - section->start)); @@ -8217,7 +8217,7 @@ display_debug_rnglists (struct dwarf_section *section) start = display_debug_rnglists_list (start, end, address_size, offset, 0, offset_size); if (start >= end) - break; + break; } start = end; @@ -8335,12 +8335,12 @@ display_debug_ranges (struct dwarf_section *section, next = section_begin + offset + debug_info_p->rnglists_base; /* If multiple DWARF entities reference the same range then we will - have multiple entries in the `range_entries' list for the same - offset. Thanks to the sort above these will all be consecutive in - the `range_entries' list, so we can easily ignore duplicates - here. */ + have multiple entries in the `range_entries' list for the same + offset. Thanks to the sort above these will all be consecutive in + the `range_entries' list, so we can easily ignore duplicates + here. */ if (i > 0 && last_offset == offset) - continue; + continue; last_offset = offset; if (dwarf_check != 0 && i > 0) @@ -10265,7 +10265,7 @@ display_debug_names (struct dwarf_section *section, void *file) printf (_("Out of %lu items there are %zu bucket clashes" " (longest of %zu entries).\n"), (unsigned long) name_count, hash_clash_count, longest_clash); - + if (name_count != buckets_filled + hash_clash_count) warn (_("The name_count (%lu) is not the same as the used bucket_count (%lu) + the hash clash count (%lu)"), (unsigned long) name_count, @@ -10369,7 +10369,7 @@ display_debug_names (struct dwarf_section *section, void *file) break; if (tagno >= 0) printf ("%s<%lu>", - (tagno == 0 && second_abbrev_tag == 0 ? " " : "\n\t"), + (tagno == 0 && second_abbrev_tag == 0 ? " " : "\n\t"), (unsigned long) abbrev_tag); for (entry = abbrev_lookup; @@ -10898,7 +10898,7 @@ process_cu_tu_index (struct dwarf_section *section, int do_display) Check for integer overflow (can occur when size_t is 32-bit) with overlarge ncols or nused values. */ if (nused == -1u - || _mul_overflow ((size_t) ncols, 4, &temp) + || _mul_overflow ((size_t) ncols, 4, &temp) || _mul_overflow ((size_t) nused + 1, temp, &total) || total > (size_t) (limit - ppool)) { @@ -10906,7 +10906,7 @@ process_cu_tu_index (struct dwarf_section *section, int do_display) section->name); return 0; } - + if (do_display) { printf (_(" Offset table\n")); @@ -11410,8 +11410,8 @@ add_separate_debug_file (const char * filename, void * handle) static bool debuginfod_fetch_separate_debug_info (struct dwarf_section * section, - char ** filename, - void * file) + char ** filename, + void * file) { size_t build_id_len; unsigned char * build_id; @@ -11429,14 +11429,14 @@ debuginfod_fetch_separate_debug_info (struct dwarf_section * section, filelen = strnlen ((const char *)section->start, section->size); if (filelen == section->size) - /* Corrupt debugaltlink. */ - return false; + /* Corrupt debugaltlink. */ + return false; build_id = section->start + filelen + 1; build_id_len = section->size - (filelen + 1); if (build_id_len == 0) - return false; + return false; } else return false; @@ -11448,25 +11448,25 @@ debuginfod_fetch_separate_debug_info (struct dwarf_section * section, client = debuginfod_begin (); if (client == NULL) - return false; + return false; /* Query debuginfod servers for the target file. If found its path - will be stored in filename. */ + will be stored in filename. */ fd = debuginfod_find_debuginfo (client, build_id, build_id_len, filename); debuginfod_end (client); /* Only free build_id if we allocated space for a hex string - in get_build_id (). */ + in get_build_id (). */ if (build_id_len == 0) - free (build_id); + free (build_id); if (fd >= 0) - { - /* File successfully retrieved. Close fd since we want to - use open_debug_file () on filename instead. */ - close (fd); - return true; - } + { + /* File successfully retrieved. Close fd since we want to + use open_debug_file () on filename instead. */ + close (fd); + return true; + } } return false; @@ -11479,7 +11479,7 @@ load_separate_debug_info (const char * main_filename, parse_func_type parse_func, check_func_type check_func, void * func_data, - void * file ATTRIBUTE_UNUSED) + void * file ATTRIBUTE_UNUSED) { const char * separate_filename; char * debug_filename; @@ -11595,11 +11595,11 @@ load_separate_debug_info (const char * main_filename, & tmp_filename, file)) { - /* File successfully downloaded from server, replace - debug_filename with the file's path. */ - free (debug_filename); - debug_filename = tmp_filename; - goto found; + /* File successfully downloaded from server, replace + debug_filename with the file's path. */ + free (debug_filename); + debug_filename = tmp_filename; + goto found; } } #endif @@ -11766,12 +11766,12 @@ load_build_id_debug_file (const char * main_filename ATTRIBUTE_UNUSED, void * ma /* In theory we should extract the contents of the section into a note structure and then check the fields. For now though just use hard coded offsets instead: - + Field Bytes Contents NSize 0...3 4 DSize 4...7 8+ Type 8..11 3 (NT_GNU_BUILD_ID) - Name 12.15 GNU\0 + Name 12.15 GNU\0 Data 16.... */ /* FIXME: Check the name size, name and type fields. */ @@ -11783,7 +11783,7 @@ load_build_id_debug_file (const char * main_filename ATTRIBUTE_UNUSED, void * ma warn (_(".note.gnu.build-id data size is too small\n")); return; } - + if (build_id_size > (section->size - 16)) { warn (_(".note.gnu.build-id data size is too bug\n")); @@ -12079,10 +12079,6 @@ free_debug_memory (void) free_all_abbrevs (); - free (cu_abbrev_map); - cu_abbrev_map = NULL; - next_free_abbrev_map_entry = 0; - free (shndx_pool); shndx_pool = NULL; shndx_pool_size = 0; |