1 /* Generate assembler source containing symbol information
3 * Copyright 2002 by Kai Germaschewski
5 * This software may be used and distributed according to the terms
6 * of the GNU General Public License, incorporated herein by reference.
8 * Usage: nm -n vmlinux | scripts/kallsyms [--all-symbols] > symbols.S
10 * Table compression uses all the unused char codes on the symbols and
11 * maps these to the most used substrings (tokens). For instance, it might
12 * map char code 0xF7 to represent "write_" and then in every symbol where
13 * "write_" appears it can be replaced by 0xF7, saving 5 bytes.
14 * The used codes themselves are also placed in the table so that the
15 * decompresion can work without "special cases".
16 * Applied to kernel symbols, this usually produces a compression ratio
28 #define ARRAY_SIZE(arr) (sizeof(arr) / sizeof(arr[0]))
30 #define _stringify_1(x) #x
31 #define _stringify(x) _stringify_1(x)
33 #define KSYM_NAME_LEN 512
36 * A substantially bigger size than the current maximum.
38 * It cannot be defined as an expression because it gets stringified
39 * for the fscanf() format string. Therefore, a _Static_assert() is
40 * used instead to maintain the relationship with KSYM_NAME_LEN.
42 #define KSYM_NAME_LEN_BUFFER 2048
44 KSYM_NAME_LEN_BUFFER == KSYM_NAME_LEN * 4,
45 "Please keep KSYM_NAME_LEN_BUFFER in sync with KSYM_NAME_LEN"
49 unsigned long long addr;
51 unsigned int start_pos;
52 unsigned int percpu_absolute;
57 const char *start_sym, *end_sym;
58 unsigned long long start, end;
61 static unsigned long long _text;
62 static unsigned long long relative_base;
63 static struct addr_range text_ranges[] = {
64 { "_stext", "_etext" },
65 { "_sinittext", "_einittext" },
67 #define text_range_text (&text_ranges[0])
68 #define text_range_inittext (&text_ranges[1])
70 static struct addr_range percpu_range = {
71 "__per_cpu_start", "__per_cpu_end", -1ULL, 0
74 static struct sym_entry **table;
75 static unsigned int table_size, table_cnt;
76 static int all_symbols;
77 static int absolute_percpu;
78 static int base_relative;
80 static int token_profit[0x10000];
82 /* the table that holds the result of the compression */
83 static unsigned char best_table[256][2];
84 static unsigned char best_table_len[256];
87 static void usage(void)
89 fprintf(stderr, "Usage: kallsyms [--all-symbols] [--absolute-percpu] "
90 "[--base-relative] < in.map > out.S\n");
94 static char *sym_name(const struct sym_entry *s)
96 return (char *)s->sym + 1;
99 static bool is_ignored_symbol(const char *name, char type)
101 /* Symbol names that exactly match to the following are ignored.*/
102 static const char * const ignored_symbols[] = {
104 * Symbols which vary between passes. Passes 1 and 2 must have
105 * identical symbol lists. The kallsyms_* symbols below are
106 * only added after pass 1, they would be included in pass 2
107 * when --all-symbols is specified so exclude them to get a
108 * stable symbol list.
110 "kallsyms_addresses",
112 "kallsyms_relative_base",
116 "kallsyms_token_table",
117 "kallsyms_token_index",
118 /* Exclude linker generated symbols which vary between passes */
119 "_SDA_BASE_", /* ppc */
120 "_SDA2_BASE_", /* ppc */
124 /* Symbol names that begin with the following are ignored.*/
125 static const char * const ignored_prefixes[] = {
126 "$", /* local symbols for ARM, MIPS, etc. */
127 ".L", /* local labels, .LBB,.Ltmpxxx,.L__unnamed_xx,.LASANPC, etc. */
128 "__crc_", /* modversions */
129 "__efistub_", /* arm64 EFI stub namespace */
130 "__kvm_nvhe_$", /* arm64 local symbols in non-VHE KVM namespace */
131 "__kvm_nvhe_.L", /* arm64 local symbols in non-VHE KVM namespace */
132 "__AArch64ADRPThunk_", /* arm64 lld */
133 "__ARMV5PILongThunk_", /* arm lld */
134 "__ARMV7PILongThunk_",
135 "__ThumbV7PILongThunk_",
136 "__LA25Thunk_", /* mips lld */
138 "__kcfi_typeid_", /* CFI type identifiers */
142 /* Symbol names that end with the following are ignored.*/
143 static const char * const ignored_suffixes[] = {
144 "_from_arm", /* arm */
145 "_from_thumb", /* arm */
150 /* Symbol names that contain the following are ignored.*/
151 static const char * const ignored_matches[] = {
152 ".long_branch.", /* ppc stub */
153 ".plt_branch.", /* ppc stub */
157 const char * const *p;
159 for (p = ignored_symbols; *p; p++)
160 if (!strcmp(name, *p))
163 for (p = ignored_prefixes; *p; p++)
164 if (!strncmp(name, *p, strlen(*p)))
167 for (p = ignored_suffixes; *p; p++) {
168 int l = strlen(name) - strlen(*p);
170 if (l >= 0 && !strcmp(name + l, *p))
174 for (p = ignored_matches; *p; p++) {
175 if (strstr(name, *p))
179 if (type == 'U' || type == 'u')
181 /* exclude debugging symbols */
182 if (type == 'N' || type == 'n')
185 if (toupper(type) == 'A') {
186 /* Keep these useful absolute symbols */
187 if (strcmp(name, "__kernel_syscall_via_break") &&
188 strcmp(name, "__kernel_syscall_via_epc") &&
189 strcmp(name, "__kernel_sigtramp") &&
190 strcmp(name, "__gp"))
197 static void check_symbol_range(const char *sym, unsigned long long addr,
198 struct addr_range *ranges, int entries)
201 struct addr_range *ar;
203 for (i = 0; i < entries; ++i) {
206 if (strcmp(sym, ar->start_sym) == 0) {
209 } else if (strcmp(sym, ar->end_sym) == 0) {
216 static struct sym_entry *read_symbol(FILE *in)
218 char name[KSYM_NAME_LEN_BUFFER+1], type;
219 unsigned long long addr;
221 struct sym_entry *sym;
224 rc = fscanf(in, "%llx %c %" _stringify(KSYM_NAME_LEN_BUFFER) "s\n", &addr, &type, name);
226 if (rc != EOF && fgets(name, ARRAY_SIZE(name), in) == NULL)
227 fprintf(stderr, "Read error or end of file.\n");
230 if (strlen(name) >= KSYM_NAME_LEN) {
231 fprintf(stderr, "Symbol %s too long for kallsyms (%zu >= %d).\n"
232 "Please increase KSYM_NAME_LEN both in kernel and kallsyms.c\n",
233 name, strlen(name), KSYM_NAME_LEN);
237 if (strcmp(name, "_text") == 0)
240 /* Ignore most absolute/undefined (?) symbols. */
241 if (is_ignored_symbol(name, type))
244 check_symbol_range(name, addr, text_ranges, ARRAY_SIZE(text_ranges));
245 check_symbol_range(name, addr, &percpu_range, 1);
247 /* include the type field in the symbol name, so that it gets
248 * compressed together */
250 len = strlen(name) + 1;
252 sym = malloc(sizeof(*sym) + len + 1);
254 fprintf(stderr, "kallsyms failure: "
255 "unable to allocate required amount of memory\n");
261 strcpy(sym_name(sym), name);
262 sym->percpu_absolute = 0;
267 static int symbol_in_range(const struct sym_entry *s,
268 const struct addr_range *ranges, int entries)
271 const struct addr_range *ar;
273 for (i = 0; i < entries; ++i) {
276 if (s->addr >= ar->start && s->addr <= ar->end)
283 static int symbol_valid(const struct sym_entry *s)
285 const char *name = sym_name(s);
287 /* if --all-symbols is not specified, then symbols outside the text
288 * and inittext sections are discarded */
290 if (symbol_in_range(s, text_ranges,
291 ARRAY_SIZE(text_ranges)) == 0)
293 /* Corner case. Discard any symbols with the same value as
294 * _etext _einittext; they can move between pass 1 and 2 when
295 * the kallsyms data are added. If these symbols move then
296 * they may get dropped in pass 2, which breaks the kallsyms
299 if ((s->addr == text_range_text->end &&
300 strcmp(name, text_range_text->end_sym)) ||
301 (s->addr == text_range_inittext->end &&
302 strcmp(name, text_range_inittext->end_sym)))
309 /* remove all the invalid symbols from the table */
310 static void shrink_table(void)
315 for (i = 0; i < table_cnt; i++) {
316 if (symbol_valid(table[i])) {
318 table[pos] = table[i];
326 /* When valid symbol is not registered, exit to error */
328 fprintf(stderr, "No valid symbol.\n");
333 static void read_map(FILE *in)
335 struct sym_entry *sym;
338 sym = read_symbol(in);
342 sym->start_pos = table_cnt;
344 if (table_cnt >= table_size) {
346 table = realloc(table, sizeof(*table) * table_size);
348 fprintf(stderr, "out of memory\n");
353 table[table_cnt++] = sym;
357 static void output_label(const char *label)
359 printf(".globl %s\n", label);
361 printf("%s:\n", label);
364 /* Provide proper symbols relocatability by their '_text' relativeness. */
365 static void output_address(unsigned long long addr)
368 printf("\tPTR\t_text + %#llx\n", addr - _text);
370 printf("\tPTR\t_text - %#llx\n", _text - addr);
373 /* uncompress a compressed symbol. When this function is called, the best table
374 * might still be compressed itself, so the function needs to be recursive */
375 static int expand_symbol(const unsigned char *data, int len, char *result)
377 int c, rlen, total=0;
381 /* if the table holds a single char that is the same as the one
382 * we are looking for, then end the search */
383 if (best_table[c][0]==c && best_table_len[c]==1) {
387 /* if not, recurse and expand */
388 rlen = expand_symbol(best_table[c], best_table_len[c], result);
400 static int symbol_absolute(const struct sym_entry *s)
402 return s->percpu_absolute;
405 static void write_src(void)
407 unsigned int i, k, off;
408 unsigned int best_idx[256];
409 unsigned int *markers;
410 char buf[KSYM_NAME_LEN];
412 printf("#include <asm/bitsperlong.h>\n");
413 printf("#if BITS_PER_LONG == 64\n");
414 printf("#define PTR .quad\n");
415 printf("#define ALGN .balign 8\n");
417 printf("#define PTR .long\n");
418 printf("#define ALGN .balign 4\n");
421 printf("\t.section .rodata, \"a\"\n");
424 output_label("kallsyms_addresses");
426 output_label("kallsyms_offsets");
428 for (i = 0; i < table_cnt; i++) {
431 * Use the offset relative to the lowest value
432 * encountered of all relative symbols, and emit
433 * non-relocatable fixed offsets that will be fixed
440 if (!absolute_percpu) {
441 offset = table[i]->addr - relative_base;
442 overflow = (offset < 0 || offset > UINT_MAX);
443 } else if (symbol_absolute(table[i])) {
444 offset = table[i]->addr;
445 overflow = (offset < 0 || offset > INT_MAX);
447 offset = relative_base - table[i]->addr - 1;
448 overflow = (offset < INT_MIN || offset >= 0);
451 fprintf(stderr, "kallsyms failure: "
452 "%s symbol value %#llx out of range in relative mode\n",
453 symbol_absolute(table[i]) ? "absolute" : "relative",
457 printf("\t.long\t%#x\n", (int)offset);
458 } else if (!symbol_absolute(table[i])) {
459 output_address(table[i]->addr);
461 printf("\tPTR\t%#llx\n", table[i]->addr);
467 output_label("kallsyms_relative_base");
468 output_address(relative_base);
472 output_label("kallsyms_num_syms");
473 printf("\t.long\t%u\n", table_cnt);
476 /* table of offset markers, that give the offset in the compressed stream
477 * every 256 symbols */
478 markers = malloc(sizeof(unsigned int) * ((table_cnt + 255) / 256));
480 fprintf(stderr, "kallsyms failure: "
481 "unable to allocate required memory\n");
485 output_label("kallsyms_names");
487 for (i = 0; i < table_cnt; i++) {
489 markers[i >> 8] = off;
491 /* There cannot be any symbol of length zero. */
492 if (table[i]->len == 0) {
493 fprintf(stderr, "kallsyms failure: "
494 "unexpected zero symbol length\n");
498 /* Only lengths that fit in up-to-two-byte ULEB128 are supported. */
499 if (table[i]->len > 0x3FFF) {
500 fprintf(stderr, "kallsyms failure: "
501 "unexpected huge symbol length\n");
505 /* Encode length with ULEB128. */
506 if (table[i]->len <= 0x7F) {
507 /* Most symbols use a single byte for the length. */
508 printf("\t.byte 0x%02x", table[i]->len);
509 off += table[i]->len + 1;
511 /* "Big" symbols use two bytes. */
512 printf("\t.byte 0x%02x, 0x%02x",
513 (table[i]->len & 0x7F) | 0x80,
514 (table[i]->len >> 7) & 0x7F);
515 off += table[i]->len + 2;
517 for (k = 0; k < table[i]->len; k++)
518 printf(", 0x%02x", table[i]->sym[k]);
523 output_label("kallsyms_markers");
524 for (i = 0; i < ((table_cnt + 255) >> 8); i++)
525 printf("\t.long\t%u\n", markers[i]);
530 output_label("kallsyms_token_table");
532 for (i = 0; i < 256; i++) {
534 expand_symbol(best_table[i], best_table_len[i], buf);
535 printf("\t.asciz\t\"%s\"\n", buf);
536 off += strlen(buf) + 1;
540 output_label("kallsyms_token_index");
541 for (i = 0; i < 256; i++)
542 printf("\t.short\t%d\n", best_idx[i]);
547 /* table lookup compression functions */
549 /* count all the possible tokens in a symbol */
550 static void learn_symbol(const unsigned char *symbol, int len)
554 for (i = 0; i < len - 1; i++)
555 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]++;
558 /* decrease the count for all the possible tokens in a symbol */
559 static void forget_symbol(const unsigned char *symbol, int len)
563 for (i = 0; i < len - 1; i++)
564 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]--;
567 /* do the initial token count */
568 static void build_initial_tok_table(void)
572 for (i = 0; i < table_cnt; i++)
573 learn_symbol(table[i]->sym, table[i]->len);
576 static unsigned char *find_token(unsigned char *str, int len,
577 const unsigned char *token)
581 for (i = 0; i < len - 1; i++) {
582 if (str[i] == token[0] && str[i+1] == token[1])
588 /* replace a given token in all the valid symbols. Use the sampled symbols
589 * to update the counts */
590 static void compress_symbols(const unsigned char *str, int idx)
592 unsigned int i, len, size;
593 unsigned char *p1, *p2;
595 for (i = 0; i < table_cnt; i++) {
600 /* find the token on the symbol */
601 p2 = find_token(p1, len, str);
604 /* decrease the counts for this symbol's tokens */
605 forget_symbol(table[i]->sym, len);
613 memmove(p2, p2 + 1, size);
619 /* find the token on the symbol */
620 p2 = find_token(p1, size, str);
626 /* increase the counts for this symbol's new tokens */
627 learn_symbol(table[i]->sym, len);
631 /* search the token with the maximum profit */
632 static int find_best_token(void)
634 int i, best, bestprofit;
639 for (i = 0; i < 0x10000; i++) {
640 if (token_profit[i] > bestprofit) {
642 bestprofit = token_profit[i];
648 /* this is the core of the algorithm: calculate the "best" table */
649 static void optimize_result(void)
653 /* using the '\0' symbol last allows compress_symbols to use standard
654 * fast string functions */
655 for (i = 255; i >= 0; i--) {
657 /* if this table slot is empty (it is not used by an actual
658 * original char code */
659 if (!best_table_len[i]) {
661 /* find the token with the best profit value */
662 best = find_best_token();
663 if (token_profit[best] == 0)
666 /* place it in the "best" table */
667 best_table_len[i] = 2;
668 best_table[i][0] = best & 0xFF;
669 best_table[i][1] = (best >> 8) & 0xFF;
671 /* replace this token in all the valid symbols */
672 compress_symbols(best_table[i], i);
677 /* start by placing the symbols that are actually used on the table */
678 static void insert_real_symbols_in_table(void)
680 unsigned int i, j, c;
682 for (i = 0; i < table_cnt; i++) {
683 for (j = 0; j < table[i]->len; j++) {
684 c = table[i]->sym[j];
691 static void optimize_token_table(void)
693 build_initial_tok_table();
695 insert_real_symbols_in_table();
700 /* guess for "linker script provide" symbol */
701 static int may_be_linker_script_provide_symbol(const struct sym_entry *se)
703 const char *symbol = sym_name(se);
704 int len = se->len - 1;
709 if (symbol[0] != '_' || symbol[1] != '_')
713 if (!memcmp(symbol + 2, "start_", 6))
717 if (!memcmp(symbol + 2, "stop_", 5))
721 if (!memcmp(symbol + 2, "end_", 4))
725 if (!memcmp(symbol + len - 6, "_start", 6))
729 if (!memcmp(symbol + len - 4, "_end", 4))
735 static int compare_symbols(const void *a, const void *b)
737 const struct sym_entry *sa = *(const struct sym_entry **)a;
738 const struct sym_entry *sb = *(const struct sym_entry **)b;
741 /* sort by address first */
742 if (sa->addr > sb->addr)
744 if (sa->addr < sb->addr)
747 /* sort by "weakness" type */
748 wa = (sa->sym[0] == 'w') || (sa->sym[0] == 'W');
749 wb = (sb->sym[0] == 'w') || (sb->sym[0] == 'W');
753 /* sort by "linker script provide" type */
754 wa = may_be_linker_script_provide_symbol(sa);
755 wb = may_be_linker_script_provide_symbol(sb);
759 /* sort by the number of prefix underscores */
760 wa = strspn(sym_name(sa), "_");
761 wb = strspn(sym_name(sb), "_");
765 /* sort by initial order, so that other symbols are left undisturbed */
766 return sa->start_pos - sb->start_pos;
769 static void sort_symbols(void)
771 qsort(table, table_cnt, sizeof(table[0]), compare_symbols);
774 static void make_percpus_absolute(void)
778 for (i = 0; i < table_cnt; i++)
779 if (symbol_in_range(table[i], &percpu_range, 1)) {
781 * Keep the 'A' override for percpu symbols to
782 * ensure consistent behavior compared to older
783 * versions of this tool.
785 table[i]->sym[0] = 'A';
786 table[i]->percpu_absolute = 1;
790 /* find the minimum non-absolute symbol address */
791 static void record_relative_base(void)
795 for (i = 0; i < table_cnt; i++)
796 if (!symbol_absolute(table[i])) {
798 * The table is sorted by address.
799 * Take the first non-absolute symbol value.
801 relative_base = table[i]->addr;
806 int main(int argc, char **argv)
810 for (i = 1; i < argc; i++) {
811 if(strcmp(argv[i], "--all-symbols") == 0)
813 else if (strcmp(argv[i], "--absolute-percpu") == 0)
815 else if (strcmp(argv[i], "--base-relative") == 0)
820 } else if (argc != 1)
826 make_percpus_absolute();
829 record_relative_base();
830 optimize_token_table();