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
27 #define ARRAY_SIZE(arr) (sizeof(arr) / sizeof(arr[0]))
29 #define KSYM_NAME_LEN 128
32 unsigned long long addr;
34 unsigned int start_pos;
36 unsigned int percpu_absolute;
40 const char *start_sym, *end_sym;
41 unsigned long long start, end;
44 static unsigned long long _text;
45 static unsigned long long relative_base;
46 static struct addr_range text_ranges[] = {
47 { "_stext", "_etext" },
48 { "_sinittext", "_einittext" },
50 #define text_range_text (&text_ranges[0])
51 #define text_range_inittext (&text_ranges[1])
53 static struct addr_range percpu_range = {
54 "__per_cpu_start", "__per_cpu_end", -1ULL, 0
57 static struct sym_entry *table;
58 static unsigned int table_size, table_cnt;
59 static int all_symbols = 0;
60 static int absolute_percpu = 0;
61 static int base_relative = 0;
63 static int token_profit[0x10000];
65 /* the table that holds the result of the compression */
66 static unsigned char best_table[256][2];
67 static unsigned char best_table_len[256];
70 static void usage(void)
72 fprintf(stderr, "Usage: kallsyms [--all-symbols] "
73 "[--base-relative] < in.map > out.S\n");
77 static char *sym_name(const struct sym_entry *s)
79 return (char *)s->sym + 1;
82 static int check_symbol_range(const char *sym, unsigned long long addr,
83 struct addr_range *ranges, int entries)
86 struct addr_range *ar;
88 for (i = 0; i < entries; ++i) {
91 if (strcmp(sym, ar->start_sym) == 0) {
94 } else if (strcmp(sym, ar->end_sym) == 0) {
103 static int read_symbol(FILE *in, struct sym_entry *s)
105 char sym[500], stype;
108 rc = fscanf(in, "%llx %c %499s\n", &s->addr, &stype, sym);
110 if (rc != EOF && fgets(sym, 500, in) == NULL)
111 fprintf(stderr, "Read error or end of file.\n");
114 if (strlen(sym) >= KSYM_NAME_LEN) {
115 fprintf(stderr, "Symbol %s too long for kallsyms (%zu >= %d).\n"
116 "Please increase KSYM_NAME_LEN both in kernel and kallsyms.c\n",
117 sym, strlen(sym), KSYM_NAME_LEN);
121 /* Ignore most absolute/undefined (?) symbols. */
122 if (strcmp(sym, "_text") == 0)
124 else if (check_symbol_range(sym, s->addr, text_ranges,
125 ARRAY_SIZE(text_ranges)) == 0)
127 else if (toupper(stype) == 'A')
129 /* Keep these useful absolute symbols */
130 if (strcmp(sym, "__kernel_syscall_via_break") &&
131 strcmp(sym, "__kernel_syscall_via_epc") &&
132 strcmp(sym, "__kernel_sigtramp") &&
137 else if (toupper(stype) == 'U')
140 * Ignore generated symbols such as:
141 * - mapping symbols in ARM ELF files ($a, $t, and $d)
142 * - MIPS ELF local symbols ($L123 instead of .L123)
144 else if (sym[0] == '$')
146 /* exclude debugging symbols */
147 else if (stype == 'N' || stype == 'n')
149 /* exclude s390 kasan local symbols */
150 else if (!strncmp(sym, ".LASANPC", 8))
153 /* include the type field in the symbol name, so that it gets
154 * compressed together */
155 s->len = strlen(sym) + 1;
156 s->sym = malloc(s->len + 1);
158 fprintf(stderr, "kallsyms failure: "
159 "unable to allocate required amount of memory\n");
162 strcpy(sym_name(s), sym);
165 s->percpu_absolute = 0;
167 /* Record if we've found __per_cpu_start/end. */
168 check_symbol_range(sym, s->addr, &percpu_range, 1);
173 static int symbol_in_range(struct sym_entry *s, struct addr_range *ranges,
177 struct addr_range *ar;
179 for (i = 0; i < entries; ++i) {
182 if (s->addr >= ar->start && s->addr <= ar->end)
189 static int symbol_valid(struct sym_entry *s)
191 /* Symbols which vary between passes. Passes 1 and 2 must have
192 * identical symbol lists. The kallsyms_* symbols below are only added
193 * after pass 1, they would be included in pass 2 when --all-symbols is
194 * specified so exclude them to get a stable symbol list.
196 static char *special_symbols[] = {
197 "kallsyms_addresses",
199 "kallsyms_relative_base",
203 "kallsyms_token_table",
204 "kallsyms_token_index",
206 /* Exclude linker generated symbols which vary between passes */
207 "_SDA_BASE_", /* ppc */
208 "_SDA2_BASE_", /* ppc */
211 static char *special_prefixes[] = {
212 "__crc_", /* modversions */
213 "__efistub_", /* arm64 EFI stub namespace */
216 static char *special_suffixes[] = {
218 "_from_arm", /* arm */
219 "_from_thumb", /* arm */
223 const char *name = sym_name(s);
225 /* if --all-symbols is not specified, then symbols outside the text
226 * and inittext sections are discarded */
228 if (symbol_in_range(s, text_ranges,
229 ARRAY_SIZE(text_ranges)) == 0)
231 /* Corner case. Discard any symbols with the same value as
232 * _etext _einittext; they can move between pass 1 and 2 when
233 * the kallsyms data are added. If these symbols move then
234 * they may get dropped in pass 2, which breaks the kallsyms
237 if ((s->addr == text_range_text->end &&
238 strcmp(name, text_range_text->end_sym)) ||
239 (s->addr == text_range_inittext->end &&
240 strcmp(name, text_range_inittext->end_sym)))
244 /* Exclude symbols which vary between passes. */
245 for (i = 0; special_symbols[i]; i++)
246 if (strcmp(name, special_symbols[i]) == 0)
249 for (i = 0; special_prefixes[i]; i++) {
250 int l = strlen(special_prefixes[i]);
252 if (strncmp(name, special_prefixes[i], l) == 0)
256 for (i = 0; special_suffixes[i]; i++) {
257 int l = strlen(name) - strlen(special_suffixes[i]);
259 if (l >= 0 && strcmp(name + l, special_suffixes[i]) == 0)
266 /* remove all the invalid symbols from the table */
267 static void shrink_table(void)
272 for (i = 0; i < table_cnt; i++) {
273 if (symbol_valid(&table[i])) {
275 table[pos] = table[i];
283 /* When valid symbol is not registered, exit to error */
285 fprintf(stderr, "No valid symbol.\n");
290 static void read_map(FILE *in)
293 if (table_cnt >= table_size) {
295 table = realloc(table, sizeof(*table) * table_size);
297 fprintf(stderr, "out of memory\n");
301 if (read_symbol(in, &table[table_cnt]) == 0) {
302 table[table_cnt].start_pos = table_cnt;
308 static void output_label(char *label)
310 printf(".globl %s\n", label);
312 printf("%s:\n", label);
315 /* uncompress a compressed symbol. When this function is called, the best table
316 * might still be compressed itself, so the function needs to be recursive */
317 static int expand_symbol(unsigned char *data, int len, char *result)
319 int c, rlen, total=0;
323 /* if the table holds a single char that is the same as the one
324 * we are looking for, then end the search */
325 if (best_table[c][0]==c && best_table_len[c]==1) {
329 /* if not, recurse and expand */
330 rlen = expand_symbol(best_table[c], best_table_len[c], result);
342 static int symbol_absolute(struct sym_entry *s)
344 return s->percpu_absolute;
347 static void write_src(void)
349 unsigned int i, k, off;
350 unsigned int best_idx[256];
351 unsigned int *markers;
352 char buf[KSYM_NAME_LEN];
354 printf("#include <asm/bitsperlong.h>\n");
355 printf("#if BITS_PER_LONG == 64\n");
356 printf("#define PTR .quad\n");
357 printf("#define ALGN .balign 8\n");
359 printf("#define PTR .long\n");
360 printf("#define ALGN .balign 4\n");
363 printf("\t.section .rodata, \"a\"\n");
365 /* Provide proper symbols relocatability by their relativeness
366 * to a fixed anchor point in the runtime image, either '_text'
367 * for absolute address tables, in which case the linker will
368 * emit the final addresses at build time. Otherwise, use the
369 * offset relative to the lowest value encountered of all relative
370 * symbols, and emit non-relocatable fixed offsets that will be fixed
373 * The symbol names cannot be used to construct normal symbol
374 * references as the list of symbols contains symbols that are
375 * declared static and are private to their .o files. This prevents
376 * .tmp_kallsyms.o or any other object from referencing them.
379 output_label("kallsyms_addresses");
381 output_label("kallsyms_offsets");
383 for (i = 0; i < table_cnt; i++) {
388 if (!absolute_percpu) {
389 offset = table[i].addr - relative_base;
390 overflow = (offset < 0 || offset > UINT_MAX);
391 } else if (symbol_absolute(&table[i])) {
392 offset = table[i].addr;
393 overflow = (offset < 0 || offset > INT_MAX);
395 offset = relative_base - table[i].addr - 1;
396 overflow = (offset < INT_MIN || offset >= 0);
399 fprintf(stderr, "kallsyms failure: "
400 "%s symbol value %#llx out of range in relative mode\n",
401 symbol_absolute(&table[i]) ? "absolute" : "relative",
405 printf("\t.long\t%#x\n", (int)offset);
406 } else if (!symbol_absolute(&table[i])) {
407 if (_text <= table[i].addr)
408 printf("\tPTR\t_text + %#llx\n",
409 table[i].addr - _text);
411 printf("\tPTR\t_text - %#llx\n",
412 _text - table[i].addr);
414 printf("\tPTR\t%#llx\n", table[i].addr);
420 output_label("kallsyms_relative_base");
421 printf("\tPTR\t_text - %#llx\n", _text - relative_base);
425 output_label("kallsyms_num_syms");
426 printf("\t.long\t%u\n", table_cnt);
429 /* table of offset markers, that give the offset in the compressed stream
430 * every 256 symbols */
431 markers = malloc(sizeof(unsigned int) * ((table_cnt + 255) / 256));
433 fprintf(stderr, "kallsyms failure: "
434 "unable to allocate required memory\n");
438 output_label("kallsyms_names");
440 for (i = 0; i < table_cnt; i++) {
442 markers[i >> 8] = off;
444 printf("\t.byte 0x%02x", table[i].len);
445 for (k = 0; k < table[i].len; k++)
446 printf(", 0x%02x", table[i].sym[k]);
449 off += table[i].len + 1;
453 output_label("kallsyms_markers");
454 for (i = 0; i < ((table_cnt + 255) >> 8); i++)
455 printf("\t.long\t%u\n", markers[i]);
460 output_label("kallsyms_token_table");
462 for (i = 0; i < 256; i++) {
464 expand_symbol(best_table[i], best_table_len[i], buf);
465 printf("\t.asciz\t\"%s\"\n", buf);
466 off += strlen(buf) + 1;
470 output_label("kallsyms_token_index");
471 for (i = 0; i < 256; i++)
472 printf("\t.short\t%d\n", best_idx[i]);
477 /* table lookup compression functions */
479 /* count all the possible tokens in a symbol */
480 static void learn_symbol(unsigned char *symbol, int len)
484 for (i = 0; i < len - 1; i++)
485 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]++;
488 /* decrease the count for all the possible tokens in a symbol */
489 static void forget_symbol(unsigned char *symbol, int len)
493 for (i = 0; i < len - 1; i++)
494 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]--;
497 /* do the initial token count */
498 static void build_initial_tok_table(void)
502 for (i = 0; i < table_cnt; i++)
503 learn_symbol(table[i].sym, table[i].len);
506 static void *find_token(unsigned char *str, int len, unsigned char *token)
510 for (i = 0; i < len - 1; i++) {
511 if (str[i] == token[0] && str[i+1] == token[1])
517 /* replace a given token in all the valid symbols. Use the sampled symbols
518 * to update the counts */
519 static void compress_symbols(unsigned char *str, int idx)
521 unsigned int i, len, size;
522 unsigned char *p1, *p2;
524 for (i = 0; i < table_cnt; i++) {
529 /* find the token on the symbol */
530 p2 = find_token(p1, len, str);
533 /* decrease the counts for this symbol's tokens */
534 forget_symbol(table[i].sym, len);
542 memmove(p2, p2 + 1, size);
548 /* find the token on the symbol */
549 p2 = find_token(p1, size, str);
555 /* increase the counts for this symbol's new tokens */
556 learn_symbol(table[i].sym, len);
560 /* search the token with the maximum profit */
561 static int find_best_token(void)
563 int i, best, bestprofit;
568 for (i = 0; i < 0x10000; i++) {
569 if (token_profit[i] > bestprofit) {
571 bestprofit = token_profit[i];
577 /* this is the core of the algorithm: calculate the "best" table */
578 static void optimize_result(void)
582 /* using the '\0' symbol last allows compress_symbols to use standard
583 * fast string functions */
584 for (i = 255; i >= 0; i--) {
586 /* if this table slot is empty (it is not used by an actual
587 * original char code */
588 if (!best_table_len[i]) {
590 /* find the token with the best profit value */
591 best = find_best_token();
592 if (token_profit[best] == 0)
595 /* place it in the "best" table */
596 best_table_len[i] = 2;
597 best_table[i][0] = best & 0xFF;
598 best_table[i][1] = (best >> 8) & 0xFF;
600 /* replace this token in all the valid symbols */
601 compress_symbols(best_table[i], i);
606 /* start by placing the symbols that are actually used on the table */
607 static void insert_real_symbols_in_table(void)
609 unsigned int i, j, c;
611 for (i = 0; i < table_cnt; i++) {
612 for (j = 0; j < table[i].len; j++) {
620 static void optimize_token_table(void)
622 build_initial_tok_table();
624 insert_real_symbols_in_table();
629 /* guess for "linker script provide" symbol */
630 static int may_be_linker_script_provide_symbol(const struct sym_entry *se)
632 const char *symbol = sym_name(se);
633 int len = se->len - 1;
638 if (symbol[0] != '_' || symbol[1] != '_')
642 if (!memcmp(symbol + 2, "start_", 6))
646 if (!memcmp(symbol + 2, "stop_", 5))
650 if (!memcmp(symbol + 2, "end_", 4))
654 if (!memcmp(symbol + len - 6, "_start", 6))
658 if (!memcmp(symbol + len - 4, "_end", 4))
664 static int compare_symbols(const void *a, const void *b)
666 const struct sym_entry *sa;
667 const struct sym_entry *sb;
673 /* sort by address first */
674 if (sa->addr > sb->addr)
676 if (sa->addr < sb->addr)
679 /* sort by "weakness" type */
680 wa = (sa->sym[0] == 'w') || (sa->sym[0] == 'W');
681 wb = (sb->sym[0] == 'w') || (sb->sym[0] == 'W');
685 /* sort by "linker script provide" type */
686 wa = may_be_linker_script_provide_symbol(sa);
687 wb = may_be_linker_script_provide_symbol(sb);
691 /* sort by the number of prefix underscores */
692 wa = strspn(sym_name(sa), "_");
693 wb = strspn(sym_name(sb), "_");
697 /* sort by initial order, so that other symbols are left undisturbed */
698 return sa->start_pos - sb->start_pos;
701 static void sort_symbols(void)
703 qsort(table, table_cnt, sizeof(struct sym_entry), compare_symbols);
706 static void make_percpus_absolute(void)
710 for (i = 0; i < table_cnt; i++)
711 if (symbol_in_range(&table[i], &percpu_range, 1)) {
713 * Keep the 'A' override for percpu symbols to
714 * ensure consistent behavior compared to older
715 * versions of this tool.
717 table[i].sym[0] = 'A';
718 table[i].percpu_absolute = 1;
722 /* find the minimum non-absolute symbol address */
723 static void record_relative_base(void)
727 for (i = 0; i < table_cnt; i++)
728 if (!symbol_absolute(&table[i])) {
730 * The table is sorted by address.
731 * Take the first non-absolute symbol value.
733 relative_base = table[i].addr;
738 int main(int argc, char **argv)
742 for (i = 1; i < argc; i++) {
743 if(strcmp(argv[i], "--all-symbols") == 0)
745 else if (strcmp(argv[i], "--absolute-percpu") == 0)
747 else if (strcmp(argv[i], "--base-relative") == 0)
752 } else if (argc != 1)
758 make_percpus_absolute();
761 record_relative_base();
762 optimize_token_table();