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 KSYM_NAME_LEN 128
33 unsigned long long addr;
35 unsigned int start_pos;
37 unsigned int percpu_absolute;
41 const char *start_sym, *end_sym;
42 unsigned long long start, end;
45 static unsigned long long _text;
46 static unsigned long long relative_base;
47 static struct addr_range text_ranges[] = {
48 { "_stext", "_etext" },
49 { "_sinittext", "_einittext" },
51 #define text_range_text (&text_ranges[0])
52 #define text_range_inittext (&text_ranges[1])
54 static struct addr_range percpu_range = {
55 "__per_cpu_start", "__per_cpu_end", -1ULL, 0
58 static struct sym_entry *table;
59 static unsigned int table_size, table_cnt;
60 static int all_symbols;
61 static int absolute_percpu;
62 static int base_relative;
64 static int token_profit[0x10000];
66 /* the table that holds the result of the compression */
67 static unsigned char best_table[256][2];
68 static unsigned char best_table_len[256];
71 static void usage(void)
73 fprintf(stderr, "Usage: kallsyms [--all-symbols] "
74 "[--base-relative] < in.map > out.S\n");
78 static char *sym_name(const struct sym_entry *s)
80 return (char *)s->sym + 1;
83 static bool is_ignored_symbol(const char *name, char type)
85 static const char * const ignored_symbols[] = {
87 * Symbols which vary between passes. Passes 1 and 2 must have
88 * identical symbol lists. The kallsyms_* symbols below are
89 * only added after pass 1, they would be included in pass 2
90 * when --all-symbols is specified so exclude them to get a
95 "kallsyms_relative_base",
99 "kallsyms_token_table",
100 "kallsyms_token_index",
101 /* Exclude linker generated symbols which vary between passes */
102 "_SDA_BASE_", /* ppc */
103 "_SDA2_BASE_", /* ppc */
107 static const char * const ignored_prefixes[] = {
108 "$", /* local symbols for ARM, MIPS, etc. */
109 ".LASANPC", /* s390 kasan local symbols */
110 "__crc_", /* modversions */
111 "__efistub_", /* arm64 EFI stub namespace */
115 static const char * const ignored_suffixes[] = {
116 "_from_arm", /* arm */
117 "_from_thumb", /* arm */
122 const char * const *p;
124 /* Exclude symbols which vary between passes. */
125 for (p = ignored_symbols; *p; p++)
126 if (!strcmp(name, *p))
129 for (p = ignored_prefixes; *p; p++)
130 if (!strncmp(name, *p, strlen(*p)))
133 for (p = ignored_suffixes; *p; p++) {
134 int l = strlen(name) - strlen(*p);
136 if (l >= 0 && !strcmp(name + l, *p))
140 if (type == 'U' || type == 'u')
142 /* exclude debugging symbols */
143 if (type == 'N' || type == 'n')
146 if (toupper(type) == 'A') {
147 /* Keep these useful absolute symbols */
148 if (strcmp(name, "__kernel_syscall_via_break") &&
149 strcmp(name, "__kernel_syscall_via_epc") &&
150 strcmp(name, "__kernel_sigtramp") &&
151 strcmp(name, "__gp"))
158 static void check_symbol_range(const char *sym, unsigned long long addr,
159 struct addr_range *ranges, int entries)
162 struct addr_range *ar;
164 for (i = 0; i < entries; ++i) {
167 if (strcmp(sym, ar->start_sym) == 0) {
170 } else if (strcmp(sym, ar->end_sym) == 0) {
177 static int read_symbol(FILE *in, struct sym_entry *s)
179 char sym[500], stype;
182 rc = fscanf(in, "%llx %c %499s\n", &s->addr, &stype, sym);
184 if (rc != EOF && fgets(sym, 500, in) == NULL)
185 fprintf(stderr, "Read error or end of file.\n");
188 if (strlen(sym) >= KSYM_NAME_LEN) {
189 fprintf(stderr, "Symbol %s too long for kallsyms (%zu >= %d).\n"
190 "Please increase KSYM_NAME_LEN both in kernel and kallsyms.c\n",
191 sym, strlen(sym), KSYM_NAME_LEN);
195 if (is_ignored_symbol(sym, stype))
198 /* Ignore most absolute/undefined (?) symbols. */
199 if (strcmp(sym, "_text") == 0)
202 check_symbol_range(sym, s->addr, text_ranges, ARRAY_SIZE(text_ranges));
203 check_symbol_range(sym, s->addr, &percpu_range, 1);
205 /* include the type field in the symbol name, so that it gets
206 * compressed together */
207 s->len = strlen(sym) + 1;
208 s->sym = malloc(s->len + 1);
210 fprintf(stderr, "kallsyms failure: "
211 "unable to allocate required amount of memory\n");
214 strcpy(sym_name(s), sym);
217 s->percpu_absolute = 0;
222 static int symbol_in_range(const struct sym_entry *s,
223 const struct addr_range *ranges, int entries)
226 const struct addr_range *ar;
228 for (i = 0; i < entries; ++i) {
231 if (s->addr >= ar->start && s->addr <= ar->end)
238 static int symbol_valid(const struct sym_entry *s)
240 const char *name = sym_name(s);
242 /* if --all-symbols is not specified, then symbols outside the text
243 * and inittext sections are discarded */
245 if (symbol_in_range(s, text_ranges,
246 ARRAY_SIZE(text_ranges)) == 0)
248 /* Corner case. Discard any symbols with the same value as
249 * _etext _einittext; they can move between pass 1 and 2 when
250 * the kallsyms data are added. If these symbols move then
251 * they may get dropped in pass 2, which breaks the kallsyms
254 if ((s->addr == text_range_text->end &&
255 strcmp(name, text_range_text->end_sym)) ||
256 (s->addr == text_range_inittext->end &&
257 strcmp(name, text_range_inittext->end_sym)))
264 /* remove all the invalid symbols from the table */
265 static void shrink_table(void)
270 for (i = 0; i < table_cnt; i++) {
271 if (symbol_valid(&table[i])) {
273 table[pos] = table[i];
281 /* When valid symbol is not registered, exit to error */
283 fprintf(stderr, "No valid symbol.\n");
288 static void read_map(FILE *in)
291 if (table_cnt >= table_size) {
293 table = realloc(table, sizeof(*table) * table_size);
295 fprintf(stderr, "out of memory\n");
299 if (read_symbol(in, &table[table_cnt]) == 0) {
300 table[table_cnt].start_pos = table_cnt;
306 static void output_label(const char *label)
308 printf(".globl %s\n", label);
310 printf("%s:\n", label);
313 /* Provide proper symbols relocatability by their '_text' relativeness. */
314 static void output_address(unsigned long long addr)
317 printf("\tPTR\t_text + %#llx\n", addr - _text);
319 printf("\tPTR\t_text - %#llx\n", _text - addr);
322 /* uncompress a compressed symbol. When this function is called, the best table
323 * might still be compressed itself, so the function needs to be recursive */
324 static int expand_symbol(const unsigned char *data, int len, char *result)
326 int c, rlen, total=0;
330 /* if the table holds a single char that is the same as the one
331 * we are looking for, then end the search */
332 if (best_table[c][0]==c && best_table_len[c]==1) {
336 /* if not, recurse and expand */
337 rlen = expand_symbol(best_table[c], best_table_len[c], result);
349 static int symbol_absolute(const struct sym_entry *s)
351 return s->percpu_absolute;
354 static void write_src(void)
356 unsigned int i, k, off;
357 unsigned int best_idx[256];
358 unsigned int *markers;
359 char buf[KSYM_NAME_LEN];
361 printf("#include <asm/bitsperlong.h>\n");
362 printf("#if BITS_PER_LONG == 64\n");
363 printf("#define PTR .quad\n");
364 printf("#define ALGN .balign 8\n");
366 printf("#define PTR .long\n");
367 printf("#define ALGN .balign 4\n");
370 printf("\t.section .rodata, \"a\"\n");
373 output_label("kallsyms_addresses");
375 output_label("kallsyms_offsets");
377 for (i = 0; i < table_cnt; i++) {
380 * Use the offset relative to the lowest value
381 * encountered of all relative symbols, and emit
382 * non-relocatable fixed offsets that will be fixed
389 if (!absolute_percpu) {
390 offset = table[i].addr - relative_base;
391 overflow = (offset < 0 || offset > UINT_MAX);
392 } else if (symbol_absolute(&table[i])) {
393 offset = table[i].addr;
394 overflow = (offset < 0 || offset > INT_MAX);
396 offset = relative_base - table[i].addr - 1;
397 overflow = (offset < INT_MIN || offset >= 0);
400 fprintf(stderr, "kallsyms failure: "
401 "%s symbol value %#llx out of range in relative mode\n",
402 symbol_absolute(&table[i]) ? "absolute" : "relative",
406 printf("\t.long\t%#x\n", (int)offset);
407 } else if (!symbol_absolute(&table[i])) {
408 output_address(table[i].addr);
410 printf("\tPTR\t%#llx\n", table[i].addr);
416 output_label("kallsyms_relative_base");
417 output_address(relative_base);
421 output_label("kallsyms_num_syms");
422 printf("\t.long\t%u\n", table_cnt);
425 /* table of offset markers, that give the offset in the compressed stream
426 * every 256 symbols */
427 markers = malloc(sizeof(unsigned int) * ((table_cnt + 255) / 256));
429 fprintf(stderr, "kallsyms failure: "
430 "unable to allocate required memory\n");
434 output_label("kallsyms_names");
436 for (i = 0; i < table_cnt; i++) {
438 markers[i >> 8] = off;
440 printf("\t.byte 0x%02x", table[i].len);
441 for (k = 0; k < table[i].len; k++)
442 printf(", 0x%02x", table[i].sym[k]);
445 off += table[i].len + 1;
449 output_label("kallsyms_markers");
450 for (i = 0; i < ((table_cnt + 255) >> 8); i++)
451 printf("\t.long\t%u\n", markers[i]);
456 output_label("kallsyms_token_table");
458 for (i = 0; i < 256; i++) {
460 expand_symbol(best_table[i], best_table_len[i], buf);
461 printf("\t.asciz\t\"%s\"\n", buf);
462 off += strlen(buf) + 1;
466 output_label("kallsyms_token_index");
467 for (i = 0; i < 256; i++)
468 printf("\t.short\t%d\n", best_idx[i]);
473 /* table lookup compression functions */
475 /* count all the possible tokens in a symbol */
476 static void learn_symbol(const unsigned char *symbol, int len)
480 for (i = 0; i < len - 1; i++)
481 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]++;
484 /* decrease the count for all the possible tokens in a symbol */
485 static void forget_symbol(const unsigned char *symbol, int len)
489 for (i = 0; i < len - 1; i++)
490 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]--;
493 /* do the initial token count */
494 static void build_initial_tok_table(void)
498 for (i = 0; i < table_cnt; i++)
499 learn_symbol(table[i].sym, table[i].len);
502 static unsigned char *find_token(unsigned char *str, int len,
503 const unsigned char *token)
507 for (i = 0; i < len - 1; i++) {
508 if (str[i] == token[0] && str[i+1] == token[1])
514 /* replace a given token in all the valid symbols. Use the sampled symbols
515 * to update the counts */
516 static void compress_symbols(const unsigned char *str, int idx)
518 unsigned int i, len, size;
519 unsigned char *p1, *p2;
521 for (i = 0; i < table_cnt; i++) {
526 /* find the token on the symbol */
527 p2 = find_token(p1, len, str);
530 /* decrease the counts for this symbol's tokens */
531 forget_symbol(table[i].sym, len);
539 memmove(p2, p2 + 1, size);
545 /* find the token on the symbol */
546 p2 = find_token(p1, size, str);
552 /* increase the counts for this symbol's new tokens */
553 learn_symbol(table[i].sym, len);
557 /* search the token with the maximum profit */
558 static int find_best_token(void)
560 int i, best, bestprofit;
565 for (i = 0; i < 0x10000; i++) {
566 if (token_profit[i] > bestprofit) {
568 bestprofit = token_profit[i];
574 /* this is the core of the algorithm: calculate the "best" table */
575 static void optimize_result(void)
579 /* using the '\0' symbol last allows compress_symbols to use standard
580 * fast string functions */
581 for (i = 255; i >= 0; i--) {
583 /* if this table slot is empty (it is not used by an actual
584 * original char code */
585 if (!best_table_len[i]) {
587 /* find the token with the best profit value */
588 best = find_best_token();
589 if (token_profit[best] == 0)
592 /* place it in the "best" table */
593 best_table_len[i] = 2;
594 best_table[i][0] = best & 0xFF;
595 best_table[i][1] = (best >> 8) & 0xFF;
597 /* replace this token in all the valid symbols */
598 compress_symbols(best_table[i], i);
603 /* start by placing the symbols that are actually used on the table */
604 static void insert_real_symbols_in_table(void)
606 unsigned int i, j, c;
608 for (i = 0; i < table_cnt; i++) {
609 for (j = 0; j < table[i].len; j++) {
617 static void optimize_token_table(void)
619 build_initial_tok_table();
621 insert_real_symbols_in_table();
626 /* guess for "linker script provide" symbol */
627 static int may_be_linker_script_provide_symbol(const struct sym_entry *se)
629 const char *symbol = sym_name(se);
630 int len = se->len - 1;
635 if (symbol[0] != '_' || symbol[1] != '_')
639 if (!memcmp(symbol + 2, "start_", 6))
643 if (!memcmp(symbol + 2, "stop_", 5))
647 if (!memcmp(symbol + 2, "end_", 4))
651 if (!memcmp(symbol + len - 6, "_start", 6))
655 if (!memcmp(symbol + len - 4, "_end", 4))
661 static int compare_symbols(const void *a, const void *b)
663 const struct sym_entry *sa;
664 const struct sym_entry *sb;
670 /* sort by address first */
671 if (sa->addr > sb->addr)
673 if (sa->addr < sb->addr)
676 /* sort by "weakness" type */
677 wa = (sa->sym[0] == 'w') || (sa->sym[0] == 'W');
678 wb = (sb->sym[0] == 'w') || (sb->sym[0] == 'W');
682 /* sort by "linker script provide" type */
683 wa = may_be_linker_script_provide_symbol(sa);
684 wb = may_be_linker_script_provide_symbol(sb);
688 /* sort by the number of prefix underscores */
689 wa = strspn(sym_name(sa), "_");
690 wb = strspn(sym_name(sb), "_");
694 /* sort by initial order, so that other symbols are left undisturbed */
695 return sa->start_pos - sb->start_pos;
698 static void sort_symbols(void)
700 qsort(table, table_cnt, sizeof(struct sym_entry), compare_symbols);
703 static void make_percpus_absolute(void)
707 for (i = 0; i < table_cnt; i++)
708 if (symbol_in_range(&table[i], &percpu_range, 1)) {
710 * Keep the 'A' override for percpu symbols to
711 * ensure consistent behavior compared to older
712 * versions of this tool.
714 table[i].sym[0] = 'A';
715 table[i].percpu_absolute = 1;
719 /* find the minimum non-absolute symbol address */
720 static void record_relative_base(void)
724 for (i = 0; i < table_cnt; i++)
725 if (!symbol_absolute(&table[i])) {
727 * The table is sorted by address.
728 * Take the first non-absolute symbol value.
730 relative_base = table[i].addr;
735 int main(int argc, char **argv)
739 for (i = 1; i < argc; i++) {
740 if(strcmp(argv[i], "--all-symbols") == 0)
742 else if (strcmp(argv[i], "--absolute-percpu") == 0)
744 else if (strcmp(argv[i], "--base-relative") == 0)
749 } else if (argc != 1)
755 make_percpus_absolute();
758 record_relative_base();
759 optimize_token_table();