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]))
31 #define KSYM_NAME_LEN 128
34 unsigned long long addr;
36 unsigned int start_pos;
38 unsigned int percpu_absolute;
42 const char *start_sym, *end_sym;
43 unsigned long long start, end;
46 static unsigned long long _text;
47 static unsigned long long relative_base;
48 static struct addr_range text_ranges[] = {
49 { "_stext", "_etext" },
50 { "_sinittext", "_einittext" },
51 { "_stext_l1", "_etext_l1" }, /* Blackfin on-chip L1 inst SRAM */
52 { "_stext_l2", "_etext_l2" }, /* Blackfin on-chip L2 SRAM */
54 #define text_range_text (&text_ranges[0])
55 #define text_range_inittext (&text_ranges[1])
57 static struct addr_range percpu_range = {
58 "__per_cpu_start", "__per_cpu_end", -1ULL, 0
61 static struct sym_entry *table;
62 static unsigned int table_size, table_cnt;
63 static int all_symbols = 0;
64 static int absolute_percpu = 0;
65 static char symbol_prefix_char = '\0';
66 static int base_relative = 0;
68 int token_profit[0x10000];
70 /* the table that holds the result of the compression */
71 unsigned char best_table[256][2];
72 unsigned char best_table_len[256];
75 static void usage(void)
77 fprintf(stderr, "Usage: kallsyms [--all-symbols] "
78 "[--symbol-prefix=<prefix char>] "
79 "[--page-offset=<CONFIG_PAGE_OFFSET>] "
80 "[--base-relative] < in.map > out.S\n");
85 * This ignores the intensely annoying "mapping symbols" found
86 * in ARM ELF files: $a, $t and $d.
88 static inline int is_arm_mapping_symbol(const char *str)
90 return str[0] == '$' && strchr("axtd", str[1])
91 && (str[2] == '\0' || str[2] == '.');
94 static int check_symbol_range(const char *sym, unsigned long long addr,
95 struct addr_range *ranges, int entries)
98 struct addr_range *ar;
100 for (i = 0; i < entries; ++i) {
103 if (strcmp(sym, ar->start_sym) == 0) {
106 } else if (strcmp(sym, ar->end_sym) == 0) {
115 static int read_symbol(FILE *in, struct sym_entry *s)
121 rc = fscanf(in, "%llx %c %499s\n", &s->addr, &stype, str);
123 if (rc != EOF && fgets(str, 500, in) == NULL)
124 fprintf(stderr, "Read error or end of file.\n");
127 if (strlen(str) > KSYM_NAME_LEN) {
128 fprintf(stderr, "Symbol %s too long for kallsyms (%zu vs %d).\n"
129 "Please increase KSYM_NAME_LEN both in kernel and kallsyms.c\n",
130 str, strlen(str), KSYM_NAME_LEN);
135 /* skip prefix char */
136 if (symbol_prefix_char && str[0] == symbol_prefix_char)
139 /* Ignore most absolute/undefined (?) symbols. */
140 if (strcmp(sym, "_text") == 0)
142 else if (check_symbol_range(sym, s->addr, text_ranges,
143 ARRAY_SIZE(text_ranges)) == 0)
145 else if (toupper(stype) == 'A')
147 /* Keep these useful absolute symbols */
148 if (strcmp(sym, "__kernel_syscall_via_break") &&
149 strcmp(sym, "__kernel_syscall_via_epc") &&
150 strcmp(sym, "__kernel_sigtramp") &&
155 else if (toupper(stype) == 'U' ||
156 is_arm_mapping_symbol(sym))
158 /* exclude also MIPS ELF local symbols ($L123 instead of .L123) */
159 else if (str[0] == '$')
161 /* exclude debugging symbols */
162 else if (stype == 'N')
165 /* include the type field in the symbol name, so that it gets
166 * compressed together */
167 s->len = strlen(str) + 1;
168 s->sym = malloc(s->len + 1);
170 fprintf(stderr, "kallsyms failure: "
171 "unable to allocate required amount of memory\n");
174 strcpy((char *)s->sym + 1, str);
177 s->percpu_absolute = 0;
179 /* Record if we've found __per_cpu_start/end. */
180 check_symbol_range(sym, s->addr, &percpu_range, 1);
185 static int symbol_in_range(struct sym_entry *s, struct addr_range *ranges,
189 struct addr_range *ar;
191 for (i = 0; i < entries; ++i) {
194 if (s->addr >= ar->start && s->addr <= ar->end)
201 static int symbol_valid(struct sym_entry *s)
203 /* Symbols which vary between passes. Passes 1 and 2 must have
204 * identical symbol lists. The kallsyms_* symbols below are only added
205 * after pass 1, they would be included in pass 2 when --all-symbols is
206 * specified so exclude them to get a stable symbol list.
208 static char *special_symbols[] = {
209 "kallsyms_addresses",
211 "kallsyms_relative_base",
215 "kallsyms_token_table",
216 "kallsyms_token_index",
218 /* Exclude linker generated symbols which vary between passes */
219 "_SDA_BASE_", /* ppc */
220 "_SDA2_BASE_", /* ppc */
223 static char *special_suffixes[] = {
225 "_from_arm", /* arm */
226 "_from_thumb", /* arm */
230 char *sym_name = (char *)s->sym + 1;
232 /* skip prefix char */
233 if (symbol_prefix_char && *sym_name == symbol_prefix_char)
237 /* if --all-symbols is not specified, then symbols outside the text
238 * and inittext sections are discarded */
240 if (symbol_in_range(s, text_ranges,
241 ARRAY_SIZE(text_ranges)) == 0)
243 /* Corner case. Discard any symbols with the same value as
244 * _etext _einittext; they can move between pass 1 and 2 when
245 * the kallsyms data are added. If these symbols move then
246 * they may get dropped in pass 2, which breaks the kallsyms
249 if ((s->addr == text_range_text->end &&
251 text_range_text->end_sym)) ||
252 (s->addr == text_range_inittext->end &&
254 text_range_inittext->end_sym)))
258 /* Exclude symbols which vary between passes. */
259 for (i = 0; special_symbols[i]; i++)
260 if (strcmp(sym_name, special_symbols[i]) == 0)
263 for (i = 0; special_suffixes[i]; i++) {
264 int l = strlen(sym_name) - strlen(special_suffixes[i]);
266 if (l >= 0 && strcmp(sym_name + l, special_suffixes[i]) == 0)
273 static void read_map(FILE *in)
276 if (table_cnt >= table_size) {
278 table = realloc(table, sizeof(*table) * table_size);
280 fprintf(stderr, "out of memory\n");
284 if (read_symbol(in, &table[table_cnt]) == 0) {
285 table[table_cnt].start_pos = table_cnt;
291 static void output_label(char *label)
293 if (symbol_prefix_char)
294 printf(".globl %c%s\n", symbol_prefix_char, label);
296 printf(".globl %s\n", label);
298 if (symbol_prefix_char)
299 printf("%c%s:\n", symbol_prefix_char, label);
301 printf("%s:\n", label);
304 /* uncompress a compressed symbol. When this function is called, the best table
305 * might still be compressed itself, so the function needs to be recursive */
306 static int expand_symbol(unsigned char *data, int len, char *result)
308 int c, rlen, total=0;
312 /* if the table holds a single char that is the same as the one
313 * we are looking for, then end the search */
314 if (best_table[c][0]==c && best_table_len[c]==1) {
318 /* if not, recurse and expand */
319 rlen = expand_symbol(best_table[c], best_table_len[c], result);
331 static int symbol_absolute(struct sym_entry *s)
333 return s->percpu_absolute;
336 static void write_src(void)
338 unsigned int i, k, off;
339 unsigned int best_idx[256];
340 unsigned int *markers;
341 char buf[KSYM_NAME_LEN];
343 printf("#include <asm/types.h>\n");
344 printf("#if BITS_PER_LONG == 64\n");
345 printf("#define PTR .quad\n");
346 printf("#define ALGN .align 8\n");
348 printf("#define PTR .long\n");
349 printf("#define ALGN .align 4\n");
352 printf("\t.section .rodata, \"a\"\n");
354 /* Provide proper symbols relocatability by their relativeness
355 * to a fixed anchor point in the runtime image, either '_text'
356 * for absolute address tables, in which case the linker will
357 * emit the final addresses at build time. Otherwise, use the
358 * offset relative to the lowest value encountered of all relative
359 * symbols, and emit non-relocatable fixed offsets that will be fixed
362 * The symbol names cannot be used to construct normal symbol
363 * references as the list of symbols contains symbols that are
364 * declared static and are private to their .o files. This prevents
365 * .tmp_kallsyms.o or any other object from referencing them.
368 output_label("kallsyms_addresses");
370 output_label("kallsyms_offsets");
372 for (i = 0; i < table_cnt; i++) {
377 if (!absolute_percpu) {
378 offset = table[i].addr - relative_base;
379 overflow = (offset < 0 || offset > UINT_MAX);
380 } else if (symbol_absolute(&table[i])) {
381 offset = table[i].addr;
382 overflow = (offset < 0 || offset > INT_MAX);
384 offset = relative_base - table[i].addr - 1;
385 overflow = (offset < INT_MIN || offset >= 0);
388 fprintf(stderr, "kallsyms failure: "
389 "%s symbol value %#llx out of range in relative mode\n",
390 symbol_absolute(&table[i]) ? "absolute" : "relative",
394 printf("\t.long\t%#x\n", (int)offset);
395 } else if (!symbol_absolute(&table[i])) {
396 if (_text <= table[i].addr)
397 printf("\tPTR\t_text + %#llx\n",
398 table[i].addr - _text);
400 printf("\tPTR\t_text - %#llx\n",
401 _text - table[i].addr);
403 printf("\tPTR\t%#llx\n", table[i].addr);
409 output_label("kallsyms_relative_base");
410 printf("\tPTR\t_text - %#llx\n", _text - relative_base);
414 output_label("kallsyms_num_syms");
415 printf("\tPTR\t%d\n", table_cnt);
418 /* table of offset markers, that give the offset in the compressed stream
419 * every 256 symbols */
420 markers = malloc(sizeof(unsigned int) * ((table_cnt + 255) / 256));
422 fprintf(stderr, "kallsyms failure: "
423 "unable to allocate required memory\n");
427 output_label("kallsyms_names");
429 for (i = 0; i < table_cnt; i++) {
431 markers[i >> 8] = off;
433 printf("\t.byte 0x%02x", table[i].len);
434 for (k = 0; k < table[i].len; k++)
435 printf(", 0x%02x", table[i].sym[k]);
438 off += table[i].len + 1;
442 output_label("kallsyms_markers");
443 for (i = 0; i < ((table_cnt + 255) >> 8); i++)
444 printf("\tPTR\t%d\n", markers[i]);
449 output_label("kallsyms_token_table");
451 for (i = 0; i < 256; i++) {
453 expand_symbol(best_table[i], best_table_len[i], buf);
454 printf("\t.asciz\t\"%s\"\n", buf);
455 off += strlen(buf) + 1;
459 output_label("kallsyms_token_index");
460 for (i = 0; i < 256; i++)
461 printf("\t.short\t%d\n", best_idx[i]);
466 /* table lookup compression functions */
468 /* count all the possible tokens in a symbol */
469 static void learn_symbol(unsigned char *symbol, int len)
473 for (i = 0; i < len - 1; i++)
474 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]++;
477 /* decrease the count for all the possible tokens in a symbol */
478 static void forget_symbol(unsigned char *symbol, int len)
482 for (i = 0; i < len - 1; i++)
483 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]--;
486 /* remove all the invalid symbols from the table and do the initial token count */
487 static void build_initial_tok_table(void)
492 for (i = 0; i < table_cnt; i++) {
493 if ( symbol_valid(&table[i]) ) {
495 table[pos] = table[i];
496 learn_symbol(table[pos].sym, table[pos].len);
503 static void *find_token(unsigned char *str, int len, 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(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 breates 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 memset(best_table, 0, sizeof(best_table));
609 memset(best_table_len, 0, sizeof(best_table_len));
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();
626 /* When valid symbol is not registered, exit to error */
628 fprintf(stderr, "No valid symbol.\n");
635 /* guess for "linker script provide" symbol */
636 static int may_be_linker_script_provide_symbol(const struct sym_entry *se)
638 const char *symbol = (char *)se->sym + 1;
639 int len = se->len - 1;
644 if (symbol[0] != '_' || symbol[1] != '_')
648 if (!memcmp(symbol + 2, "start_", 6))
652 if (!memcmp(symbol + 2, "stop_", 5))
656 if (!memcmp(symbol + 2, "end_", 4))
660 if (!memcmp(symbol + len - 6, "_start", 6))
664 if (!memcmp(symbol + len - 4, "_end", 4))
670 static int prefix_underscores_count(const char *str)
672 const char *tail = str;
680 static int compare_symbols(const void *a, const void *b)
682 const struct sym_entry *sa;
683 const struct sym_entry *sb;
689 /* sort by address first */
690 if (sa->addr > sb->addr)
692 if (sa->addr < sb->addr)
695 /* sort by "weakness" type */
696 wa = (sa->sym[0] == 'w') || (sa->sym[0] == 'W');
697 wb = (sb->sym[0] == 'w') || (sb->sym[0] == 'W');
701 /* sort by "linker script provide" type */
702 wa = may_be_linker_script_provide_symbol(sa);
703 wb = may_be_linker_script_provide_symbol(sb);
707 /* sort by the number of prefix underscores */
708 wa = prefix_underscores_count((const char *)sa->sym + 1);
709 wb = prefix_underscores_count((const char *)sb->sym + 1);
713 /* sort by initial order, so that other symbols are left undisturbed */
714 return sa->start_pos - sb->start_pos;
717 static void sort_symbols(void)
719 qsort(table, table_cnt, sizeof(struct sym_entry), compare_symbols);
722 static void make_percpus_absolute(void)
726 for (i = 0; i < table_cnt; i++)
727 if (symbol_in_range(&table[i], &percpu_range, 1)) {
729 * Keep the 'A' override for percpu symbols to
730 * ensure consistent behavior compared to older
731 * versions of this tool.
733 table[i].sym[0] = 'A';
734 table[i].percpu_absolute = 1;
738 /* find the minimum non-absolute symbol address */
739 static void record_relative_base(void)
743 relative_base = -1ULL;
744 for (i = 0; i < table_cnt; i++)
745 if (!symbol_absolute(&table[i]) &&
746 table[i].addr < relative_base)
747 relative_base = table[i].addr;
750 int main(int argc, char **argv)
754 for (i = 1; i < argc; i++) {
755 if(strcmp(argv[i], "--all-symbols") == 0)
757 else if (strcmp(argv[i], "--absolute-percpu") == 0)
759 else if (strncmp(argv[i], "--symbol-prefix=", 16) == 0) {
760 char *p = &argv[i][16];
762 if ((*p == '"' && *(p+2) == '"') || (*p == '\'' && *(p+2) == '\''))
764 symbol_prefix_char = *p;
765 } else if (strcmp(argv[i], "--base-relative") == 0)
770 } else if (argc != 1)
775 make_percpus_absolute();
777 record_relative_base();
779 optimize_token_table();