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
26 #define KSYM_NAME_LEN 128
29 unsigned long long addr;
31 unsigned int start_pos;
35 static struct sym_entry *table;
36 static unsigned int table_size, table_cnt;
37 static unsigned long long _text, _stext, _etext, _sinittext, _einittext;
38 static unsigned long long _stext_l1, _etext_l1, _stext_l2, _etext_l2;
39 static int all_symbols = 0;
40 static char symbol_prefix_char = '\0';
42 int token_profit[0x10000];
44 /* the table that holds the result of the compression */
45 unsigned char best_table[256][2];
46 unsigned char best_table_len[256];
49 static void usage(void)
51 fprintf(stderr, "Usage: kallsyms [--all-symbols] [--symbol-prefix=<prefix char>] < in.map > out.S\n");
56 * This ignores the intensely annoying "mapping symbols" found
57 * in ARM ELF files: $a, $t and $d.
59 static inline int is_arm_mapping_symbol(const char *str)
61 return str[0] == '$' && strchr("atd", str[1])
62 && (str[2] == '\0' || str[2] == '.');
65 static int read_symbol(FILE *in, struct sym_entry *s)
71 rc = fscanf(in, "%llx %c %499s\n", &s->addr, &stype, str);
81 /* skip prefix char */
82 if (symbol_prefix_char && str[0] == symbol_prefix_char)
85 /* Ignore most absolute/undefined (?) symbols. */
86 if (strcmp(sym, "_text") == 0)
88 else if (strcmp(sym, "_stext") == 0)
90 else if (strcmp(sym, "_etext") == 0)
92 else if (strcmp(sym, "_sinittext") == 0)
94 else if (strcmp(sym, "_einittext") == 0)
96 else if (strcmp(sym, "_stext_l1") == 0)
98 else if (strcmp(sym, "_etext_l1") == 0)
100 else if (strcmp(sym, "_stext_l2") == 0)
102 else if (strcmp(sym, "_etext_l2") == 0)
104 else if (toupper(stype) == 'A')
106 /* Keep these useful absolute symbols */
107 if (strcmp(sym, "__kernel_syscall_via_break") &&
108 strcmp(sym, "__kernel_syscall_via_epc") &&
109 strcmp(sym, "__kernel_sigtramp") &&
114 else if (toupper(stype) == 'U' ||
115 is_arm_mapping_symbol(sym))
117 /* exclude also MIPS ELF local symbols ($L123 instead of .L123) */
118 else if (str[0] == '$')
120 /* exclude debugging symbols */
121 else if (stype == 'N')
124 /* include the type field in the symbol name, so that it gets
125 * compressed together */
126 s->len = strlen(str) + 1;
127 s->sym = malloc(s->len + 1);
129 fprintf(stderr, "kallsyms failure: "
130 "unable to allocate required amount of memory\n");
133 strcpy((char *)s->sym + 1, str);
139 static int symbol_valid(struct sym_entry *s)
141 /* Symbols which vary between passes. Passes 1 and 2 must have
142 * identical symbol lists. The kallsyms_* symbols below are only added
143 * after pass 1, they would be included in pass 2 when --all-symbols is
144 * specified so exclude them to get a stable symbol list.
146 static char *special_symbols[] = {
147 "kallsyms_addresses",
151 "kallsyms_token_table",
152 "kallsyms_token_index",
154 /* Exclude linker generated symbols which vary between passes */
155 "_SDA_BASE_", /* ppc */
156 "_SDA2_BASE_", /* ppc */
161 /* skip prefix char */
162 if (symbol_prefix_char && *(s->sym + 1) == symbol_prefix_char)
165 /* if --all-symbols is not specified, then symbols outside the text
166 * and inittext sections are discarded */
168 if ((s->addr < _stext || s->addr > _etext)
169 && (s->addr < _sinittext || s->addr > _einittext)
170 && (s->addr < _stext_l1 || s->addr > _etext_l1)
171 && (s->addr < _stext_l2 || s->addr > _etext_l2))
173 /* Corner case. Discard any symbols with the same value as
174 * _etext _einittext; they can move between pass 1 and 2 when
175 * the kallsyms data are added. If these symbols move then
176 * they may get dropped in pass 2, which breaks the kallsyms
179 if ((s->addr == _etext &&
180 strcmp((char *)s->sym + offset, "_etext")) ||
181 (s->addr == _einittext &&
182 strcmp((char *)s->sym + offset, "_einittext")))
186 /* Exclude symbols which vary between passes. */
187 if (strstr((char *)s->sym + offset, "_compiled."))
190 for (i = 0; special_symbols[i]; i++)
191 if( strcmp((char *)s->sym + offset, special_symbols[i]) == 0 )
197 static void read_map(FILE *in)
200 if (table_cnt >= table_size) {
202 table = realloc(table, sizeof(*table) * table_size);
204 fprintf(stderr, "out of memory\n");
208 if (read_symbol(in, &table[table_cnt]) == 0) {
209 table[table_cnt].start_pos = table_cnt;
215 static void output_label(char *label)
217 if (symbol_prefix_char)
218 printf(".globl %c%s\n", symbol_prefix_char, label);
220 printf(".globl %s\n", label);
222 if (symbol_prefix_char)
223 printf("%c%s:\n", symbol_prefix_char, label);
225 printf("%s:\n", label);
228 /* uncompress a compressed symbol. When this function is called, the best table
229 * might still be compressed itself, so the function needs to be recursive */
230 static int expand_symbol(unsigned char *data, int len, char *result)
232 int c, rlen, total=0;
236 /* if the table holds a single char that is the same as the one
237 * we are looking for, then end the search */
238 if (best_table[c][0]==c && best_table_len[c]==1) {
242 /* if not, recurse and expand */
243 rlen = expand_symbol(best_table[c], best_table_len[c], result);
255 static void write_src(void)
257 unsigned int i, k, off;
258 unsigned int best_idx[256];
259 unsigned int *markers;
260 char buf[KSYM_NAME_LEN];
262 printf("#include <asm/types.h>\n");
263 printf("#if BITS_PER_LONG == 64\n");
264 printf("#define PTR .quad\n");
265 printf("#define ALGN .align 8\n");
267 printf("#define PTR .long\n");
268 printf("#define ALGN .align 4\n");
271 printf("\t.section .rodata, \"a\"\n");
273 /* Provide proper symbols relocatability by their '_text'
274 * relativeness. The symbol names cannot be used to construct
275 * normal symbol references as the list of symbols contains
276 * symbols that are declared static and are private to their
277 * .o files. This prevents .tmp_kallsyms.o or any other
278 * object from referencing them.
280 output_label("kallsyms_addresses");
281 for (i = 0; i < table_cnt; i++) {
282 if (toupper(table[i].sym[0]) != 'A') {
283 if (_text <= table[i].addr)
284 printf("\tPTR\t_text + %#llx\n",
285 table[i].addr - _text);
287 printf("\tPTR\t_text - %#llx\n",
288 _text - table[i].addr);
290 printf("\tPTR\t%#llx\n", table[i].addr);
295 output_label("kallsyms_num_syms");
296 printf("\tPTR\t%d\n", table_cnt);
299 /* table of offset markers, that give the offset in the compressed stream
300 * every 256 symbols */
301 markers = malloc(sizeof(unsigned int) * ((table_cnt + 255) / 256));
303 fprintf(stderr, "kallsyms failure: "
304 "unable to allocate required memory\n");
308 output_label("kallsyms_names");
310 for (i = 0; i < table_cnt; i++) {
312 markers[i >> 8] = off;
314 printf("\t.byte 0x%02x", table[i].len);
315 for (k = 0; k < table[i].len; k++)
316 printf(", 0x%02x", table[i].sym[k]);
319 off += table[i].len + 1;
323 output_label("kallsyms_markers");
324 for (i = 0; i < ((table_cnt + 255) >> 8); i++)
325 printf("\tPTR\t%d\n", markers[i]);
330 output_label("kallsyms_token_table");
332 for (i = 0; i < 256; i++) {
334 expand_symbol(best_table[i], best_table_len[i], buf);
335 printf("\t.asciz\t\"%s\"\n", buf);
336 off += strlen(buf) + 1;
340 output_label("kallsyms_token_index");
341 for (i = 0; i < 256; i++)
342 printf("\t.short\t%d\n", best_idx[i]);
347 /* table lookup compression functions */
349 /* count all the possible tokens in a symbol */
350 static void learn_symbol(unsigned char *symbol, int len)
354 for (i = 0; i < len - 1; i++)
355 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]++;
358 /* decrease the count for all the possible tokens in a symbol */
359 static void forget_symbol(unsigned char *symbol, int len)
363 for (i = 0; i < len - 1; i++)
364 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]--;
367 /* remove all the invalid symbols from the table and do the initial token count */
368 static void build_initial_tok_table(void)
373 for (i = 0; i < table_cnt; i++) {
374 if ( symbol_valid(&table[i]) ) {
376 table[pos] = table[i];
377 learn_symbol(table[pos].sym, table[pos].len);
384 static void *find_token(unsigned char *str, int len, unsigned char *token)
388 for (i = 0; i < len - 1; i++) {
389 if (str[i] == token[0] && str[i+1] == token[1])
395 /* replace a given token in all the valid symbols. Use the sampled symbols
396 * to update the counts */
397 static void compress_symbols(unsigned char *str, int idx)
399 unsigned int i, len, size;
400 unsigned char *p1, *p2;
402 for (i = 0; i < table_cnt; i++) {
407 /* find the token on the symbol */
408 p2 = find_token(p1, len, str);
411 /* decrease the counts for this symbol's tokens */
412 forget_symbol(table[i].sym, len);
420 memmove(p2, p2 + 1, size);
426 /* find the token on the symbol */
427 p2 = find_token(p1, size, str);
433 /* increase the counts for this symbol's new tokens */
434 learn_symbol(table[i].sym, len);
438 /* search the token with the maximum profit */
439 static int find_best_token(void)
441 int i, best, bestprofit;
446 for (i = 0; i < 0x10000; i++) {
447 if (token_profit[i] > bestprofit) {
449 bestprofit = token_profit[i];
455 /* this is the core of the algorithm: calculate the "best" table */
456 static void optimize_result(void)
460 /* using the '\0' symbol last allows compress_symbols to use standard
461 * fast string functions */
462 for (i = 255; i >= 0; i--) {
464 /* if this table slot is empty (it is not used by an actual
465 * original char code */
466 if (!best_table_len[i]) {
468 /* find the token with the breates profit value */
469 best = find_best_token();
471 /* place it in the "best" table */
472 best_table_len[i] = 2;
473 best_table[i][0] = best & 0xFF;
474 best_table[i][1] = (best >> 8) & 0xFF;
476 /* replace this token in all the valid symbols */
477 compress_symbols(best_table[i], i);
482 /* start by placing the symbols that are actually used on the table */
483 static void insert_real_symbols_in_table(void)
485 unsigned int i, j, c;
487 memset(best_table, 0, sizeof(best_table));
488 memset(best_table_len, 0, sizeof(best_table_len));
490 for (i = 0; i < table_cnt; i++) {
491 for (j = 0; j < table[i].len; j++) {
499 static void optimize_token_table(void)
501 build_initial_tok_table();
503 insert_real_symbols_in_table();
505 /* When valid symbol is not registered, exit to error */
507 fprintf(stderr, "No valid symbol.\n");
514 /* guess for "linker script provide" symbol */
515 static int may_be_linker_script_provide_symbol(const struct sym_entry *se)
517 const char *symbol = (char *)se->sym + 1;
518 int len = se->len - 1;
523 if (symbol[0] != '_' || symbol[1] != '_')
527 if (!memcmp(symbol + 2, "start_", 6))
531 if (!memcmp(symbol + 2, "stop_", 5))
535 if (!memcmp(symbol + 2, "end_", 4))
539 if (!memcmp(symbol + len - 6, "_start", 6))
543 if (!memcmp(symbol + len - 4, "_end", 4))
549 static int prefix_underscores_count(const char *str)
551 const char *tail = str;
559 static int compare_symbols(const void *a, const void *b)
561 const struct sym_entry *sa;
562 const struct sym_entry *sb;
568 /* sort by address first */
569 if (sa->addr > sb->addr)
571 if (sa->addr < sb->addr)
574 /* sort by "weakness" type */
575 wa = (sa->sym[0] == 'w') || (sa->sym[0] == 'W');
576 wb = (sb->sym[0] == 'w') || (sb->sym[0] == 'W');
580 /* sort by "linker script provide" type */
581 wa = may_be_linker_script_provide_symbol(sa);
582 wb = may_be_linker_script_provide_symbol(sb);
586 /* sort by the number of prefix underscores */
587 wa = prefix_underscores_count((const char *)sa->sym + 1);
588 wb = prefix_underscores_count((const char *)sb->sym + 1);
592 /* sort by initial order, so that other symbols are left undisturbed */
593 return sa->start_pos - sb->start_pos;
596 static void sort_symbols(void)
598 qsort(table, table_cnt, sizeof(struct sym_entry), compare_symbols);
601 int main(int argc, char **argv)
605 for (i = 1; i < argc; i++) {
606 if(strcmp(argv[i], "--all-symbols") == 0)
608 else if (strncmp(argv[i], "--symbol-prefix=", 16) == 0) {
609 char *p = &argv[i][16];
611 if ((*p == '"' && *(p+2) == '"') || (*p == '\'' && *(p+2) == '\''))
613 symbol_prefix_char = *p;
617 } else if (argc != 1)
622 optimize_token_table();