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]))
30 #define KSYM_NAME_LEN 128
33 unsigned long long addr;
35 unsigned int start_pos;
40 const char *stext, *etext;
41 unsigned long long start, end;
44 static unsigned long long _text;
45 static struct text_range text_ranges[] = {
46 { "_stext", "_etext" },
47 { "_sinittext", "_einittext" },
48 { "_stext_l1", "_etext_l1" }, /* Blackfin on-chip L1 inst SRAM */
49 { "_stext_l2", "_etext_l2" }, /* Blackfin on-chip L2 SRAM */
51 #define text_range_text (&text_ranges[0])
52 #define text_range_inittext (&text_ranges[1])
54 static struct sym_entry *table;
55 static unsigned int table_size, table_cnt;
56 static int all_symbols = 0;
57 static char symbol_prefix_char = '\0';
59 int token_profit[0x10000];
61 /* the table that holds the result of the compression */
62 unsigned char best_table[256][2];
63 unsigned char best_table_len[256];
66 static void usage(void)
68 fprintf(stderr, "Usage: kallsyms [--all-symbols] [--symbol-prefix=<prefix char>] < in.map > out.S\n");
73 * This ignores the intensely annoying "mapping symbols" found
74 * in ARM ELF files: $a, $t and $d.
76 static inline int is_arm_mapping_symbol(const char *str)
78 return str[0] == '$' && strchr("atd", str[1])
79 && (str[2] == '\0' || str[2] == '.');
82 static int read_symbol_tr(const char *sym, unsigned long long addr)
85 struct text_range *tr;
87 for (i = 0; i < ARRAY_SIZE(text_ranges); ++i) {
90 if (strcmp(sym, tr->stext) == 0) {
93 } else if (strcmp(sym, tr->etext) == 0) {
102 static int read_symbol(FILE *in, struct sym_entry *s)
108 rc = fscanf(in, "%llx %c %499s\n", &s->addr, &stype, str);
118 /* skip prefix char */
119 if (symbol_prefix_char && str[0] == symbol_prefix_char)
122 /* Ignore most absolute/undefined (?) symbols. */
123 if (strcmp(sym, "_text") == 0)
125 else if (read_symbol_tr(sym, s->addr) == 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' ||
138 is_arm_mapping_symbol(sym))
140 /* exclude also MIPS ELF local symbols ($L123 instead of .L123) */
141 else if (str[0] == '$')
143 /* exclude debugging symbols */
144 else if (stype == 'N')
147 /* include the type field in the symbol name, so that it gets
148 * compressed together */
149 s->len = strlen(str) + 1;
150 s->sym = malloc(s->len + 1);
152 fprintf(stderr, "kallsyms failure: "
153 "unable to allocate required amount of memory\n");
156 strcpy((char *)s->sym + 1, str);
162 static int symbol_valid_tr(struct sym_entry *s)
165 struct text_range *tr;
167 for (i = 0; i < ARRAY_SIZE(text_ranges); ++i) {
168 tr = &text_ranges[i];
170 if (s->addr >= tr->start && s->addr <= tr->end)
177 static int symbol_valid(struct sym_entry *s)
179 /* Symbols which vary between passes. Passes 1 and 2 must have
180 * identical symbol lists. The kallsyms_* symbols below are only added
181 * after pass 1, they would be included in pass 2 when --all-symbols is
182 * specified so exclude them to get a stable symbol list.
184 static char *special_symbols[] = {
185 "kallsyms_addresses",
189 "kallsyms_token_table",
190 "kallsyms_token_index",
192 /* Exclude linker generated symbols which vary between passes */
193 "_SDA_BASE_", /* ppc */
194 "_SDA2_BASE_", /* ppc */
199 /* skip prefix char */
200 if (symbol_prefix_char && *(s->sym + 1) == symbol_prefix_char)
203 /* if --all-symbols is not specified, then symbols outside the text
204 * and inittext sections are discarded */
206 if (symbol_valid_tr(s) == 0)
208 /* Corner case. Discard any symbols with the same value as
209 * _etext _einittext; they can move between pass 1 and 2 when
210 * the kallsyms data are added. If these symbols move then
211 * they may get dropped in pass 2, which breaks the kallsyms
214 if ((s->addr == text_range_text->end &&
215 strcmp((char *)s->sym + offset, text_range_text->etext)) ||
216 (s->addr == text_range_inittext->end &&
217 strcmp((char *)s->sym + offset, text_range_inittext->etext)))
221 /* Exclude symbols which vary between passes. */
222 if (strstr((char *)s->sym + offset, "_compiled."))
225 for (i = 0; special_symbols[i]; i++)
226 if( strcmp((char *)s->sym + offset, special_symbols[i]) == 0 )
232 static void read_map(FILE *in)
235 if (table_cnt >= table_size) {
237 table = realloc(table, sizeof(*table) * table_size);
239 fprintf(stderr, "out of memory\n");
243 if (read_symbol(in, &table[table_cnt]) == 0) {
244 table[table_cnt].start_pos = table_cnt;
250 static void output_label(char *label)
252 if (symbol_prefix_char)
253 printf(".globl %c%s\n", symbol_prefix_char, label);
255 printf(".globl %s\n", label);
257 if (symbol_prefix_char)
258 printf("%c%s:\n", symbol_prefix_char, label);
260 printf("%s:\n", label);
263 /* uncompress a compressed symbol. When this function is called, the best table
264 * might still be compressed itself, so the function needs to be recursive */
265 static int expand_symbol(unsigned char *data, int len, char *result)
267 int c, rlen, total=0;
271 /* if the table holds a single char that is the same as the one
272 * we are looking for, then end the search */
273 if (best_table[c][0]==c && best_table_len[c]==1) {
277 /* if not, recurse and expand */
278 rlen = expand_symbol(best_table[c], best_table_len[c], result);
290 static void write_src(void)
292 unsigned int i, k, off;
293 unsigned int best_idx[256];
294 unsigned int *markers;
295 char buf[KSYM_NAME_LEN];
297 printf("#include <asm/types.h>\n");
298 printf("#if BITS_PER_LONG == 64\n");
299 printf("#define PTR .quad\n");
300 printf("#define ALGN .align 8\n");
302 printf("#define PTR .long\n");
303 printf("#define ALGN .align 4\n");
306 printf("\t.section .rodata, \"a\"\n");
308 /* Provide proper symbols relocatability by their '_text'
309 * relativeness. The symbol names cannot be used to construct
310 * normal symbol references as the list of symbols contains
311 * symbols that are declared static and are private to their
312 * .o files. This prevents .tmp_kallsyms.o or any other
313 * object from referencing them.
315 output_label("kallsyms_addresses");
316 for (i = 0; i < table_cnt; i++) {
317 if (toupper(table[i].sym[0]) != 'A') {
318 if (_text <= table[i].addr)
319 printf("\tPTR\t_text + %#llx\n",
320 table[i].addr - _text);
322 printf("\tPTR\t_text - %#llx\n",
323 _text - table[i].addr);
325 printf("\tPTR\t%#llx\n", table[i].addr);
330 output_label("kallsyms_num_syms");
331 printf("\tPTR\t%d\n", table_cnt);
334 /* table of offset markers, that give the offset in the compressed stream
335 * every 256 symbols */
336 markers = malloc(sizeof(unsigned int) * ((table_cnt + 255) / 256));
338 fprintf(stderr, "kallsyms failure: "
339 "unable to allocate required memory\n");
343 output_label("kallsyms_names");
345 for (i = 0; i < table_cnt; i++) {
347 markers[i >> 8] = off;
349 printf("\t.byte 0x%02x", table[i].len);
350 for (k = 0; k < table[i].len; k++)
351 printf(", 0x%02x", table[i].sym[k]);
354 off += table[i].len + 1;
358 output_label("kallsyms_markers");
359 for (i = 0; i < ((table_cnt + 255) >> 8); i++)
360 printf("\tPTR\t%d\n", markers[i]);
365 output_label("kallsyms_token_table");
367 for (i = 0; i < 256; i++) {
369 expand_symbol(best_table[i], best_table_len[i], buf);
370 printf("\t.asciz\t\"%s\"\n", buf);
371 off += strlen(buf) + 1;
375 output_label("kallsyms_token_index");
376 for (i = 0; i < 256; i++)
377 printf("\t.short\t%d\n", best_idx[i]);
382 /* table lookup compression functions */
384 /* count all the possible tokens in a symbol */
385 static void learn_symbol(unsigned char *symbol, int len)
389 for (i = 0; i < len - 1; i++)
390 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]++;
393 /* decrease the count for all the possible tokens in a symbol */
394 static void forget_symbol(unsigned char *symbol, int len)
398 for (i = 0; i < len - 1; i++)
399 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]--;
402 /* remove all the invalid symbols from the table and do the initial token count */
403 static void build_initial_tok_table(void)
408 for (i = 0; i < table_cnt; i++) {
409 if ( symbol_valid(&table[i]) ) {
411 table[pos] = table[i];
412 learn_symbol(table[pos].sym, table[pos].len);
419 static void *find_token(unsigned char *str, int len, unsigned char *token)
423 for (i = 0; i < len - 1; i++) {
424 if (str[i] == token[0] && str[i+1] == token[1])
430 /* replace a given token in all the valid symbols. Use the sampled symbols
431 * to update the counts */
432 static void compress_symbols(unsigned char *str, int idx)
434 unsigned int i, len, size;
435 unsigned char *p1, *p2;
437 for (i = 0; i < table_cnt; i++) {
442 /* find the token on the symbol */
443 p2 = find_token(p1, len, str);
446 /* decrease the counts for this symbol's tokens */
447 forget_symbol(table[i].sym, len);
455 memmove(p2, p2 + 1, size);
461 /* find the token on the symbol */
462 p2 = find_token(p1, size, str);
468 /* increase the counts for this symbol's new tokens */
469 learn_symbol(table[i].sym, len);
473 /* search the token with the maximum profit */
474 static int find_best_token(void)
476 int i, best, bestprofit;
481 for (i = 0; i < 0x10000; i++) {
482 if (token_profit[i] > bestprofit) {
484 bestprofit = token_profit[i];
490 /* this is the core of the algorithm: calculate the "best" table */
491 static void optimize_result(void)
495 /* using the '\0' symbol last allows compress_symbols to use standard
496 * fast string functions */
497 for (i = 255; i >= 0; i--) {
499 /* if this table slot is empty (it is not used by an actual
500 * original char code */
501 if (!best_table_len[i]) {
503 /* find the token with the breates profit value */
504 best = find_best_token();
506 /* place it in the "best" table */
507 best_table_len[i] = 2;
508 best_table[i][0] = best & 0xFF;
509 best_table[i][1] = (best >> 8) & 0xFF;
511 /* replace this token in all the valid symbols */
512 compress_symbols(best_table[i], i);
517 /* start by placing the symbols that are actually used on the table */
518 static void insert_real_symbols_in_table(void)
520 unsigned int i, j, c;
522 memset(best_table, 0, sizeof(best_table));
523 memset(best_table_len, 0, sizeof(best_table_len));
525 for (i = 0; i < table_cnt; i++) {
526 for (j = 0; j < table[i].len; j++) {
534 static void optimize_token_table(void)
536 build_initial_tok_table();
538 insert_real_symbols_in_table();
540 /* When valid symbol is not registered, exit to error */
542 fprintf(stderr, "No valid symbol.\n");
549 /* guess for "linker script provide" symbol */
550 static int may_be_linker_script_provide_symbol(const struct sym_entry *se)
552 const char *symbol = (char *)se->sym + 1;
553 int len = se->len - 1;
558 if (symbol[0] != '_' || symbol[1] != '_')
562 if (!memcmp(symbol + 2, "start_", 6))
566 if (!memcmp(symbol + 2, "stop_", 5))
570 if (!memcmp(symbol + 2, "end_", 4))
574 if (!memcmp(symbol + len - 6, "_start", 6))
578 if (!memcmp(symbol + len - 4, "_end", 4))
584 static int prefix_underscores_count(const char *str)
586 const char *tail = str;
594 static int compare_symbols(const void *a, const void *b)
596 const struct sym_entry *sa;
597 const struct sym_entry *sb;
603 /* sort by address first */
604 if (sa->addr > sb->addr)
606 if (sa->addr < sb->addr)
609 /* sort by "weakness" type */
610 wa = (sa->sym[0] == 'w') || (sa->sym[0] == 'W');
611 wb = (sb->sym[0] == 'w') || (sb->sym[0] == 'W');
615 /* sort by "linker script provide" type */
616 wa = may_be_linker_script_provide_symbol(sa);
617 wb = may_be_linker_script_provide_symbol(sb);
621 /* sort by the number of prefix underscores */
622 wa = prefix_underscores_count((const char *)sa->sym + 1);
623 wb = prefix_underscores_count((const char *)sb->sym + 1);
627 /* sort by initial order, so that other symbols are left undisturbed */
628 return sa->start_pos - sb->start_pos;
631 static void sort_symbols(void)
633 qsort(table, table_cnt, sizeof(struct sym_entry), compare_symbols);
636 int main(int argc, char **argv)
640 for (i = 1; i < argc; i++) {
641 if(strcmp(argv[i], "--all-symbols") == 0)
643 else if (strncmp(argv[i], "--symbol-prefix=", 16) == 0) {
644 char *p = &argv[i][16];
646 if ((*p == '"' && *(p+2) == '"') || (*p == '\'' && *(p+2) == '\''))
648 symbol_prefix_char = *p;
652 } else if (argc != 1)
657 optimize_token_table();