scripts/kallsyms.c

   1 /* Generate assembler source containing symbol information
   2  *
   3  * Copyright 2002       by Kai Germaschewski
   4  *
   5  * This software may be used and distributed according to the terms
   6  * of the GNU General Public License, incorporated herein by reference.
   7  *
   8  * Usage: nm -n vmlinux | scripts/kallsyms [--all-symbols] > symbols.S
   9  *
  10  * ChangeLog:
  11  *
  12  * (25/Aug/2004) Paulo Marques <pmarques@grupopie.com>
  13  *      Changed the compression method from stem compression to "table lookup"
  14  *      compression
  15  *
  16  *      Table compression uses all the unused char codes on the symbols and
  17  *  maps these to the most used substrings (tokens). For instance, it might
  18  *  map char code 0xF7 to represent "write_" and then in every symbol where
  19  *  "write_" appears it can be replaced by 0xF7, saving 5 bytes.
  20  *      The used codes themselves are also placed in the table so that the
  21  *  decompresion can work without "special cases".
  22  *      Applied to kernel symbols, this usually produces a compression ratio
  23  *  of about 50%.
  24  *
  25  */
  26
  27 #include <stdio.h>
  28 #include <stdlib.h>
  29 #include <string.h>
  30 #include <ctype.h>
  31
  32 #define KSYM_NAME_LEN           128
  33
  34
  35 struct sym_entry {
  36         unsigned long long addr;
  37         unsigned int len;
  38         unsigned char *sym;
  39 };
  40
  41
  42 static struct sym_entry *table;
  43 static unsigned int table_size, table_cnt;
  44 static unsigned long long _text, _stext, _etext, _sinittext, _einittext, _sextratext, _eextratext;
  45 static int all_symbols = 0;
  46 static char symbol_prefix_char = '\0';
  47
  48 int token_profit[0x10000];
  49
  50 /* the table that holds the result of the compression */
  51 unsigned char best_table[256][2];
  52 unsigned char best_table_len[256];
  53
  54
  55 static void usage(void)
  56 {
  57         fprintf(stderr, "Usage: kallsyms [--all-symbols] [--symbol-prefix=<prefix char>] < in.map > out.S\n");
  58         exit(1);
  59 }
  60
  61 /*
  62  * This ignores the intensely annoying "mapping symbols" found
  63  * in ARM ELF files: $a, $t and $d.
  64  */
  65 static inline int is_arm_mapping_symbol(const char *str)
  66 {
  67         return str[0] == '$' && strchr("atd", str[1])
  68                && (str[2] == '\0' || str[2] == '.');
  69 }
  70
  71 static int read_symbol(FILE *in, struct sym_entry *s)
  72 {
  73         char str[500];
  74         char *sym, stype;
  75         int rc;
  76
  77         rc = fscanf(in, "%llx %c %499s\n", &s->addr, &stype, str);
  78         if (rc != 3) {
  79                 if (rc != EOF) {
  80                         /* skip line */
  81                         fgets(str, 500, in);
  82                 }
  83                 return -1;
  84         }
  85
  86         sym = str;
  87         /* skip prefix char */
  88         if (symbol_prefix_char && str[0] == symbol_prefix_char)
  89                 sym++;
  90
  91         /* Ignore most absolute/undefined (?) symbols. */
  92         if (strcmp(sym, "_text") == 0)
  93                 _text = s->addr;
  94         else if (strcmp(sym, "_stext") == 0)
  95                 _stext = s->addr;
  96         else if (strcmp(sym, "_etext") == 0)
  97                 _etext = s->addr;
  98         else if (strcmp(sym, "_sinittext") == 0)
  99                 _sinittext = s->addr;
 100         else if (strcmp(sym, "_einittext") == 0)
 101                 _einittext = s->addr;
 102         else if (strcmp(sym, "_sextratext") == 0)
 103                 _sextratext = s->addr;
 104         else if (strcmp(sym, "_eextratext") == 0)
 105                 _eextratext = s->addr;
 106         else if (toupper(stype) == 'A')
 107         {
 108                 /* Keep these useful absolute symbols */
 109                 if (strcmp(sym, "__kernel_syscall_via_break") &&
 110                     strcmp(sym, "__kernel_syscall_via_epc") &&
 111                     strcmp(sym, "__kernel_sigtramp") &&
 112                     strcmp(sym, "__gp"))
 113                         return -1;
 114
 115         }
 116         else if (toupper(stype) == 'U' ||
 117                  is_arm_mapping_symbol(sym))
 118                 return -1;
 119         /* exclude also MIPS ELF local symbols ($L123 instead of .L123) */
 120         else if (str[0] == '$')
 121                 return -1;
 122
 123         /* include the type field in the symbol name, so that it gets
 124          * compressed together */
 125         s->len = strlen(str) + 1;
 126         s->sym = malloc(s->len + 1);
 127         if (!s->sym) {
 128                 fprintf(stderr, "kallsyms failure: "
 129                         "unable to allocate required amount of memory\n");
 130                 exit(EXIT_FAILURE);
 131         }
 132         strcpy((char *)s->sym + 1, str);
 133         s->sym[0] = stype;
 134
 135         return 0;
 136 }
 137
 138 static int symbol_valid(struct sym_entry *s)
 139 {
 140         /* Symbols which vary between passes.  Passes 1 and 2 must have
 141          * identical symbol lists.  The kallsyms_* symbols below are only added
 142          * after pass 1, they would be included in pass 2 when --all-symbols is
 143          * specified so exclude them to get a stable symbol list.
 144          */
 145         static char *special_symbols[] = {
 146                 "kallsyms_addresses",
 147                 "kallsyms_num_syms",
 148                 "kallsyms_names",
 149                 "kallsyms_markers",
 150                 "kallsyms_token_table",
 151                 "kallsyms_token_index",
 152
 153         /* Exclude linker generated symbols which vary between passes */
 154                 "_SDA_BASE_",           /* ppc */
 155                 "_SDA2_BASE_",          /* ppc */
 156                 NULL };
 157         int i;
 158         int offset = 1;
 159
 160         /* skip prefix char */
 161         if (symbol_prefix_char && *(s->sym + 1) == symbol_prefix_char)
 162                 offset++;
 163
 164         /* if --all-symbols is not specified, then symbols outside the text
 165          * and inittext sections are discarded */
 166         if (!all_symbols) {
 167                 if ((s->addr < _stext || s->addr > _etext)
 168                     && (s->addr < _sinittext || s->addr > _einittext)
 169                     && (s->addr < _sextratext || s->addr > _eextratext))
 170                         return 0;
 171                 /* Corner case.  Discard any symbols with the same value as
 172                  * _etext _einittext or _eextratext; they can move between pass
 173                  * 1 and 2 when the kallsyms data are added.  If these symbols
 174                  * move then they may get dropped in pass 2, which breaks the
 175                  * kallsyms rules.
 176                  */
 177                 if ((s->addr == _etext && strcmp((char*)s->sym + offset, "_etext")) ||
 178                     (s->addr == _einittext && strcmp((char*)s->sym + offset, "_einittext")) ||
 179                     (s->addr == _eextratext && strcmp((char*)s->sym + offset, "_eextratext")))
 180                         return 0;
 181         }
 182
 183         /* Exclude symbols which vary between passes. */
 184         if (strstr((char *)s->sym + offset, "_compiled."))
 185                 return 0;
 186
 187         for (i = 0; special_symbols[i]; i++)
 188                 if( strcmp((char *)s->sym + offset, special_symbols[i]) == 0 )
 189                         return 0;
 190
 191         return 1;
 192 }
 193
 194 static void read_map(FILE *in)
 195 {
 196         while (!feof(in)) {
 197                 if (table_cnt >= table_size) {
 198                         table_size += 10000;
 199                         table = realloc(table, sizeof(*table) * table_size);
 200                         if (!table) {
 201                                 fprintf(stderr, "out of memory\n");
 202                                 exit (1);
 203                         }
 204                 }
 205                 if (read_symbol(in, &table[table_cnt]) == 0)
 206                         table_cnt++;
 207         }
 208 }
 209
 210 static void output_label(char *label)
 211 {
 212         if (symbol_prefix_char)
 213                 printf(".globl %c%s\n", symbol_prefix_char, label);
 214         else
 215                 printf(".globl %s\n", label);
 216         printf("\tALGN\n");
 217         if (symbol_prefix_char)
 218                 printf("%c%s:\n", symbol_prefix_char, label);
 219         else
 220                 printf("%s:\n", label);
 221 }
 222
 223 /* uncompress a compressed symbol. When this function is called, the best table
 224  * might still be compressed itself, so the function needs to be recursive */
 225 static int expand_symbol(unsigned char *data, int len, char *result)
 226 {
 227         int c, rlen, total=0;
 228
 229         while (len) {
 230                 c = *data;
 231                 /* if the table holds a single char that is the same as the one
 232                  * we are looking for, then end the search */
 233                 if (best_table[c][0]==c && best_table_len[c]==1) {
 234                         *result++ = c;
 235                         total++;
 236                 } else {
 237                         /* if not, recurse and expand */
 238                         rlen = expand_symbol(best_table[c], best_table_len[c], result);
 239                         total += rlen;
 240                         result += rlen;
 241                 }
 242                 data++;
 243                 len--;
 244         }
 245         *result=0;
 246
 247         return total;
 248 }
 249
 250 static void write_src(void)
 251 {
 252         unsigned int i, k, off;
 253         unsigned int best_idx[256];
 254         unsigned int *markers;
 255         char buf[KSYM_NAME_LEN];
 256
 257         printf("#include <asm/types.h>\n");
 258         printf("#if BITS_PER_LONG == 64\n");
 259         printf("#define PTR .quad\n");
 260         printf("#define ALGN .align 8\n");
 261         printf("#else\n");
 262         printf("#define PTR .long\n");
 263         printf("#define ALGN .align 4\n");
 264         printf("#endif\n");
 265
 266         printf("\t.section .rodata, \"a\"\n");
 267
 268         /* Provide proper symbols relocatability by their '_text'
 269          * relativeness.  The symbol names cannot be used to construct
 270          * normal symbol references as the list of symbols contains
 271          * symbols that are declared static and are private to their
 272          * .o files.  This prevents .tmp_kallsyms.o or any other
 273          * object from referencing them.
 274          */
 275         output_label("kallsyms_addresses");
 276         for (i = 0; i < table_cnt; i++) {
 277                 if (toupper(table[i].sym[0]) != 'A') {
 278                         if (_text <= table[i].addr)
 279                                 printf("\tPTR\t_text + %#llx\n",
 280                                         table[i].addr - _text);
 281                         else
 282                                 printf("\tPTR\t_text - %#llx\n",
 283                                         _text - table[i].addr);
 284                 } else {
 285                         printf("\tPTR\t%#llx\n", table[i].addr);
 286                 }
 287         }
 288         printf("\n");
 289
 290         output_label("kallsyms_num_syms");
 291         printf("\tPTR\t%d\n", table_cnt);
 292         printf("\n");
 293
 294         /* table of offset markers, that give the offset in the compressed stream
 295          * every 256 symbols */
 296         markers = malloc(sizeof(unsigned int) * ((table_cnt + 255) / 256));
 297         if (!markers) {
 298                 fprintf(stderr, "kallsyms failure: "
 299                         "unable to allocate required memory\n");
 300                 exit(EXIT_FAILURE);
 301         }
 302
 303         output_label("kallsyms_names");
 304         off = 0;
 305         for (i = 0; i < table_cnt; i++) {
 306                 if ((i & 0xFF) == 0)
 307                         markers[i >> 8] = off;
 308
 309                 printf("\t.byte 0x%02x", table[i].len);
 310                 for (k = 0; k < table[i].len; k++)
 311                         printf(", 0x%02x", table[i].sym[k]);
 312                 printf("\n");
 313
 314                 off += table[i].len + 1;
 315         }
 316         printf("\n");
 317
 318         output_label("kallsyms_markers");
 319         for (i = 0; i < ((table_cnt + 255) >> 8); i++)
 320                 printf("\tPTR\t%d\n", markers[i]);
 321         printf("\n");
 322
 323         free(markers);
 324
 325         output_label("kallsyms_token_table");
 326         off = 0;
 327         for (i = 0; i < 256; i++) {
 328                 best_idx[i] = off;
 329                 expand_symbol(best_table[i], best_table_len[i], buf);
 330                 printf("\t.asciz\t\"%s\"\n", buf);
 331                 off += strlen(buf) + 1;
 332         }
 333         printf("\n");
 334
 335         output_label("kallsyms_token_index");
 336         for (i = 0; i < 256; i++)
 337                 printf("\t.short\t%d\n", best_idx[i]);
 338         printf("\n");
 339 }
 340
 341
 342 /* table lookup compression functions */
 343
 344 /* count all the possible tokens in a symbol */
 345 static void learn_symbol(unsigned char *symbol, int len)
 346 {
 347         int i;
 348
 349         for (i = 0; i < len - 1; i++)
 350                 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]++;
 351 }
 352
 353 /* decrease the count for all the possible tokens in a symbol */
 354 static void forget_symbol(unsigned char *symbol, int len)
 355 {
 356         int i;
 357
 358         for (i = 0; i < len - 1; i++)
 359                 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]--;
 360 }
 361
 362 /* remove all the invalid symbols from the table and do the initial token count */
 363 static void build_initial_tok_table(void)
 364 {
 365         unsigned int i, pos;
 366
 367         pos = 0;
 368         for (i = 0; i < table_cnt; i++) {
 369                 if ( symbol_valid(&table[i]) ) {
 370                         if (pos != i)
 371                                 table[pos] = table[i];
 372                         learn_symbol(table[pos].sym, table[pos].len);
 373                         pos++;
 374                 }
 375         }
 376         table_cnt = pos;
 377 }
 378
 379 static void *find_token(unsigned char *str, int len, unsigned char *token)
 380 {
 381         int i;
 382
 383         for (i = 0; i < len - 1; i++) {
 384                 if (str[i] == token[0] && str[i+1] == token[1])
 385                         return &str[i];
 386         }
 387         return NULL;
 388 }
 389
 390 /* replace a given token in all the valid symbols. Use the sampled symbols
 391  * to update the counts */
 392 static void compress_symbols(unsigned char *str, int idx)
 393 {
 394         unsigned int i, len, size;
 395         unsigned char *p1, *p2;
 396
 397         for (i = 0; i < table_cnt; i++) {
 398
 399                 len = table[i].len;
 400                 p1 = table[i].sym;
 401
 402                 /* find the token on the symbol */
 403                 p2 = find_token(p1, len, str);
 404                 if (!p2) continue;
 405
 406                 /* decrease the counts for this symbol's tokens */
 407                 forget_symbol(table[i].sym, len);
 408
 409                 size = len;
 410
 411                 do {
 412                         *p2 = idx;
 413                         p2++;
 414                         size -= (p2 - p1);
 415                         memmove(p2, p2 + 1, size);
 416                         p1 = p2;
 417                         len--;
 418
 419                         if (size < 2) break;
 420
 421                         /* find the token on the symbol */
 422                         p2 = find_token(p1, size, str);
 423
 424                 } while (p2);
 425
 426                 table[i].len = len;
 427
 428                 /* increase the counts for this symbol's new tokens */
 429                 learn_symbol(table[i].sym, len);
 430         }
 431 }
 432
 433 /* search the token with the maximum profit */
 434 static int find_best_token(void)
 435 {
 436         int i, best, bestprofit;
 437
 438         bestprofit=-10000;
 439         best = 0;
 440
 441         for (i = 0; i < 0x10000; i++) {
 442                 if (token_profit[i] > bestprofit) {
 443                         best = i;
 444                         bestprofit = token_profit[i];
 445                 }
 446         }
 447         return best;
 448 }
 449
 450 /* this is the core of the algorithm: calculate the "best" table */
 451 static void optimize_result(void)
 452 {
 453         int i, best;
 454
 455         /* using the '\0' symbol last allows compress_symbols to use standard
 456          * fast string functions */
 457         for (i = 255; i >= 0; i--) {
 458
 459                 /* if this table slot is empty (it is not used by an actual
 460                  * original char code */
 461                 if (!best_table_len[i]) {
 462
 463                         /* find the token with the breates profit value */
 464                         best = find_best_token();
 465
 466                         /* place it in the "best" table */
 467                         best_table_len[i] = 2;
 468                         best_table[i][0] = best & 0xFF;
 469                         best_table[i][1] = (best >> 8) & 0xFF;
 470
 471                         /* replace this token in all the valid symbols */
 472                         compress_symbols(best_table[i], i);
 473                 }
 474         }
 475 }
 476
 477 /* start by placing the symbols that are actually used on the table */
 478 static void insert_real_symbols_in_table(void)
 479 {
 480         unsigned int i, j, c;
 481
 482         memset(best_table, 0, sizeof(best_table));
 483         memset(best_table_len, 0, sizeof(best_table_len));
 484
 485         for (i = 0; i < table_cnt; i++) {
 486                 for (j = 0; j < table[i].len; j++) {
 487                         c = table[i].sym[j];
 488                         best_table[c][0]=c;
 489                         best_table_len[c]=1;
 490                 }
 491         }
 492 }
 493
 494 static void optimize_token_table(void)
 495 {
 496         build_initial_tok_table();
 497
 498         insert_real_symbols_in_table();
 499
 500         /* When valid symbol is not registered, exit to error */
 501         if (!table_cnt) {
 502                 fprintf(stderr, "No valid symbol.\n");
 503                 exit(1);
 504         }
 505
 506         optimize_result();
 507 }
 508
 509
 510 int main(int argc, char **argv)
 511 {
 512         if (argc >= 2) {
 513                 int i;
 514                 for (i = 1; i < argc; i++) {
 515                         if(strcmp(argv[i], "--all-symbols") == 0)
 516                                 all_symbols = 1;
 517                         else if (strncmp(argv[i], "--symbol-prefix=", 16) == 0) {
 518                                 char *p = &argv[i][16];
 519                                 /* skip quote */
 520                                 if ((*p == '"' && *(p+2) == '"') || (*p == '\'' && *(p+2) == '\''))
 521                                         p++;
 522                                 symbol_prefix_char = *p;
 523                         } else
 524                                 usage();
 525                 }
 526         } else if (argc != 1)
 527                 usage();
 528
 529         read_map(stdin);
 530         optimize_token_table();
 531         write_src();
 532
 533         return 0;
 534 }