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 #define _GNU_SOURCE
  28
  29 #include <stdio.h>
  30 #include <stdlib.h>
  31 #include <string.h>
  32 #include <ctype.h>
  33
  34 #define KSYM_NAME_LEN           127
  35
  36
  37 struct sym_entry {
  38         unsigned long long addr;
  39         unsigned int len;
  40         unsigned char *sym;
  41 };
  42
  43
  44 static struct sym_entry *table;
  45 static unsigned int table_size, table_cnt;
  46 static unsigned long long _stext, _etext, _sinittext, _einittext, _sextratext, _eextratext;
  47 static int all_symbols = 0;
  48 static char symbol_prefix_char = '\0';
  49
  50 int token_profit[0x10000];
  51
  52 /* the table that holds the result of the compression */
  53 unsigned char best_table[256][2];
  54 unsigned char best_table_len[256];
  55
  56
  57 static void usage(void)
  58 {
  59         fprintf(stderr, "Usage: kallsyms [--all-symbols] [--symbol-prefix=<prefix char>] < in.map > out.S\n");
  60         exit(1);
  61 }
  62
  63 /*
  64  * This ignores the intensely annoying "mapping symbols" found
  65  * in ARM ELF files: $a, $t and $d.
  66  */
  67 static inline int is_arm_mapping_symbol(const char *str)
  68 {
  69         return str[0] == '$' && strchr("atd", str[1])
  70                && (str[2] == '\0' || str[2] == '.');
  71 }
  72
  73 static int read_symbol(FILE *in, struct sym_entry *s)
  74 {
  75         char str[500];
  76         char *sym, stype;
  77         int rc;
  78
  79         rc = fscanf(in, "%llx %c %499s\n", &s->addr, &stype, str);
  80         if (rc != 3) {
  81                 if (rc != EOF) {
  82                         /* skip line */
  83                         fgets(str, 500, in);
  84                 }
  85                 return -1;
  86         }
  87
  88         sym = str;
  89         /* skip prefix char */
  90         if (symbol_prefix_char && str[0] == symbol_prefix_char)
  91                 sym++;
  92
  93         /* Ignore most absolute/undefined (?) symbols. */
  94         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+1];
 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(".data\n");
 267
 268         output_label("kallsyms_addresses");
 269         for (i = 0; i < table_cnt; i++) {
 270                 printf("\tPTR\t%#llx\n", table[i].addr);
 271         }
 272         printf("\n");
 273
 274         output_label("kallsyms_num_syms");
 275         printf("\tPTR\t%d\n", table_cnt);
 276         printf("\n");
 277
 278         /* table of offset markers, that give the offset in the compressed stream
 279          * every 256 symbols */
 280         markers = malloc(sizeof(unsigned int) * ((table_cnt + 255) / 256));
 281         if (!markers) {
 282                 fprintf(stderr, "kallsyms failure: "
 283                         "unable to allocate required memory\n");
 284                 exit(EXIT_FAILURE);
 285         }
 286
 287         output_label("kallsyms_names");
 288         off = 0;
 289         for (i = 0; i < table_cnt; i++) {
 290                 if ((i & 0xFF) == 0)
 291                         markers[i >> 8] = off;
 292
 293                 printf("\t.byte 0x%02x", table[i].len);
 294                 for (k = 0; k < table[i].len; k++)
 295                         printf(", 0x%02x", table[i].sym[k]);
 296                 printf("\n");
 297
 298                 off += table[i].len + 1;
 299         }
 300         printf("\n");
 301
 302         output_label("kallsyms_markers");
 303         for (i = 0; i < ((table_cnt + 255) >> 8); i++)
 304                 printf("\tPTR\t%d\n", markers[i]);
 305         printf("\n");
 306
 307         free(markers);
 308
 309         output_label("kallsyms_token_table");
 310         off = 0;
 311         for (i = 0; i < 256; i++) {
 312                 best_idx[i] = off;
 313                 expand_symbol(best_table[i], best_table_len[i], buf);
 314                 printf("\t.asciz\t\"%s\"\n", buf);
 315                 off += strlen(buf) + 1;
 316         }
 317         printf("\n");
 318
 319         output_label("kallsyms_token_index");
 320         for (i = 0; i < 256; i++)
 321                 printf("\t.short\t%d\n", best_idx[i]);
 322         printf("\n");
 323 }
 324
 325
 326 /* table lookup compression functions */
 327
 328 /* count all the possible tokens in a symbol */
 329 static void learn_symbol(unsigned char *symbol, int len)
 330 {
 331         int i;
 332
 333         for (i = 0; i < len - 1; i++)
 334                 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]++;
 335 }
 336
 337 /* decrease the count for all the possible tokens in a symbol */
 338 static void forget_symbol(unsigned char *symbol, int len)
 339 {
 340         int i;
 341
 342         for (i = 0; i < len - 1; i++)
 343                 token_profit[ symbol[i] + (symbol[i + 1] << 8) ]--;
 344 }
 345
 346 /* remove all the invalid symbols from the table and do the initial token count */
 347 static void build_initial_tok_table(void)
 348 {
 349         unsigned int i, pos;
 350
 351         pos = 0;
 352         for (i = 0; i < table_cnt; i++) {
 353                 if ( symbol_valid(&table[i]) ) {
 354                         if (pos != i)
 355                                 table[pos] = table[i];
 356                         learn_symbol(table[pos].sym, table[pos].len);
 357                         pos++;
 358                 }
 359         }
 360         table_cnt = pos;
 361 }
 362
 363 /* replace a given token in all the valid symbols. Use the sampled symbols
 364  * to update the counts */
 365 static void compress_symbols(unsigned char *str, int idx)
 366 {
 367         unsigned int i, len, size;
 368         unsigned char *p1, *p2;
 369
 370         for (i = 0; i < table_cnt; i++) {
 371
 372                 len = table[i].len;
 373                 p1 = table[i].sym;
 374
 375                 /* find the token on the symbol */
 376                 p2 = memmem(p1, len, str, 2);
 377                 if (!p2) continue;
 378
 379                 /* decrease the counts for this symbol's tokens */
 380                 forget_symbol(table[i].sym, len);
 381
 382                 size = len;
 383
 384                 do {
 385                         *p2 = idx;
 386                         p2++;
 387                         size -= (p2 - p1);
 388                         memmove(p2, p2 + 1, size);
 389                         p1 = p2;
 390                         len--;
 391
 392                         if (size < 2) break;
 393
 394                         /* find the token on the symbol */
 395                         p2 = memmem(p1, size, str, 2);
 396
 397                 } while (p2);
 398
 399                 table[i].len = len;
 400
 401                 /* increase the counts for this symbol's new tokens */
 402                 learn_symbol(table[i].sym, len);
 403         }
 404 }
 405
 406 /* search the token with the maximum profit */
 407 static int find_best_token(void)
 408 {
 409         int i, best, bestprofit;
 410
 411         bestprofit=-10000;
 412         best = 0;
 413
 414         for (i = 0; i < 0x10000; i++) {
 415                 if (token_profit[i] > bestprofit) {
 416                         best = i;
 417                         bestprofit = token_profit[i];
 418                 }
 419         }
 420         return best;
 421 }
 422
 423 /* this is the core of the algorithm: calculate the "best" table */
 424 static void optimize_result(void)
 425 {
 426         int i, best;
 427
 428         /* using the '\0' symbol last allows compress_symbols to use standard
 429          * fast string functions */
 430         for (i = 255; i >= 0; i--) {
 431
 432                 /* if this table slot is empty (it is not used by an actual
 433                  * original char code */
 434                 if (!best_table_len[i]) {
 435
 436                         /* find the token with the breates profit value */
 437                         best = find_best_token();
 438
 439                         /* place it in the "best" table */
 440                         best_table_len[i] = 2;
 441                         best_table[i][0] = best & 0xFF;
 442                         best_table[i][1] = (best >> 8) & 0xFF;
 443
 444                         /* replace this token in all the valid symbols */
 445                         compress_symbols(best_table[i], i);
 446                 }
 447         }
 448 }
 449
 450 /* start by placing the symbols that are actually used on the table */
 451 static void insert_real_symbols_in_table(void)
 452 {
 453         unsigned int i, j, c;
 454
 455         memset(best_table, 0, sizeof(best_table));
 456         memset(best_table_len, 0, sizeof(best_table_len));
 457
 458         for (i = 0; i < table_cnt; i++) {
 459                 for (j = 0; j < table[i].len; j++) {
 460                         c = table[i].sym[j];
 461                         best_table[c][0]=c;
 462                         best_table_len[c]=1;
 463                 }
 464         }
 465 }
 466
 467 static void optimize_token_table(void)
 468 {
 469         build_initial_tok_table();
 470
 471         insert_real_symbols_in_table();
 472
 473         /* When valid symbol is not registered, exit to error */
 474         if (!table_cnt) {
 475                 fprintf(stderr, "No valid symbol.\n");
 476                 exit(1);
 477         }
 478
 479         optimize_result();
 480 }
 481
 482
 483 int main(int argc, char **argv)
 484 {
 485         if (argc >= 2) {
 486                 int i;
 487                 for (i = 1; i < argc; i++) {
 488                         if(strcmp(argv[i], "--all-symbols") == 0)
 489                                 all_symbols = 1;
 490                         else if (strncmp(argv[i], "--symbol-prefix=", 16) == 0) {
 491                                 char *p = &argv[i][16];
 492                                 /* skip quote */
 493                                 if ((*p == '"' && *(p+2) == '"') || (*p == '\'' && *(p+2) == '\''))
 494                                         p++;
 495                                 symbol_prefix_char = *p;
 496                         } else
 497                                 usage();
 498                 }
 499         } else if (argc != 1)
 500                 usage();
 501
 502         read_map(stdin);
 503         optimize_token_table();
 504         write_src();
 505
 506         return 0;
 507 }