Commit | Line | Data |
---|---|---|
1da177e4 LT |
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 | * | |
1da177e4 LT |
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 | |
17 | * of about 50%. | |
18 | * | |
19 | */ | |
20 | ||
a41333e0 | 21 | #include <stdbool.h> |
1da177e4 LT |
22 | #include <stdio.h> |
23 | #include <stdlib.h> | |
24 | #include <string.h> | |
25 | #include <ctype.h> | |
2213e9a6 | 26 | #include <limits.h> |
1da177e4 | 27 | |
17b1f0de | 28 | #define ARRAY_SIZE(arr) (sizeof(arr) / sizeof(arr[0])) |
17b1f0de | 29 | |
b471927e BF |
30 | #define _stringify_1(x) #x |
31 | #define _stringify(x) _stringify_1(x) | |
32 | ||
9281acea | 33 | #define KSYM_NAME_LEN 128 |
1da177e4 | 34 | |
6e8c5bbd MO |
35 | /* |
36 | * A substantially bigger size than the current maximum. | |
37 | * | |
38 | * It cannot be defined as an expression because it gets stringified | |
39 | * for the fscanf() format string. Therefore, a _Static_assert() is | |
40 | * used instead to maintain the relationship with KSYM_NAME_LEN. | |
41 | */ | |
42 | #define KSYM_NAME_LEN_BUFFER 512 | |
43 | _Static_assert( | |
44 | KSYM_NAME_LEN_BUFFER == KSYM_NAME_LEN * 4, | |
45 | "Please keep KSYM_NAME_LEN_BUFFER in sync with KSYM_NAME_LEN" | |
46 | ); | |
b471927e | 47 | |
1da177e4 LT |
48 | struct sym_entry { |
49 | unsigned long long addr; | |
b3dbb4ec | 50 | unsigned int len; |
f2df3f65 | 51 | unsigned int start_pos; |
8c996940 | 52 | unsigned int percpu_absolute; |
9d82973e | 53 | unsigned char sym[]; |
1da177e4 LT |
54 | }; |
55 | ||
78eb7159 KC |
56 | struct addr_range { |
57 | const char *start_sym, *end_sym; | |
17b1f0de MF |
58 | unsigned long long start, end; |
59 | }; | |
60 | ||
61 | static unsigned long long _text; | |
2213e9a6 | 62 | static unsigned long long relative_base; |
78eb7159 | 63 | static struct addr_range text_ranges[] = { |
17b1f0de MF |
64 | { "_stext", "_etext" }, |
65 | { "_sinittext", "_einittext" }, | |
17b1f0de MF |
66 | }; |
67 | #define text_range_text (&text_ranges[0]) | |
68 | #define text_range_inittext (&text_ranges[1]) | |
69 | ||
c6bda7c9 RR |
70 | static struct addr_range percpu_range = { |
71 | "__per_cpu_start", "__per_cpu_end", -1ULL, 0 | |
72 | }; | |
73 | ||
8d605269 | 74 | static struct sym_entry **table; |
b3dbb4ec | 75 | static unsigned int table_size, table_cnt; |
831362fc MY |
76 | static int all_symbols; |
77 | static int absolute_percpu; | |
78 | static int base_relative; | |
1da177e4 | 79 | |
f43e9daa | 80 | static int token_profit[0x10000]; |
1da177e4 LT |
81 | |
82 | /* the table that holds the result of the compression */ | |
f43e9daa MY |
83 | static unsigned char best_table[256][2]; |
84 | static unsigned char best_table_len[256]; | |
1da177e4 LT |
85 | |
86 | ||
b3dbb4ec | 87 | static void usage(void) |
1da177e4 | 88 | { |
8d3a7507 | 89 | fprintf(stderr, "Usage: kallsyms [--all-symbols] [--absolute-percpu] " |
2213e9a6 | 90 | "[--base-relative] < in.map > out.S\n"); |
1da177e4 LT |
91 | exit(1); |
92 | } | |
93 | ||
29e55ad3 MY |
94 | static char *sym_name(const struct sym_entry *s) |
95 | { | |
96 | return (char *)s->sym + 1; | |
97 | } | |
98 | ||
a41333e0 MY |
99 | static bool is_ignored_symbol(const char *name, char type) |
100 | { | |
516d980f | 101 | /* Symbol names that exactly match to the following are ignored.*/ |
a41333e0 MY |
102 | static const char * const ignored_symbols[] = { |
103 | /* | |
104 | * Symbols which vary between passes. Passes 1 and 2 must have | |
105 | * identical symbol lists. The kallsyms_* symbols below are | |
106 | * only added after pass 1, they would be included in pass 2 | |
107 | * when --all-symbols is specified so exclude them to get a | |
108 | * stable symbol list. | |
109 | */ | |
110 | "kallsyms_addresses", | |
111 | "kallsyms_offsets", | |
112 | "kallsyms_relative_base", | |
113 | "kallsyms_num_syms", | |
114 | "kallsyms_names", | |
115 | "kallsyms_markers", | |
116 | "kallsyms_token_table", | |
117 | "kallsyms_token_index", | |
118 | /* Exclude linker generated symbols which vary between passes */ | |
119 | "_SDA_BASE_", /* ppc */ | |
120 | "_SDA2_BASE_", /* ppc */ | |
121 | NULL | |
122 | }; | |
123 | ||
516d980f | 124 | /* Symbol names that begin with the following are ignored.*/ |
a41333e0 | 125 | static const char * const ignored_prefixes[] = { |
97261e1e | 126 | "$", /* local symbols for ARM, MIPS, etc. */ |
d4c85864 | 127 | ".L", /* local labels, .LBB,.Ltmpxxx,.L__unnamed_xx,.LASANPC, etc. */ |
a41333e0 MY |
128 | "__crc_", /* modversions */ |
129 | "__efistub_", /* arm64 EFI stub namespace */ | |
6ccf9cb5 KS |
130 | "__kvm_nvhe_$", /* arm64 local symbols in non-VHE KVM namespace */ |
131 | "__kvm_nvhe_.L", /* arm64 local symbols in non-VHE KVM namespace */ | |
efe6e306 AB |
132 | "__AArch64ADRPThunk_", /* arm64 lld */ |
133 | "__ARMV5PILongThunk_", /* arm lld */ | |
134 | "__ARMV7PILongThunk_", | |
135 | "__ThumbV7PILongThunk_", | |
136 | "__LA25Thunk_", /* mips lld */ | |
137 | "__microLA25Thunk_", | |
a41333e0 MY |
138 | NULL |
139 | }; | |
140 | ||
516d980f | 141 | /* Symbol names that end with the following are ignored.*/ |
a41333e0 MY |
142 | static const char * const ignored_suffixes[] = { |
143 | "_from_arm", /* arm */ | |
144 | "_from_thumb", /* arm */ | |
145 | "_veneer", /* arm */ | |
146 | NULL | |
147 | }; | |
148 | ||
516d980f MY |
149 | /* Symbol names that contain the following are ignored.*/ |
150 | static const char * const ignored_matches[] = { | |
151 | ".long_branch.", /* ppc stub */ | |
152 | ".plt_branch.", /* ppc stub */ | |
153 | NULL | |
154 | }; | |
155 | ||
a41333e0 MY |
156 | const char * const *p; |
157 | ||
a41333e0 MY |
158 | for (p = ignored_symbols; *p; p++) |
159 | if (!strcmp(name, *p)) | |
160 | return true; | |
161 | ||
162 | for (p = ignored_prefixes; *p; p++) | |
163 | if (!strncmp(name, *p, strlen(*p))) | |
164 | return true; | |
165 | ||
166 | for (p = ignored_suffixes; *p; p++) { | |
167 | int l = strlen(name) - strlen(*p); | |
168 | ||
169 | if (l >= 0 && !strcmp(name + l, *p)) | |
170 | return true; | |
171 | } | |
172 | ||
516d980f MY |
173 | for (p = ignored_matches; *p; p++) { |
174 | if (strstr(name, *p)) | |
175 | return true; | |
176 | } | |
177 | ||
887df76d MY |
178 | if (type == 'U' || type == 'u') |
179 | return true; | |
180 | /* exclude debugging symbols */ | |
181 | if (type == 'N' || type == 'n') | |
182 | return true; | |
183 | ||
184 | if (toupper(type) == 'A') { | |
185 | /* Keep these useful absolute symbols */ | |
186 | if (strcmp(name, "__kernel_syscall_via_break") && | |
187 | strcmp(name, "__kernel_syscall_via_epc") && | |
188 | strcmp(name, "__kernel_sigtramp") && | |
189 | strcmp(name, "__gp")) | |
190 | return true; | |
191 | } | |
192 | ||
a41333e0 MY |
193 | return false; |
194 | } | |
195 | ||
b6233d0d MY |
196 | static void check_symbol_range(const char *sym, unsigned long long addr, |
197 | struct addr_range *ranges, int entries) | |
17b1f0de MF |
198 | { |
199 | size_t i; | |
78eb7159 | 200 | struct addr_range *ar; |
17b1f0de | 201 | |
78eb7159 KC |
202 | for (i = 0; i < entries; ++i) { |
203 | ar = &ranges[i]; | |
17b1f0de | 204 | |
78eb7159 KC |
205 | if (strcmp(sym, ar->start_sym) == 0) { |
206 | ar->start = addr; | |
b6233d0d | 207 | return; |
78eb7159 KC |
208 | } else if (strcmp(sym, ar->end_sym) == 0) { |
209 | ar->end = addr; | |
b6233d0d | 210 | return; |
17b1f0de MF |
211 | } |
212 | } | |
17b1f0de MF |
213 | } |
214 | ||
8d605269 | 215 | static struct sym_entry *read_symbol(FILE *in) |
1da177e4 | 216 | { |
b471927e | 217 | char name[KSYM_NAME_LEN_BUFFER+1], type; |
8d605269 MY |
218 | unsigned long long addr; |
219 | unsigned int len; | |
220 | struct sym_entry *sym; | |
1da177e4 LT |
221 | int rc; |
222 | ||
b471927e | 223 | rc = fscanf(in, "%llx %c %" _stringify(KSYM_NAME_LEN_BUFFER) "s\n", &addr, &type, name); |
1da177e4 | 224 | if (rc != 3) { |
b66c874f | 225 | if (rc != EOF && fgets(name, ARRAY_SIZE(name), in) == NULL) |
ef894870 | 226 | fprintf(stderr, "Read error or end of file.\n"); |
8d605269 | 227 | return NULL; |
1da177e4 | 228 | } |
be9f6133 | 229 | if (strlen(name) >= KSYM_NAME_LEN) { |
6db2983c | 230 | fprintf(stderr, "Symbol %s too long for kallsyms (%zu >= %d).\n" |
bb66fc67 | 231 | "Please increase KSYM_NAME_LEN both in kernel and kallsyms.c\n", |
be9f6133 | 232 | name, strlen(name), KSYM_NAME_LEN); |
8d605269 | 233 | return NULL; |
f3462aa9 | 234 | } |
1da177e4 | 235 | |
be9f6133 | 236 | if (strcmp(name, "_text") == 0) |
8d605269 | 237 | _text = addr; |
b6233d0d | 238 | |
7883a143 MP |
239 | /* Ignore most absolute/undefined (?) symbols. */ |
240 | if (is_ignored_symbol(name, type)) | |
241 | return NULL; | |
242 | ||
8d605269 MY |
243 | check_symbol_range(name, addr, text_ranges, ARRAY_SIZE(text_ranges)); |
244 | check_symbol_range(name, addr, &percpu_range, 1); | |
1da177e4 LT |
245 | |
246 | /* include the type field in the symbol name, so that it gets | |
247 | * compressed together */ | |
8d605269 MY |
248 | |
249 | len = strlen(name) + 1; | |
250 | ||
9d1b3895 | 251 | sym = malloc(sizeof(*sym) + len + 1); |
8d605269 | 252 | if (!sym) { |
f1a136e0 JJ |
253 | fprintf(stderr, "kallsyms failure: " |
254 | "unable to allocate required amount of memory\n"); | |
255 | exit(EXIT_FAILURE); | |
256 | } | |
8d605269 MY |
257 | sym->addr = addr; |
258 | sym->len = len; | |
259 | sym->sym[0] = type; | |
9d1b3895 | 260 | strcpy(sym_name(sym), name); |
8d605269 | 261 | sym->percpu_absolute = 0; |
8c996940 | 262 | |
8d605269 | 263 | return sym; |
1da177e4 LT |
264 | } |
265 | ||
4bfe2b78 MY |
266 | static int symbol_in_range(const struct sym_entry *s, |
267 | const struct addr_range *ranges, int entries) | |
17b1f0de MF |
268 | { |
269 | size_t i; | |
4bfe2b78 | 270 | const struct addr_range *ar; |
17b1f0de | 271 | |
78eb7159 KC |
272 | for (i = 0; i < entries; ++i) { |
273 | ar = &ranges[i]; | |
17b1f0de | 274 | |
78eb7159 | 275 | if (s->addr >= ar->start && s->addr <= ar->end) |
ac6ca5c8 | 276 | return 1; |
17b1f0de MF |
277 | } |
278 | ||
ac6ca5c8 | 279 | return 0; |
17b1f0de MF |
280 | } |
281 | ||
4bfe2b78 | 282 | static int symbol_valid(const struct sym_entry *s) |
1da177e4 | 283 | { |
29e55ad3 | 284 | const char *name = sym_name(s); |
bd8b22d2 | 285 | |
1da177e4 LT |
286 | /* if --all-symbols is not specified, then symbols outside the text |
287 | * and inittext sections are discarded */ | |
288 | if (!all_symbols) { | |
78eb7159 KC |
289 | if (symbol_in_range(s, text_ranges, |
290 | ARRAY_SIZE(text_ranges)) == 0) | |
1da177e4 LT |
291 | return 0; |
292 | /* Corner case. Discard any symbols with the same value as | |
a3b81113 RG |
293 | * _etext _einittext; they can move between pass 1 and 2 when |
294 | * the kallsyms data are added. If these symbols move then | |
295 | * they may get dropped in pass 2, which breaks the kallsyms | |
296 | * rules. | |
1da177e4 | 297 | */ |
17b1f0de | 298 | if ((s->addr == text_range_text->end && |
29e55ad3 | 299 | strcmp(name, text_range_text->end_sym)) || |
17b1f0de | 300 | (s->addr == text_range_inittext->end && |
29e55ad3 | 301 | strcmp(name, text_range_inittext->end_sym))) |
1da177e4 LT |
302 | return 0; |
303 | } | |
304 | ||
1da177e4 LT |
305 | return 1; |
306 | } | |
307 | ||
5e5c4fa7 MY |
308 | /* remove all the invalid symbols from the table */ |
309 | static void shrink_table(void) | |
310 | { | |
311 | unsigned int i, pos; | |
312 | ||
313 | pos = 0; | |
314 | for (i = 0; i < table_cnt; i++) { | |
8d605269 | 315 | if (symbol_valid(table[i])) { |
5e5c4fa7 MY |
316 | if (pos != i) |
317 | table[pos] = table[i]; | |
318 | pos++; | |
319 | } else { | |
8d605269 | 320 | free(table[i]); |
5e5c4fa7 MY |
321 | } |
322 | } | |
323 | table_cnt = pos; | |
324 | ||
325 | /* When valid symbol is not registered, exit to error */ | |
326 | if (!table_cnt) { | |
327 | fprintf(stderr, "No valid symbol.\n"); | |
328 | exit(1); | |
329 | } | |
330 | } | |
331 | ||
b3dbb4ec | 332 | static void read_map(FILE *in) |
1da177e4 | 333 | { |
8d605269 MY |
334 | struct sym_entry *sym; |
335 | ||
1da177e4 | 336 | while (!feof(in)) { |
8d605269 MY |
337 | sym = read_symbol(in); |
338 | if (!sym) | |
339 | continue; | |
340 | ||
341 | sym->start_pos = table_cnt; | |
342 | ||
b3dbb4ec PM |
343 | if (table_cnt >= table_size) { |
344 | table_size += 10000; | |
345 | table = realloc(table, sizeof(*table) * table_size); | |
1da177e4 LT |
346 | if (!table) { |
347 | fprintf(stderr, "out of memory\n"); | |
348 | exit (1); | |
349 | } | |
350 | } | |
8d605269 MY |
351 | |
352 | table[table_cnt++] = sym; | |
1da177e4 LT |
353 | } |
354 | } | |
355 | ||
4bfe2b78 | 356 | static void output_label(const char *label) |
1da177e4 | 357 | { |
534c9f2e | 358 | printf(".globl %s\n", label); |
1da177e4 | 359 | printf("\tALGN\n"); |
534c9f2e | 360 | printf("%s:\n", label); |
1da177e4 LT |
361 | } |
362 | ||
fd2ab2f6 MY |
363 | /* Provide proper symbols relocatability by their '_text' relativeness. */ |
364 | static void output_address(unsigned long long addr) | |
365 | { | |
366 | if (_text <= addr) | |
367 | printf("\tPTR\t_text + %#llx\n", addr - _text); | |
368 | else | |
369 | printf("\tPTR\t_text - %#llx\n", _text - addr); | |
370 | } | |
371 | ||
1da177e4 LT |
372 | /* uncompress a compressed symbol. When this function is called, the best table |
373 | * might still be compressed itself, so the function needs to be recursive */ | |
4bfe2b78 | 374 | static int expand_symbol(const unsigned char *data, int len, char *result) |
1da177e4 LT |
375 | { |
376 | int c, rlen, total=0; | |
377 | ||
378 | while (len) { | |
379 | c = *data; | |
380 | /* if the table holds a single char that is the same as the one | |
381 | * we are looking for, then end the search */ | |
382 | if (best_table[c][0]==c && best_table_len[c]==1) { | |
383 | *result++ = c; | |
384 | total++; | |
385 | } else { | |
386 | /* if not, recurse and expand */ | |
387 | rlen = expand_symbol(best_table[c], best_table_len[c], result); | |
388 | total += rlen; | |
389 | result += rlen; | |
390 | } | |
391 | data++; | |
392 | len--; | |
393 | } | |
394 | *result=0; | |
395 | ||
396 | return total; | |
397 | } | |
398 | ||
4bfe2b78 | 399 | static int symbol_absolute(const struct sym_entry *s) |
78eb7159 | 400 | { |
8c996940 | 401 | return s->percpu_absolute; |
78eb7159 KC |
402 | } |
403 | ||
b3dbb4ec | 404 | static void write_src(void) |
1da177e4 | 405 | { |
b3dbb4ec | 406 | unsigned int i, k, off; |
1da177e4 LT |
407 | unsigned int best_idx[256]; |
408 | unsigned int *markers; | |
9281acea | 409 | char buf[KSYM_NAME_LEN]; |
1da177e4 | 410 | |
500193ec | 411 | printf("#include <asm/bitsperlong.h>\n"); |
1da177e4 LT |
412 | printf("#if BITS_PER_LONG == 64\n"); |
413 | printf("#define PTR .quad\n"); | |
72d3ebb9 | 414 | printf("#define ALGN .balign 8\n"); |
1da177e4 LT |
415 | printf("#else\n"); |
416 | printf("#define PTR .long\n"); | |
72d3ebb9 | 417 | printf("#define ALGN .balign 4\n"); |
1da177e4 LT |
418 | printf("#endif\n"); |
419 | ||
aad09470 | 420 | printf("\t.section .rodata, \"a\"\n"); |
1da177e4 | 421 | |
2213e9a6 AB |
422 | if (!base_relative) |
423 | output_label("kallsyms_addresses"); | |
424 | else | |
425 | output_label("kallsyms_offsets"); | |
426 | ||
b3dbb4ec | 427 | for (i = 0; i < table_cnt; i++) { |
2213e9a6 | 428 | if (base_relative) { |
fd2ab2f6 MY |
429 | /* |
430 | * Use the offset relative to the lowest value | |
431 | * encountered of all relative symbols, and emit | |
432 | * non-relocatable fixed offsets that will be fixed | |
433 | * up at runtime. | |
434 | */ | |
435 | ||
2213e9a6 AB |
436 | long long offset; |
437 | int overflow; | |
438 | ||
439 | if (!absolute_percpu) { | |
8d605269 | 440 | offset = table[i]->addr - relative_base; |
2213e9a6 | 441 | overflow = (offset < 0 || offset > UINT_MAX); |
8d605269 MY |
442 | } else if (symbol_absolute(table[i])) { |
443 | offset = table[i]->addr; | |
2213e9a6 AB |
444 | overflow = (offset < 0 || offset > INT_MAX); |
445 | } else { | |
8d605269 | 446 | offset = relative_base - table[i]->addr - 1; |
2213e9a6 AB |
447 | overflow = (offset < INT_MIN || offset >= 0); |
448 | } | |
449 | if (overflow) { | |
450 | fprintf(stderr, "kallsyms failure: " | |
451 | "%s symbol value %#llx out of range in relative mode\n", | |
8d605269 MY |
452 | symbol_absolute(table[i]) ? "absolute" : "relative", |
453 | table[i]->addr); | |
2213e9a6 AB |
454 | exit(EXIT_FAILURE); |
455 | } | |
456 | printf("\t.long\t%#x\n", (int)offset); | |
8d605269 MY |
457 | } else if (!symbol_absolute(table[i])) { |
458 | output_address(table[i]->addr); | |
fd593d12 | 459 | } else { |
8d605269 | 460 | printf("\tPTR\t%#llx\n", table[i]->addr); |
fd593d12 | 461 | } |
1da177e4 LT |
462 | } |
463 | printf("\n"); | |
464 | ||
2213e9a6 AB |
465 | if (base_relative) { |
466 | output_label("kallsyms_relative_base"); | |
fd2ab2f6 | 467 | output_address(relative_base); |
2213e9a6 AB |
468 | printf("\n"); |
469 | } | |
470 | ||
1da177e4 | 471 | output_label("kallsyms_num_syms"); |
80ffbaa5 | 472 | printf("\t.long\t%u\n", table_cnt); |
1da177e4 LT |
473 | printf("\n"); |
474 | ||
475 | /* table of offset markers, that give the offset in the compressed stream | |
476 | * every 256 symbols */ | |
f1a136e0 JJ |
477 | markers = malloc(sizeof(unsigned int) * ((table_cnt + 255) / 256)); |
478 | if (!markers) { | |
479 | fprintf(stderr, "kallsyms failure: " | |
480 | "unable to allocate required memory\n"); | |
481 | exit(EXIT_FAILURE); | |
482 | } | |
1da177e4 LT |
483 | |
484 | output_label("kallsyms_names"); | |
1da177e4 | 485 | off = 0; |
b3dbb4ec PM |
486 | for (i = 0; i < table_cnt; i++) { |
487 | if ((i & 0xFF) == 0) | |
488 | markers[i >> 8] = off; | |
1da177e4 | 489 | |
73bbb944 MO |
490 | /* There cannot be any symbol of length zero. */ |
491 | if (table[i]->len == 0) { | |
492 | fprintf(stderr, "kallsyms failure: " | |
493 | "unexpected zero symbol length\n"); | |
494 | exit(EXIT_FAILURE); | |
495 | } | |
496 | ||
497 | /* Only lengths that fit in up-to-two-byte ULEB128 are supported. */ | |
498 | if (table[i]->len > 0x3FFF) { | |
499 | fprintf(stderr, "kallsyms failure: " | |
500 | "unexpected huge symbol length\n"); | |
501 | exit(EXIT_FAILURE); | |
502 | } | |
503 | ||
504 | /* Encode length with ULEB128. */ | |
505 | if (table[i]->len <= 0x7F) { | |
506 | /* Most symbols use a single byte for the length. */ | |
507 | printf("\t.byte 0x%02x", table[i]->len); | |
508 | off += table[i]->len + 1; | |
509 | } else { | |
510 | /* "Big" symbols use two bytes. */ | |
511 | printf("\t.byte 0x%02x, 0x%02x", | |
512 | (table[i]->len & 0x7F) | 0x80, | |
513 | (table[i]->len >> 7) & 0x7F); | |
514 | off += table[i]->len + 2; | |
515 | } | |
8d605269 MY |
516 | for (k = 0; k < table[i]->len; k++) |
517 | printf(", 0x%02x", table[i]->sym[k]); | |
1da177e4 | 518 | printf("\n"); |
1da177e4 LT |
519 | } |
520 | printf("\n"); | |
521 | ||
522 | output_label("kallsyms_markers"); | |
b3dbb4ec | 523 | for (i = 0; i < ((table_cnt + 255) >> 8); i++) |
80ffbaa5 | 524 | printf("\t.long\t%u\n", markers[i]); |
1da177e4 LT |
525 | printf("\n"); |
526 | ||
527 | free(markers); | |
528 | ||
529 | output_label("kallsyms_token_table"); | |
530 | off = 0; | |
531 | for (i = 0; i < 256; i++) { | |
532 | best_idx[i] = off; | |
b3dbb4ec | 533 | expand_symbol(best_table[i], best_table_len[i], buf); |
1da177e4 LT |
534 | printf("\t.asciz\t\"%s\"\n", buf); |
535 | off += strlen(buf) + 1; | |
536 | } | |
537 | printf("\n"); | |
538 | ||
539 | output_label("kallsyms_token_index"); | |
540 | for (i = 0; i < 256; i++) | |
541 | printf("\t.short\t%d\n", best_idx[i]); | |
542 | printf("\n"); | |
543 | } | |
544 | ||
545 | ||
546 | /* table lookup compression functions */ | |
547 | ||
1da177e4 | 548 | /* count all the possible tokens in a symbol */ |
4bfe2b78 | 549 | static void learn_symbol(const unsigned char *symbol, int len) |
1da177e4 LT |
550 | { |
551 | int i; | |
552 | ||
553 | for (i = 0; i < len - 1; i++) | |
b3dbb4ec | 554 | token_profit[ symbol[i] + (symbol[i + 1] << 8) ]++; |
1da177e4 LT |
555 | } |
556 | ||
557 | /* decrease the count for all the possible tokens in a symbol */ | |
4bfe2b78 | 558 | static void forget_symbol(const unsigned char *symbol, int len) |
1da177e4 LT |
559 | { |
560 | int i; | |
561 | ||
562 | for (i = 0; i < len - 1; i++) | |
b3dbb4ec | 563 | token_profit[ symbol[i] + (symbol[i + 1] << 8) ]--; |
1da177e4 LT |
564 | } |
565 | ||
5e5c4fa7 | 566 | /* do the initial token count */ |
1da177e4 LT |
567 | static void build_initial_tok_table(void) |
568 | { | |
5e5c4fa7 | 569 | unsigned int i; |
1da177e4 | 570 | |
5e5c4fa7 | 571 | for (i = 0; i < table_cnt; i++) |
8d605269 | 572 | learn_symbol(table[i]->sym, table[i]->len); |
1da177e4 LT |
573 | } |
574 | ||
2558c138 | 575 | static unsigned char *find_token(unsigned char *str, int len, |
4bfe2b78 | 576 | const unsigned char *token) |
7c5d249a PM |
577 | { |
578 | int i; | |
579 | ||
580 | for (i = 0; i < len - 1; i++) { | |
581 | if (str[i] == token[0] && str[i+1] == token[1]) | |
582 | return &str[i]; | |
583 | } | |
584 | return NULL; | |
585 | } | |
586 | ||
1da177e4 LT |
587 | /* replace a given token in all the valid symbols. Use the sampled symbols |
588 | * to update the counts */ | |
4bfe2b78 | 589 | static void compress_symbols(const unsigned char *str, int idx) |
1da177e4 | 590 | { |
b3dbb4ec PM |
591 | unsigned int i, len, size; |
592 | unsigned char *p1, *p2; | |
1da177e4 | 593 | |
b3dbb4ec | 594 | for (i = 0; i < table_cnt; i++) { |
1da177e4 | 595 | |
8d605269 MY |
596 | len = table[i]->len; |
597 | p1 = table[i]->sym; | |
b3dbb4ec PM |
598 | |
599 | /* find the token on the symbol */ | |
7c5d249a | 600 | p2 = find_token(p1, len, str); |
b3dbb4ec PM |
601 | if (!p2) continue; |
602 | ||
603 | /* decrease the counts for this symbol's tokens */ | |
8d605269 | 604 | forget_symbol(table[i]->sym, len); |
b3dbb4ec PM |
605 | |
606 | size = len; | |
1da177e4 LT |
607 | |
608 | do { | |
b3dbb4ec PM |
609 | *p2 = idx; |
610 | p2++; | |
611 | size -= (p2 - p1); | |
612 | memmove(p2, p2 + 1, size); | |
613 | p1 = p2; | |
614 | len--; | |
615 | ||
616 | if (size < 2) break; | |
617 | ||
1da177e4 | 618 | /* find the token on the symbol */ |
7c5d249a | 619 | p2 = find_token(p1, size, str); |
1da177e4 | 620 | |
b3dbb4ec | 621 | } while (p2); |
1da177e4 | 622 | |
8d605269 | 623 | table[i]->len = len; |
1da177e4 | 624 | |
b3dbb4ec | 625 | /* increase the counts for this symbol's new tokens */ |
8d605269 | 626 | learn_symbol(table[i]->sym, len); |
1da177e4 LT |
627 | } |
628 | } | |
629 | ||
630 | /* search the token with the maximum profit */ | |
b3dbb4ec | 631 | static int find_best_token(void) |
1da177e4 | 632 | { |
b3dbb4ec | 633 | int i, best, bestprofit; |
1da177e4 LT |
634 | |
635 | bestprofit=-10000; | |
b3dbb4ec | 636 | best = 0; |
1da177e4 | 637 | |
b3dbb4ec PM |
638 | for (i = 0; i < 0x10000; i++) { |
639 | if (token_profit[i] > bestprofit) { | |
640 | best = i; | |
641 | bestprofit = token_profit[i]; | |
1da177e4 | 642 | } |
1da177e4 | 643 | } |
1da177e4 LT |
644 | return best; |
645 | } | |
646 | ||
647 | /* this is the core of the algorithm: calculate the "best" table */ | |
648 | static void optimize_result(void) | |
649 | { | |
b3dbb4ec | 650 | int i, best; |
1da177e4 LT |
651 | |
652 | /* using the '\0' symbol last allows compress_symbols to use standard | |
653 | * fast string functions */ | |
654 | for (i = 255; i >= 0; i--) { | |
655 | ||
656 | /* if this table slot is empty (it is not used by an actual | |
657 | * original char code */ | |
658 | if (!best_table_len[i]) { | |
659 | ||
cbf7a90e | 660 | /* find the token with the best profit value */ |
1da177e4 | 661 | best = find_best_token(); |
e0a04b11 XW |
662 | if (token_profit[best] == 0) |
663 | break; | |
1da177e4 LT |
664 | |
665 | /* place it in the "best" table */ | |
b3dbb4ec PM |
666 | best_table_len[i] = 2; |
667 | best_table[i][0] = best & 0xFF; | |
668 | best_table[i][1] = (best >> 8) & 0xFF; | |
1da177e4 LT |
669 | |
670 | /* replace this token in all the valid symbols */ | |
b3dbb4ec | 671 | compress_symbols(best_table[i], i); |
1da177e4 LT |
672 | } |
673 | } | |
674 | } | |
675 | ||
676 | /* start by placing the symbols that are actually used on the table */ | |
677 | static void insert_real_symbols_in_table(void) | |
678 | { | |
b3dbb4ec | 679 | unsigned int i, j, c; |
1da177e4 | 680 | |
b3dbb4ec | 681 | for (i = 0; i < table_cnt; i++) { |
8d605269 MY |
682 | for (j = 0; j < table[i]->len; j++) { |
683 | c = table[i]->sym[j]; | |
b3dbb4ec PM |
684 | best_table[c][0]=c; |
685 | best_table_len[c]=1; | |
1da177e4 LT |
686 | } |
687 | } | |
688 | } | |
689 | ||
690 | static void optimize_token_table(void) | |
691 | { | |
1da177e4 LT |
692 | build_initial_tok_table(); |
693 | ||
694 | insert_real_symbols_in_table(); | |
695 | ||
696 | optimize_result(); | |
697 | } | |
698 | ||
b478b782 LJ |
699 | /* guess for "linker script provide" symbol */ |
700 | static int may_be_linker_script_provide_symbol(const struct sym_entry *se) | |
701 | { | |
29e55ad3 | 702 | const char *symbol = sym_name(se); |
b478b782 LJ |
703 | int len = se->len - 1; |
704 | ||
705 | if (len < 8) | |
706 | return 0; | |
707 | ||
708 | if (symbol[0] != '_' || symbol[1] != '_') | |
709 | return 0; | |
710 | ||
711 | /* __start_XXXXX */ | |
712 | if (!memcmp(symbol + 2, "start_", 6)) | |
713 | return 1; | |
714 | ||
715 | /* __stop_XXXXX */ | |
716 | if (!memcmp(symbol + 2, "stop_", 5)) | |
717 | return 1; | |
718 | ||
719 | /* __end_XXXXX */ | |
720 | if (!memcmp(symbol + 2, "end_", 4)) | |
721 | return 1; | |
722 | ||
723 | /* __XXXXX_start */ | |
724 | if (!memcmp(symbol + len - 6, "_start", 6)) | |
725 | return 1; | |
726 | ||
727 | /* __XXXXX_end */ | |
728 | if (!memcmp(symbol + len - 4, "_end", 4)) | |
729 | return 1; | |
730 | ||
731 | return 0; | |
732 | } | |
733 | ||
f2df3f65 PM |
734 | static int compare_symbols(const void *a, const void *b) |
735 | { | |
8d605269 MY |
736 | const struct sym_entry *sa = *(const struct sym_entry **)a; |
737 | const struct sym_entry *sb = *(const struct sym_entry **)b; | |
f2df3f65 PM |
738 | int wa, wb; |
739 | ||
f2df3f65 PM |
740 | /* sort by address first */ |
741 | if (sa->addr > sb->addr) | |
742 | return 1; | |
743 | if (sa->addr < sb->addr) | |
744 | return -1; | |
745 | ||
746 | /* sort by "weakness" type */ | |
747 | wa = (sa->sym[0] == 'w') || (sa->sym[0] == 'W'); | |
748 | wb = (sb->sym[0] == 'w') || (sb->sym[0] == 'W'); | |
749 | if (wa != wb) | |
750 | return wa - wb; | |
751 | ||
b478b782 LJ |
752 | /* sort by "linker script provide" type */ |
753 | wa = may_be_linker_script_provide_symbol(sa); | |
754 | wb = may_be_linker_script_provide_symbol(sb); | |
755 | if (wa != wb) | |
756 | return wa - wb; | |
757 | ||
758 | /* sort by the number of prefix underscores */ | |
aa915245 MY |
759 | wa = strspn(sym_name(sa), "_"); |
760 | wb = strspn(sym_name(sb), "_"); | |
b478b782 LJ |
761 | if (wa != wb) |
762 | return wa - wb; | |
763 | ||
f2df3f65 PM |
764 | /* sort by initial order, so that other symbols are left undisturbed */ |
765 | return sa->start_pos - sb->start_pos; | |
766 | } | |
767 | ||
768 | static void sort_symbols(void) | |
769 | { | |
8d605269 | 770 | qsort(table, table_cnt, sizeof(table[0]), compare_symbols); |
f2df3f65 | 771 | } |
1da177e4 | 772 | |
c6bda7c9 RR |
773 | static void make_percpus_absolute(void) |
774 | { | |
775 | unsigned int i; | |
776 | ||
777 | for (i = 0; i < table_cnt; i++) | |
8d605269 | 778 | if (symbol_in_range(table[i], &percpu_range, 1)) { |
8c996940 AB |
779 | /* |
780 | * Keep the 'A' override for percpu symbols to | |
781 | * ensure consistent behavior compared to older | |
782 | * versions of this tool. | |
783 | */ | |
8d605269 MY |
784 | table[i]->sym[0] = 'A'; |
785 | table[i]->percpu_absolute = 1; | |
8c996940 | 786 | } |
c6bda7c9 RR |
787 | } |
788 | ||
2213e9a6 AB |
789 | /* find the minimum non-absolute symbol address */ |
790 | static void record_relative_base(void) | |
791 | { | |
792 | unsigned int i; | |
793 | ||
2213e9a6 | 794 | for (i = 0; i < table_cnt; i++) |
8d605269 | 795 | if (!symbol_absolute(table[i])) { |
f34ea029 MY |
796 | /* |
797 | * The table is sorted by address. | |
798 | * Take the first non-absolute symbol value. | |
799 | */ | |
8d605269 | 800 | relative_base = table[i]->addr; |
f34ea029 MY |
801 | return; |
802 | } | |
2213e9a6 AB |
803 | } |
804 | ||
b3dbb4ec | 805 | int main(int argc, char **argv) |
1da177e4 | 806 | { |
41f11a4f YS |
807 | if (argc >= 2) { |
808 | int i; | |
809 | for (i = 1; i < argc; i++) { | |
810 | if(strcmp(argv[i], "--all-symbols") == 0) | |
811 | all_symbols = 1; | |
c6bda7c9 RR |
812 | else if (strcmp(argv[i], "--absolute-percpu") == 0) |
813 | absolute_percpu = 1; | |
534c9f2e | 814 | else if (strcmp(argv[i], "--base-relative") == 0) |
2213e9a6 AB |
815 | base_relative = 1; |
816 | else | |
41f11a4f YS |
817 | usage(); |
818 | } | |
819 | } else if (argc != 1) | |
1da177e4 LT |
820 | usage(); |
821 | ||
822 | read_map(stdin); | |
5e5c4fa7 | 823 | shrink_table(); |
c6bda7c9 RR |
824 | if (absolute_percpu) |
825 | make_percpus_absolute(); | |
f34ea029 | 826 | sort_symbols(); |
2213e9a6 AB |
827 | if (base_relative) |
828 | record_relative_base(); | |
2ea03891 | 829 | optimize_token_table(); |
1da177e4 LT |
830 | write_src(); |
831 | ||
832 | return 0; | |
833 | } |