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
28 #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
;
37 unsigned int percpu_absolute
;
41 const char *start_sym
, *end_sym
;
42 unsigned long long start
, end
;
45 static unsigned long long _text
;
46 static unsigned long long relative_base
;
47 static struct addr_range text_ranges
[] = {
48 { "_stext", "_etext" },
49 { "_sinittext", "_einittext" },
51 #define text_range_text (&text_ranges[0])
52 #define text_range_inittext (&text_ranges[1])
54 static struct addr_range percpu_range
= {
55 "__per_cpu_start", "__per_cpu_end", -1ULL, 0
58 static struct sym_entry
*table
;
59 static unsigned int table_size
, table_cnt
;
60 static int all_symbols
;
61 static int absolute_percpu
;
62 static int base_relative
;
64 static int token_profit
[0x10000];
66 /* the table that holds the result of the compression */
67 static unsigned char best_table
[256][2];
68 static unsigned char best_table_len
[256];
71 static void usage(void)
73 fprintf(stderr
, "Usage: kallsyms [--all-symbols] "
74 "[--base-relative] < in.map > out.S\n");
78 static char *sym_name(const struct sym_entry
*s
)
80 return (char *)s
->sym
+ 1;
83 static bool is_ignored_symbol(const char *name
, char type
)
85 static const char * const ignored_symbols
[] = {
87 * Symbols which vary between passes. Passes 1 and 2 must have
88 * identical symbol lists. The kallsyms_* symbols below are
89 * only added after pass 1, they would be included in pass 2
90 * when --all-symbols is specified so exclude them to get a
95 "kallsyms_relative_base",
99 "kallsyms_token_table",
100 "kallsyms_token_index",
101 /* Exclude linker generated symbols which vary between passes */
102 "_SDA_BASE_", /* ppc */
103 "_SDA2_BASE_", /* ppc */
107 static const char * const ignored_prefixes
[] = {
108 "$", /* local symbols for ARM, MIPS, etc. */
109 ".LASANPC", /* s390 kasan local symbols */
110 "__crc_", /* modversions */
111 "__efistub_", /* arm64 EFI stub namespace */
115 static const char * const ignored_suffixes
[] = {
116 "_from_arm", /* arm */
117 "_from_thumb", /* arm */
122 const char * const *p
;
124 /* Exclude symbols which vary between passes. */
125 for (p
= ignored_symbols
; *p
; p
++)
126 if (!strcmp(name
, *p
))
129 for (p
= ignored_prefixes
; *p
; p
++)
130 if (!strncmp(name
, *p
, strlen(*p
)))
133 for (p
= ignored_suffixes
; *p
; p
++) {
134 int l
= strlen(name
) - strlen(*p
);
136 if (l
>= 0 && !strcmp(name
+ l
, *p
))
140 if (type
== 'U' || type
== 'u')
142 /* exclude debugging symbols */
143 if (type
== 'N' || type
== 'n')
146 if (toupper(type
) == 'A') {
147 /* Keep these useful absolute symbols */
148 if (strcmp(name
, "__kernel_syscall_via_break") &&
149 strcmp(name
, "__kernel_syscall_via_epc") &&
150 strcmp(name
, "__kernel_sigtramp") &&
151 strcmp(name
, "__gp"))
158 static void check_symbol_range(const char *sym
, unsigned long long addr
,
159 struct addr_range
*ranges
, int entries
)
162 struct addr_range
*ar
;
164 for (i
= 0; i
< entries
; ++i
) {
167 if (strcmp(sym
, ar
->start_sym
) == 0) {
170 } else if (strcmp(sym
, ar
->end_sym
) == 0) {
177 static int read_symbol(FILE *in
, struct sym_entry
*s
)
179 char sym
[500], stype
;
182 rc
= fscanf(in
, "%llx %c %499s\n", &s
->addr
, &stype
, sym
);
184 if (rc
!= EOF
&& fgets(sym
, 500, in
) == NULL
)
185 fprintf(stderr
, "Read error or end of file.\n");
188 if (strlen(sym
) >= KSYM_NAME_LEN
) {
189 fprintf(stderr
, "Symbol %s too long for kallsyms (%zu >= %d).\n"
190 "Please increase KSYM_NAME_LEN both in kernel and kallsyms.c\n",
191 sym
, strlen(sym
), KSYM_NAME_LEN
);
195 if (is_ignored_symbol(sym
, stype
))
198 /* Ignore most absolute/undefined (?) symbols. */
199 if (strcmp(sym
, "_text") == 0)
202 check_symbol_range(sym
, s
->addr
, text_ranges
, ARRAY_SIZE(text_ranges
));
203 check_symbol_range(sym
, s
->addr
, &percpu_range
, 1);
205 /* include the type field in the symbol name, so that it gets
206 * compressed together */
207 s
->len
= strlen(sym
) + 1;
208 s
->sym
= malloc(s
->len
+ 1);
210 fprintf(stderr
, "kallsyms failure: "
211 "unable to allocate required amount of memory\n");
214 strcpy(sym_name(s
), sym
);
217 s
->percpu_absolute
= 0;
222 static int symbol_in_range(const struct sym_entry
*s
,
223 const struct addr_range
*ranges
, int entries
)
226 const struct addr_range
*ar
;
228 for (i
= 0; i
< entries
; ++i
) {
231 if (s
->addr
>= ar
->start
&& s
->addr
<= ar
->end
)
238 static int symbol_valid(const struct sym_entry
*s
)
240 const char *name
= sym_name(s
);
242 /* if --all-symbols is not specified, then symbols outside the text
243 * and inittext sections are discarded */
245 if (symbol_in_range(s
, text_ranges
,
246 ARRAY_SIZE(text_ranges
)) == 0)
248 /* Corner case. Discard any symbols with the same value as
249 * _etext _einittext; they can move between pass 1 and 2 when
250 * the kallsyms data are added. If these symbols move then
251 * they may get dropped in pass 2, which breaks the kallsyms
254 if ((s
->addr
== text_range_text
->end
&&
255 strcmp(name
, text_range_text
->end_sym
)) ||
256 (s
->addr
== text_range_inittext
->end
&&
257 strcmp(name
, text_range_inittext
->end_sym
)))
264 /* remove all the invalid symbols from the table */
265 static void shrink_table(void)
270 for (i
= 0; i
< table_cnt
; i
++) {
271 if (symbol_valid(&table
[i
])) {
273 table
[pos
] = table
[i
];
281 /* When valid symbol is not registered, exit to error */
283 fprintf(stderr
, "No valid symbol.\n");
288 static void read_map(FILE *in
)
291 if (table_cnt
>= table_size
) {
293 table
= realloc(table
, sizeof(*table
) * table_size
);
295 fprintf(stderr
, "out of memory\n");
299 if (read_symbol(in
, &table
[table_cnt
]) == 0) {
300 table
[table_cnt
].start_pos
= table_cnt
;
306 static void output_label(const char *label
)
308 printf(".globl %s\n", label
);
310 printf("%s:\n", label
);
313 /* uncompress a compressed symbol. When this function is called, the best table
314 * might still be compressed itself, so the function needs to be recursive */
315 static int expand_symbol(const unsigned char *data
, int len
, char *result
)
317 int c
, rlen
, total
=0;
321 /* if the table holds a single char that is the same as the one
322 * we are looking for, then end the search */
323 if (best_table
[c
][0]==c
&& best_table_len
[c
]==1) {
327 /* if not, recurse and expand */
328 rlen
= expand_symbol(best_table
[c
], best_table_len
[c
], result
);
340 static int symbol_absolute(const struct sym_entry
*s
)
342 return s
->percpu_absolute
;
345 static void write_src(void)
347 unsigned int i
, k
, off
;
348 unsigned int best_idx
[256];
349 unsigned int *markers
;
350 char buf
[KSYM_NAME_LEN
];
352 printf("#include <asm/bitsperlong.h>\n");
353 printf("#if BITS_PER_LONG == 64\n");
354 printf("#define PTR .quad\n");
355 printf("#define ALGN .balign 8\n");
357 printf("#define PTR .long\n");
358 printf("#define ALGN .balign 4\n");
361 printf("\t.section .rodata, \"a\"\n");
363 /* Provide proper symbols relocatability by their relativeness
364 * to a fixed anchor point in the runtime image, either '_text'
365 * for absolute address tables, in which case the linker will
366 * emit the final addresses at build time. Otherwise, use the
367 * offset relative to the lowest value encountered of all relative
368 * symbols, and emit non-relocatable fixed offsets that will be fixed
371 * The symbol names cannot be used to construct normal symbol
372 * references as the list of symbols contains symbols that are
373 * declared static and are private to their .o files. This prevents
374 * .tmp_kallsyms.o or any other object from referencing them.
377 output_label("kallsyms_addresses");
379 output_label("kallsyms_offsets");
381 for (i
= 0; i
< table_cnt
; i
++) {
386 if (!absolute_percpu
) {
387 offset
= table
[i
].addr
- relative_base
;
388 overflow
= (offset
< 0 || offset
> UINT_MAX
);
389 } else if (symbol_absolute(&table
[i
])) {
390 offset
= table
[i
].addr
;
391 overflow
= (offset
< 0 || offset
> INT_MAX
);
393 offset
= relative_base
- table
[i
].addr
- 1;
394 overflow
= (offset
< INT_MIN
|| offset
>= 0);
397 fprintf(stderr
, "kallsyms failure: "
398 "%s symbol value %#llx out of range in relative mode\n",
399 symbol_absolute(&table
[i
]) ? "absolute" : "relative",
403 printf("\t.long\t%#x\n", (int)offset
);
404 } else if (!symbol_absolute(&table
[i
])) {
405 if (_text
<= table
[i
].addr
)
406 printf("\tPTR\t_text + %#llx\n",
407 table
[i
].addr
- _text
);
409 printf("\tPTR\t_text - %#llx\n",
410 _text
- table
[i
].addr
);
412 printf("\tPTR\t%#llx\n", table
[i
].addr
);
418 output_label("kallsyms_relative_base");
419 printf("\tPTR\t_text - %#llx\n", _text
- relative_base
);
423 output_label("kallsyms_num_syms");
424 printf("\t.long\t%u\n", table_cnt
);
427 /* table of offset markers, that give the offset in the compressed stream
428 * every 256 symbols */
429 markers
= malloc(sizeof(unsigned int) * ((table_cnt
+ 255) / 256));
431 fprintf(stderr
, "kallsyms failure: "
432 "unable to allocate required memory\n");
436 output_label("kallsyms_names");
438 for (i
= 0; i
< table_cnt
; i
++) {
440 markers
[i
>> 8] = off
;
442 printf("\t.byte 0x%02x", table
[i
].len
);
443 for (k
= 0; k
< table
[i
].len
; k
++)
444 printf(", 0x%02x", table
[i
].sym
[k
]);
447 off
+= table
[i
].len
+ 1;
451 output_label("kallsyms_markers");
452 for (i
= 0; i
< ((table_cnt
+ 255) >> 8); i
++)
453 printf("\t.long\t%u\n", markers
[i
]);
458 output_label("kallsyms_token_table");
460 for (i
= 0; i
< 256; i
++) {
462 expand_symbol(best_table
[i
], best_table_len
[i
], buf
);
463 printf("\t.asciz\t\"%s\"\n", buf
);
464 off
+= strlen(buf
) + 1;
468 output_label("kallsyms_token_index");
469 for (i
= 0; i
< 256; i
++)
470 printf("\t.short\t%d\n", best_idx
[i
]);
475 /* table lookup compression functions */
477 /* count all the possible tokens in a symbol */
478 static void learn_symbol(const unsigned char *symbol
, int len
)
482 for (i
= 0; i
< len
- 1; i
++)
483 token_profit
[ symbol
[i
] + (symbol
[i
+ 1] << 8) ]++;
486 /* decrease the count for all the possible tokens in a symbol */
487 static void forget_symbol(const unsigned char *symbol
, int len
)
491 for (i
= 0; i
< len
- 1; i
++)
492 token_profit
[ symbol
[i
] + (symbol
[i
+ 1] << 8) ]--;
495 /* do the initial token count */
496 static void build_initial_tok_table(void)
500 for (i
= 0; i
< table_cnt
; i
++)
501 learn_symbol(table
[i
].sym
, table
[i
].len
);
504 static unsigned char *find_token(unsigned char *str
, int len
,
505 const unsigned char *token
)
509 for (i
= 0; i
< len
- 1; i
++) {
510 if (str
[i
] == token
[0] && str
[i
+1] == token
[1])
516 /* replace a given token in all the valid symbols. Use the sampled symbols
517 * to update the counts */
518 static void compress_symbols(const unsigned char *str
, int idx
)
520 unsigned int i
, len
, size
;
521 unsigned char *p1
, *p2
;
523 for (i
= 0; i
< table_cnt
; i
++) {
528 /* find the token on the symbol */
529 p2
= find_token(p1
, len
, str
);
532 /* decrease the counts for this symbol's tokens */
533 forget_symbol(table
[i
].sym
, len
);
541 memmove(p2
, p2
+ 1, size
);
547 /* find the token on the symbol */
548 p2
= find_token(p1
, size
, str
);
554 /* increase the counts for this symbol's new tokens */
555 learn_symbol(table
[i
].sym
, len
);
559 /* search the token with the maximum profit */
560 static int find_best_token(void)
562 int i
, best
, bestprofit
;
567 for (i
= 0; i
< 0x10000; i
++) {
568 if (token_profit
[i
] > bestprofit
) {
570 bestprofit
= token_profit
[i
];
576 /* this is the core of the algorithm: calculate the "best" table */
577 static void optimize_result(void)
581 /* using the '\0' symbol last allows compress_symbols to use standard
582 * fast string functions */
583 for (i
= 255; i
>= 0; i
--) {
585 /* if this table slot is empty (it is not used by an actual
586 * original char code */
587 if (!best_table_len
[i
]) {
589 /* find the token with the best profit value */
590 best
= find_best_token();
591 if (token_profit
[best
] == 0)
594 /* place it in the "best" table */
595 best_table_len
[i
] = 2;
596 best_table
[i
][0] = best
& 0xFF;
597 best_table
[i
][1] = (best
>> 8) & 0xFF;
599 /* replace this token in all the valid symbols */
600 compress_symbols(best_table
[i
], i
);
605 /* start by placing the symbols that are actually used on the table */
606 static void insert_real_symbols_in_table(void)
608 unsigned int i
, j
, c
;
610 for (i
= 0; i
< table_cnt
; i
++) {
611 for (j
= 0; j
< table
[i
].len
; j
++) {
619 static void optimize_token_table(void)
621 build_initial_tok_table();
623 insert_real_symbols_in_table();
628 /* guess for "linker script provide" symbol */
629 static int may_be_linker_script_provide_symbol(const struct sym_entry
*se
)
631 const char *symbol
= sym_name(se
);
632 int len
= se
->len
- 1;
637 if (symbol
[0] != '_' || symbol
[1] != '_')
641 if (!memcmp(symbol
+ 2, "start_", 6))
645 if (!memcmp(symbol
+ 2, "stop_", 5))
649 if (!memcmp(symbol
+ 2, "end_", 4))
653 if (!memcmp(symbol
+ len
- 6, "_start", 6))
657 if (!memcmp(symbol
+ len
- 4, "_end", 4))
663 static int compare_symbols(const void *a
, const void *b
)
665 const struct sym_entry
*sa
;
666 const struct sym_entry
*sb
;
672 /* sort by address first */
673 if (sa
->addr
> sb
->addr
)
675 if (sa
->addr
< sb
->addr
)
678 /* sort by "weakness" type */
679 wa
= (sa
->sym
[0] == 'w') || (sa
->sym
[0] == 'W');
680 wb
= (sb
->sym
[0] == 'w') || (sb
->sym
[0] == 'W');
684 /* sort by "linker script provide" type */
685 wa
= may_be_linker_script_provide_symbol(sa
);
686 wb
= may_be_linker_script_provide_symbol(sb
);
690 /* sort by the number of prefix underscores */
691 wa
= strspn(sym_name(sa
), "_");
692 wb
= strspn(sym_name(sb
), "_");
696 /* sort by initial order, so that other symbols are left undisturbed */
697 return sa
->start_pos
- sb
->start_pos
;
700 static void sort_symbols(void)
702 qsort(table
, table_cnt
, sizeof(struct sym_entry
), compare_symbols
);
705 static void make_percpus_absolute(void)
709 for (i
= 0; i
< table_cnt
; i
++)
710 if (symbol_in_range(&table
[i
], &percpu_range
, 1)) {
712 * Keep the 'A' override for percpu symbols to
713 * ensure consistent behavior compared to older
714 * versions of this tool.
716 table
[i
].sym
[0] = 'A';
717 table
[i
].percpu_absolute
= 1;
721 /* find the minimum non-absolute symbol address */
722 static void record_relative_base(void)
726 for (i
= 0; i
< table_cnt
; i
++)
727 if (!symbol_absolute(&table
[i
])) {
729 * The table is sorted by address.
730 * Take the first non-absolute symbol value.
732 relative_base
= table
[i
].addr
;
737 int main(int argc
, char **argv
)
741 for (i
= 1; i
< argc
; i
++) {
742 if(strcmp(argv
[i
], "--all-symbols") == 0)
744 else if (strcmp(argv
[i
], "--absolute-percpu") == 0)
746 else if (strcmp(argv
[i
], "--base-relative") == 0)
751 } else if (argc
!= 1)
757 make_percpus_absolute();
760 record_relative_base();
761 optimize_token_table();